MASM611 - Programmer's Guide

Programmer’s Guide
Microsoft MASM ®
Assembly-Language Development System

Version 6.1
For MS-DOS ® and Windows ™ Operating Systems
Microsoft Corporation
Filename: LMAPGTTL.DOC Project:

Template: FRONTA1.DOT Author: Bart Simpson, Who the Hell Are You? Last Saved By: Mike Eddy
Revision #: 16 Page: 1 of 1 Printed: 10/02/00 04:19 PM
Information in this document is subject to change without notice. Companies, names, and data used in
examples herein are fictitious unless otherwise noted. No part of this document maybe reproduced or
transmitted in any form or by any means, electronic or mechanical, for any purpose, without the
express written permission of Microsoft Corporation.
©1992 Microsoft Corporation. All rights reserved.
Microsoft, MS, MS-DOS, XENIX, CodeView, and QuickC are registered trademarks and Microsoft
QuickBasic, QuickPascal, Windows and Windows NT are trademarks of Microsoft Corporation in
the USA and other countries.
U.S. Patent No. 4,955,066
Hercules is a registered trademark of Hercules Computer Technology.

IBM, PS/2, and OS/2 are registered trademarks of International Business Machines Corporation.
Intel is a registered trademark of Intel Corporation.
NEC and V25 are registered trademarks and V35 is a trademark of NEC Corporation.
Document No. DB35747-1292

Printed in the United States of America.
Filename: LMAPGCPY.DOC Project:

Template: FRONTA1.DOT Author: Ruth L Silverio Last Saved By: Mike Eddy
iii
Contents
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
New and Extended Features in MASM 6.1 . . . . . . . . . . . . . . . . . . . . . . . . xiii
MASM Features New Since Version 5.1 . . . . . . . . . . . . . . . . . . . . . . . . xiv
MASM Features New Since Version 6.0 . . . . . . . . . . . . . . . . . . . . . . . . xv
ML and MASM Command Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi
Compatibility with Earlier Versions of MASM . . . . . . . . . . . . . . . . . . . . xvi
A Word About Instruction Timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Books for Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii
Document Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
Getting Assistance and Reporting Problems . . . . . . . . . . . . . . . . . . . . . . . . xx
Chapter 1 Understanding Global Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

The Processing Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
8086-Based Processors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Operating Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Segmented Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Segment Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Segmented Addressing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Segment Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Language Components of MASM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Reserved Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Identifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Predefined Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Integer Constants and Constant Expressions . . . . . . . . . . . . . . . . . . . . . 11
Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Data Types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
The Assembly Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Generating and Running Executable Programs . . . . . . . . . . . . . . . . . . . . 23
Using the OPTION Directive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Conditional Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Chapter 2 Organizing Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Physical Memory Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Logical Segments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Using Simplified Segment Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Filename: LMAPGTOC.DOC Project:

Template: FRONTA1.DOT Author: Don Hayward Last Saved By: Ruth L Silverio
iv Contents
Defining Basic Attributes with .MODEL . . . . . . . . . . . . . . . . . . . . . . . . 34

Specifying a Processor and Coprocessor . . . . . . . . . . . . . . . . . . . . . . . . 38
Creating a Stack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Creating Data Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Creating Code Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Starting and Ending Code with .STARTUP and .EXIT . . . . . . . . . . . . . . 41
Using Full Segment Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Defining Segments with the SEGMENT Directive. . . . . . . . . . . . . . . . . . 44
Controlling the Segment Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Setting the ASSUME Directive for Segment Registers . . . . . . . . . . . . . . . 49
Defining Segment Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Chapter 3 Using Addresses and Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Programming Segmented Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Initializing Default Segment Registers. . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Near and Far Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Operands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Register Operands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Immediate Operands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Direct Memory Operands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Indirect Memory Operands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
The Program Stack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Saving Operands on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Saving Flags on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Saving Registers on the Stack (80186–80486 Only). . . . . . . . . . . . . . . . . 74
Accessing Data with Pointers and Addresses . . . . . . . . . . . . . . . . . . . . . . . . 74
Defining Pointer Types with TYPEDEF . . . . . . . . . . . . . . . . . . . . . . . . 75
Defining Register Types with ASSUME . . . . . . . . . . . . . . . . . . . . . . . . . 77
Basic Pointer and Address Operations . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Chapter 4 Defining and Using Simple Data Types . . . . . . . . . . . . . . . . . . . . . . . 85

Declaring Integer Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Allocating Memory for Integer Variables . . . . . . . . . . . . . . . . . . . . . . . . 85
Data Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Working with Simple Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Copying Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Adding and Subtracting Integers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Multiplying and Dividing Integers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Manipulating Numbers at the Bit Level. . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Logical Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Shifting and Rotating Bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Contents v
Multiplying and Dividing with Shift Instructions . . . . . . . . . . . . . . . . . . 102
Chapter 5 Defining and Using Complex Data Types. . . . . . . . . . . . . . . . . . . . . 105

Arrays and Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Declaring and Referencing Arrays. . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Declaring and Initializing Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Processing Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Structures and Unions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Declaring Structure and Union Types . . . . . . . . . . . . . . . . . . . . . . . . . 118
Defining Structure and Union Variables. . . . . . . . . . . . . . . . . . . . . . . . 121
Referencing Structures, Unions, and Fields . . . . . . . . . . . . . . . . . . . . . 126
Nested Structures and Unions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Declaring Record Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Defining Record Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Record Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Chapter 6 Using Floating-Point and Binary Coded Decimal Numbers . . . . . . . 135

Using Floating-Point Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Declaring Floating-Point Variables and Constants . . . . . . . . . . . . . . . . . 136
Storing Numbers in Floating-Point Format. . . . . . . . . . . . . . . . . . . . . . 138
Using a Math Coprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Coprocessor Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Instruction and Operand Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Coordinating Memory Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Using Coprocessor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Using An Emulator Library. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Using Binary Coded Decimal Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Defining BCD Constants and Variables . . . . . . . . . . . . . . . . . . . . . . . . 157
BCD Calculations on a Coprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . 157
BCD Calculations on the Main Processor . . . . . . . . . . . . . . . . . . . . . . 158
Chapter 7 Controlling Program Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Jumps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Unconditional Jumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Conditional Jumps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Loop-Generating Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Writing Loop Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Defining Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

vi Contents
Passing Arguments on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

Declaring Parameters with the PROC Directive . . . . . . . . . . . . . . . . . . 184
Using Local Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Creating Local Variables Automatically . . . . . . . . . . . . . . . . . . . . . . . . 190
Declaring Procedure Prototypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Calling Procedures with INVOKE . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Generating Prologue and Epilogue Code. . . . . . . . . . . . . . . . . . . . . . . . 198
MS-DOS Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Calling MS-DOS and ROM-BIOS Interrupts . . . . . . . . . . . . . . . . . . . . 204
Replacing an Interrupt Routine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Chapter 8 Sharing Data and Procedures Among Modules and Libraries . . . . . 211
Selecting Data-Sharing Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Sharing Symbols with Include Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Organizing Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Declaring Symbols Public and External . . . . . . . . . . . . . . . . . . . . . . . . 214
Positioning External Declarations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Using Alternatives to Include Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
PUBLIC and EXTERN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Other Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Developing Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Associating Libraries with Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Using EXTERN with Library Routines . . . . . . . . . . . . . . . . . . . . . . . . 223
Chapter 9 Using Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Text Macros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Macro Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Creating Macro Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Passing Arguments to Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Specifying Required and Default Parameters . . . . . . . . . . . . . . . . . . . . 229
Defining Local Symbols in Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Assembly-Time Variables and Macro Operators . . . . . . . . . . . . . . . . . . . . 233
Text Delimiters and the Literal-Character Operator . . . . . . . . . . . . . . . . 234
Expansion Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Substitution Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Defining Repeat Blocks with Loop Directives . . . . . . . . . . . . . . . . . . . . . . 239
REPEAT Loops. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
WHILE Loops. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
FOR Loops and Variable-Length Parameters . . . . . . . . . . . . . . . . . . . . 242
FORC Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
String Directives and Predefined Functions . . . . . . . . . . . . . . . . . . . . . . . . 245

Contents vii
Returning Values with Macro Functions. . . . . . . . . . . . . . . . . . . . . . . . . . 248

Returning Values with EXITM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Using Macro Functions with Variable-Length Parameter Lists . . . . . . . . 249
Expansion Operator in Macro Functions . . . . . . . . . . . . . . . . . . . . . . . 251
Advanced Macro Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Defining Macros within Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Testing for Argument Type and Environment . . . . . . . . . . . . . . . . . . . 252
Using Recursive Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Chapter 10 Writing a Dynamic-Link Library For Windows . . . . . . . . . . . . . . . . 257

Overview of DLLs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Loading a DLL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Building a DLL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
DLL Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
DLL Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
DLL Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
DLL Extension Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Example of a DLL: SYSINFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Entry Routine for SYSINFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Expanding SYSINFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Chapter 11 Writing Memory-Resident Software . . . . . . . . . . . . . . . . . . . . . . . . 273

Terminate-and-Stay-Resident Programs. . . . . . . . . . . . . . . . . . . . . . . . . . 273
Structure of a TSR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Passive TSRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Active TSRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Interrupt Handlers in Active TSRs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Auditing Hardware Events for TSR Requests . . . . . . . . . . . . . . . . . . . 275
Monitoring System Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Determining Whether to Invoke the TSR . . . . . . . . . . . . . . . . . . . . . . 279
Example of a Simple TSR: ALARM . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Using MS-DOS in Active TSRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Understanding MS-DOS Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Determining MS-DOS Activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Interrupting MS-DOS Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Monitoring the Critical Error Flag . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Preventing Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Trapping Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Preserving an Existing Condition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Preserving Existing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290

viii Contents
Communicating Through the Multiplex Interrupt . . . . . . . . . . . . . . . . . . . . 290

The Multiplex Handler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Using the Multiplex Interrupt Under MS-DOS Version 2.x. . . . . . . . . . . 292
Deinstalling a TSR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
Example of an Advanced TSR: SNAP . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Building SNAP.EXE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
Outline of SNAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Chapter 12 Mixed-Language Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Naming and Calling Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
The C Calling Convention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
The Pascal Calling Convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
The STDCALL and SYSCALL Calling Conventions. . . . . . . . . . . . . . . 311
Writing an Assembly Procedure For a Mixed-Language Program . . . . . . . . 312
The MASM/High-Level–Language Interface . . . . . . . . . . . . . . . . . . . . . . . 313
The C/MASM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
The C++/MASM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
The FORTRAN/MASM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
The Basic/MASM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Chapter 13 Writing 32-Bit Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

32-Bit Memory Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
MASM Directives for 32-Bit Programming . . . . . . . . . . . . . . . . . . . . . . . . 336
Sample Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
Appendixes
Appendix A Differences Between MASM 6.1 and 5.1. . . . . . . . . . . . . . . . . . . . 341
New Features of Version 6.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
The Assembler, Environment, and Utilities . . . . . . . . . . . . . . . . . . . . . . 342
Segment Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Data Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
Procedures, Loops, and Jumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Simplifying Multiple-Module Projects . . . . . . . . . . . . . . . . . . . . . . . . . 348
Expanded State Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
New Processor Instructions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Renamed Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Macro Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
MASM 6.1 Programming Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Compatibility Between MASM 5.1 and 6.1. . . . . . . . . . . . . . . . . . . . . . . . 352

Contents ix
Rewriting Code for Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

Using the OPTION Directive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
Changes to Instruction Encodings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Appendix B BNF Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
Appendix Generating and Reading Assembly Listings. . . . . . . . . . . . . . . . . . 397

Generating Listing Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Precedence of Command-Line Options and Listing Directives. . . . . . . . 399
Reading the Listing File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Generated Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
Symbols and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
Reading Tables in a Listing File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
Appendix D MASM Reserved Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

Operands and Symbols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Special Operands for the 80386/486 . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Predefined Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Operators and Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Processor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
8086/8088 Processor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
80186 Processor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
80286 and 80386 Privileged-Mode Instructions . . . . . . . . . . . . . . . . . . 413
Instruction Prefixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Coprocessor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
8087 Coprocessor Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
80287 Privileged-Mode Instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
80387 Instructions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Appendix E Default Segment Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

x Contents
Figures and Tables

Figures
1.1 Segment Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Calculating Physical Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Registers for 8088-80286 Processors . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4 Extended Registers for the 80386/486 Processors . . . . . . . . . . . . . . . . . 18
1.5 Flags for 8088-80486 Processors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1 Stack Status Before and After Pushes and Pops . . . . . . . . . . . . . . . . . . 72
4.1 Integer Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.2 Shifts and Rotates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.1 Encoding for Real Numbers in IEEE Format . . . . . . . . . . . . . . . . . . . 138
6.2 Coprocessor Data Registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.3 Status of the Register Stack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.4 Status of the Register Stack and Memory Locations . . . . . . . . . . . . . . 143
6.5 Status of the Previously Initialized Register Stack . . . . . . . . . . . . . . . . 144
6.6 Status of the Already Initialized Register Stack . . . . . . . . . . . . . . . . . . 144
6.7 Status of the Register Stack: Main Memory and Coprocessor. . . . . . . . 148
6.8 Coprocessor Control Registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.9 Coprocessor and Processor Control Flags . . . . . . . . . . . . . . . . . . . . . . 155
7.1 Program Arguments on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
7.2 Local Variables on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
7.3 Operation of Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
8.1 Using EXTERNDEF for Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . 215
8.2 Using PROTO and INVOKE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
8.3 Using PUBLIC and EXTERN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.1 Time Line of Interaction Between Interrupt Handlers
for a Typical TSR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
11.2 Flowchart for SNAP.EXE: Installation Phase . . . . . . . . . . . . . . . . . . 296
11.3 Flowchart for SNAP.EXE Resident Phase . . . . . . . . . . . . . . . . . . . . 297
11.4 Flowchart for SNAP.EXE Deinstallation Phase. . . . . . . . . . . . . . . . . 298
12.1 C String Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
12.2 C Stack Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
12.3 FORTRAN String Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
12.4 FORTRAN Stack Frame. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
12.5 Basic String Descriptor Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
12.6 Basic Stack Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
B.1 BNF Definition of the TYPEDEF Directive . . . . . . . . . . . . . . . . . . . . 380

Contents xi
Tables
1.1 8086 Family of Processors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 The MS-DOS and Windows Operating Systems Compared . . . . . . . . . . 4
1.3 Operator Precedence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1 Attributes of Memory Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1 Indirect Addressing with 16-Bit Registers . . . . . . . . . . . . . . . . . . . . . . 68
4.1 Division Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.1 Requirements for String Instructions . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.1 Ranges of Floating-Point Variables . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.2 Coprocessor Operand Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
6.3 Control-Flag Settings After Comparison or Test . . . . . . . . . . . . . . . . . 151
7.1 Conditional Jumps Based on Comparisons of Two Values . . . . . . . . . 167
9.1 MASM Macro Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
11.1 MS-DOS Internal Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
12.1 Naming and Calling Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.2 Register Conventions for Simple Return Values . . . . . . . . . . . . . . . . 317
A.1 Requirements for String Instructions. . . . . . . . . . . . . . . . . . . . . . . . . 353
C.1 Options for Generating or Modifying Listing Files . . . . . . . . . . . . . . . 398
C.2 Symbols and Abbreviations in Listings . . . . . . . . . . . . . . . . . . . . . . . 400
C.3 Symbols in Timing Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401

xiii
Introduction
The Microsoft® Macro Assembler Programmer’s Guide provides the
information you need to write and debug assembly-language programs with the
Microsoft Macro Assembler (MASM), version 6.1. This book documents
enhanced features of the language and the programming environment for
MASM 6.1.
This Programmer’s Guide is written for experienced programmers who know
assembly language and are familiar with an assembler. The book does not teach
the basics of assembly language; it does explain Microsoft-specific features. If
you want to learn or review the basics of assembly language, refer to “Books for
Further Reading” in this introduction.
This book teaches you how to write efficient code with the new and advanced
features of MASM. Getting Started explains how to set up MASM 6.1.
Environment and Tools introduces the integrated development environment
called the Programmer’s WorkBench (PWB). It also includes a detailed
reference to Microsoft tools and utilities such as Microsoft ® CodeView ®,
LINK, and NMAKE. The Microsoft Macro Assembler Reference provides a full
listing of all MASM instructions, directives, statements, and operators, and it
serves as a quick reference to utility commands.
For more information on these same topics, see the online Microsoft Advisor,
which is a complete reference to Macro Assembler language topics, to the
utilities, and to PWB. You should be able to find most of the information you
need in the Microsoft Advisor.
New and Extended Features in MASM 6.1

MASM 6.1 continues the break with tradition established by version 6.0. It
incorporates conveniences of high-level languages while offering all the
traditional advantages of assembly-language programming.
For example, MASM 6.1 includes the Programmer’s WorkBench, which
provides the same integrated software development environment enjoyed by
programmers of Microsoft high-level languages such as C and Basic. From
within PWB you can edit, build, debug, or run a program. You can perform
most of these operations with either menu selections or keyboard commands.
You can also customize PWB to suit your individual programming and editing
requirements and preferences.
Filename: LMAPGINT.DOC Project:

Template: MSGRIDA1.DOT Author: Terri Sharkey Last Saved By: Ruth L Silverio
xiv Programmer’s Guide
MASM Features New Since Version 5.1

MASM 6.1 includes several features designed to make programming more
efficient and productive. The following list briefly describes how MASM 6.1
improves on the language features of the popular version 5.1.
u MASM 6.1 has many enhancements related to types. You can now use the
same type specifiers in initializations as in other contexts (BYTE instead of
DB). You can also define your own types, including pointer types, with the
new TYPEDEF directive. See Chapter 3, “Using Addresses and Pointers,”
and Chapter 4, “Defining and Using Simple Data Types.”
u The syntax for defining and using structures and records has been enhanced
since version 5.1. You can also define unions with the new UNION
directive. See Chapter 5, “Defining and Using Complex Data Types.”
u MASM now generates complete CodeView information for all types. See
Chapter 3, “Using Addresses and Pointers,” and Chapter 4, “Defining and
Using Simple Data Types.”
u New control-flow directives let you use high-level – language constructs such
as loops and if-then-else blocks defined with .REPEAT and .UNTIL (or
.UNTILCXZ); .WHILE and .ENDW; and .IF, .ELSE, and .ELSEIF. The
assembler generates the appropriate code to implement the control structure.
See Chapter 7, “Controlling Program Flow.”
u MASM now has more powerful features for defining and calling procedures.
The extended PROC syntax for generating stack frames has been enhanced
since version 5.1. You can also use the PROTO directive to prototype a
procedure, which you can then call with the INVOKE directive. INVOKE
automatically generates code to pass arguments (converting them to a related
type, if appropriate) and makes the call according to the specified calling
convention. See Chapter 7, “Controlling Program Flow.”
u MASM optimizes jumps by automatically determining the most efficient
coding for a jump and then generating the appropriate code. See Chapter 7,
“Controlling Program Flow.”
u Maintaining multiple-module programs is easier in MASM 6.1 than in version
5.1. The EXTERNDEF and PROTO directives make it easy to maintain all
global definitions in include files shared by all the source modules of a
project. See Chapter 8, “Sharing Data and Procedures Among Modules and
Libraries.”
The assembler has many new macro features that make complex macros clearer
and easier to write:
u You can specify default values for macro arguments or mark arguments as
required. And with the VARARG keyword, one parameter can accept a
variable number of arguments.

Introduction xv
u You can implement loops inside of macros in various ways. For example, the
new WHILE directive expands the statements in a macro body while an
expression is not zero.
u You can define macro functions, which return text macros. Several
predefined text macros are also provided for processing strings. Macro
operators and other features related to processing text macros and macro
arguments have been enhanced. For more information on all these macro
features, see Chapter 9, “Using Macros.”
MASM 6.1 has other improved capabilities, such as:

u The .STARTUP and .EXIT directives automatically generate appropriate
startup and exit code for your assembly-language programs. See Chapter 2,
“Organizing Segments.”
u MASM 6.1 supports flat memory model, available with the new Microsoft ®
Windows NT ™ operating system. Flat model allows segments as large as 4
gigabytes instead of 64K (kilobytes). Offsets are 32 bits instead of 16 bits.
See Chapter 2, “Organizing Segments.”
u The program H2INC.EXE converts C include files to MASM include files
and translates data structures and declarations. See Chapter 20 in
Environment and Tools.
u MASM 6.1 provides a library of assembly routines that let you create a
terminate-and-stay-resident program (TSR) in a high-level language.
MASM 6.1 includes many other minor new features as well as extensive support
for features of earlier versions of MASM. For a complete list of enhancements,
refer to Appendix A, “Differences between MASM 6.1 and 5.1.” The cross-
references in Appendix A guide you to the chapters where the new features are
described in detail.
MASM Features New Since Version 6.0

MASM 6.1 offers several new features:
u ML now runs in 32-bit protected mode under MS-DOS, giving it direct
access to extended memory for assembling very large source files.
u A collection of tools lets you write a dynamic-link library (DLL) for the
Microsoft ® Windows ™ operating system without the Windows Software
Development Kit. The LIBW.LIB library provides access to all functions in
the Windows application programming interface (API), so your DLL can
display menus, dialog boxes, and scroll bars. Chapter 10, “Writing a
Dynamic-Link Library for Windows,” shows you how.

xvi Programmer’s Guide
u Program listings now show instruction timings. The number of required

processor cycles appears adjacent to each instruction in the listing, based on
the selected processor. For an example listing and instructions on how to use
this feature, see Appendix C, “Generating and Reading Assembly Listings.”
u All utilities have been updated for version 6.1. Documentation is clearer and
better arranged, with a new Environment and Tools reference book.
u Version 6.1 generates debugging information for CodeView version 4.0 and
later.
u MASM 6.1 provides even greater compatibility with version 5.1 than does
MASM 6.0. Many programs written with version 5.1 will assemble
unchanged under MASM 6.1.
ML and MASM Command Lines

MASM 6.1 provides an updated version of the command-line driver, ML,
introduced in version 6.0. ML is more powerful and flexible than the MASM
driver of version 5.1. ML assembles and links with one command. It recognizes
all the old MASM driver command syntax, however, to support existing batch
files and makefiles that use MASM command lines.
Note The name MASM has traditionally referred to the Microsoft Macro
Assembler. It is used in that context throughout this book. However, MASM
also refers to MASM.EXE, which has been replaced by ML.EXE. In MASM
6.1, MASM.EXE is a small utility that translates command-line options to those
accepted by ML.EXE, and then calls ML.EXE. The distinction between
ML.EXE and MASM.EXE is made whenever necessary. Otherwise, MASM
refers to the assembler and its features.
Compatibility with Earlier Versions of MASM

MASM 6.1 is fully compatible with version 6.0 and, in many cases, with version
5.1. Code written for MASM 5.1 will often assemble correctly without
modification under MASM 6.1. However, MASM 6.1 provides the OPTION
directive to let you selectively modify the assembly process. In particular, you
can use the M510 argument with OPTION or the /Zm command-line option to
set most features to be compatible with version 5.1 code.
For information about obsolete features that will not assemble correctly under
MASM 6.1, see Appendix A, “Differences Between MASM 6.1 and 5.1.” The
appendix also explains how to update code to use the new features.

Introduction xvii
A Word About Instruction Timings

As an assembly-language programmer, whether novice or expert, you are
probably interested in producing lightning-fast code. After all, one of the main
reasons to program in assembly is to take advantage of its ability to streamline
execution speeds to the limit of the processor. This book will help you write
efficient and fast programs.
When discussing the speed of individual instructions, the chapters in this book
often speak of “timing,” which is the number of processor cycles required to
carry out an instruction. The Reference lists instruction timings for processors in
the 8086 family. It is tempting to use timing as the only criterion when judging
an instruction’s actual execution speed, but the world within the processor is not
so simple.
The clock for instruction timing does not begin ticking until the processor has
read and begins to execute an instruction. When you read about instruction
timings (in this book or any other), keep in mind that other factors also influence
the real speed of an instruction: the instruction’s size, whether it resides in cache
memory, whether it accesses memory, its position in the processor’s prefetch
queue, and the processor type. These factors make it impossible to say precisely
how fast an instruction executes. Accept the references to timing in this book as
guidelines, but use these simple rules to write fast code:
u Whenever possible, use registers rather than constant values, and constant
values rather than memory.
u Minimize changes in program flow.
u Smaller is often better. For example, the instructions
dec bx
sub bx, 1
accomplish the same thing and have the same timings on 80386/486
processors. But the first instruction is 3 bytes smaller than the second, and so
may reach the processor faster.
• When possible, use the string instructions described in Chapter 5, “Defining
and Using Complex Data Types.”

xviii Programmer’s Guide
Books for Further Reading

The following books may help you learn to program in assembly language or
write specialized programs. These books are listed only for your convenience.
Microsoft makes no specific recommendations concerning any of these books.
Books About Programming in Assembly Language

Abrash, Michael. Zen of Assembly Language. Glenview, IL: Scott, Foresman
and Co., 1990. Out of print.
Duntemann, Jeff. Assembly Language from Square One: For the PC AT and
Compatibles. Glenview, IL: Scott, Foresman and Co., 1990. Out of print.
Fernandez, Judi N., and Ruth Ashley. Assembly Language Programming for
the 80386. New York: McGraw-Hill, 1990.
Miller, Alan R. DOS Assembly Language Programming. San Francisco:
SYBEX, 1988. Out of print.
Scanlon, Leo J. 80286 Assembly Language Programming on MS-DOS
Computers. New York: Brady Communications, 1986. Out of print.
Turley, James L. Advanced 80386 Programming Techniques. Berkeley, CA:
Osborne McGraw-Hill, 1988.
Books About MS-DOS and BIOS

“Terminate-and-Stay-Resident Utilities.” MS-DOS Encyclopedia. Redmond,
WA: Microsoft Press, 1989.
Duncan, Ray. Advanced MS-DOS Programming: The Microsoft Guide for
Assembly Language and C Programmers. 2d ed. Redmond, WA: Microsoft
Press, 1988.
Duncan, Ray. Extending DOS: Programmer’s Guide to Protected-Mode DOS.
Redding, MA: Addison-Wesley. 1991.
Jourdain, Robert. Programmer’s Problem Solver for the IBM PC, XT and AT.
New York: Brady Communications, 1985. Out of print.
Microsoft MS-DOS Programmer’s Reference. Redmond, WA: Microsoft Press,
1991.
Norton, Peter and Richard Wilton. The New Peter Norton Programmer’s Guide
to the IBM PC and PS/2. Redmond, WA: Microsoft Press, 1988.
Wilton, Richard. Programmer’s Guide to PC & PS/2 Video Systems: Maximum
Video Performance from the EGA, VGA, HGC, and MCGA. Redmond, WA:
Microsoft Press, 1987. Out of print.

Introduction xix
Books and Articles About Windows

Kauler, Barry. Windows Assembly Language & Systems Programming: Object-
Oriented & Systems Programming in Assembly Language for Windows 3.0
and 3.1. New York, NY: Prentice Hall, 1993.
Klein, Mike. Windows Programmer’s Guide to DLLs & Memory Management.
Carmel, IN: Sams, 1992.
Petzold, Charles. Programming Windows. 3d ed. Redmond, WA: Microsoft
Press, 1992.
Petzold, Charles. “Environments.” PC Magazine. New York, NY: Ziff-Davis
Publishing Company, June 1990–1992.
Programmer’s Reference. 4 vols. Microsoft Windows Software Development
Kit (SDK). Redmond, WA: Microsoft Press, 1992.
Books About Other Topics

Nelson, Ross P. The 80386/80486 Programming Guide. 2d ed. Redmond, WA:
Microsoft Press, 1991.
Startz, Richard. 8087/80287/80387 for the IBM PC and Compatibles:
Applications and Programming with Intel’s Math Coprocessors. Bowie, MD:
Robert J. Brady Co., 1988. Out of print.
Document Conventions
The following document conventions are used throughout this manual:
Example of
Convention Description
SAMPLE2.ASM Uppercase letters indicate filenames, segment names, registers, and
terms used at the command level.
.MODEL Boldface type indicates assembly-language directives, instructions,
type specifiers, and predefined macros, as well as keywords in other
programming languages.
placeholder Italic letters indicate placeholders for information you must supply,
such as a filename. Italics are used occasionally for emphasis in the
text.
target This font is used to indicate example programs, user input, and screen
output.
; A semicolon in the first column of an example signals illegal code. A
semicolon also marks a comment.

xx Programmer’s Guide
SHIFT Small capital letters signify names of keys on the keyboard. Notice
that a plus (+) indicates a combination of keys. For example, CTRL+E
means to hold down the CTRL key while pressing the E key.
[[argument]] Items inside double square brackets are optional.
{register|memory} Braces and a vertical bar indicate a choice between two or more
items. You must choose one of the items unless double square
brackets surround the braces.
Repeating elements... A horizontal ellipsis (...) following an item indicates that more items
having the same form may appear.
Program A vertical ellipsis tells you that part of a program has been intentionally
. omitted.
.
.
Fragment
Getting Assistance and Reporting Problems

If you need help or think you have discovered a problem in the software, please
provide the following information to help us locate the source of the problem:
u The version of MS-DOS or Windows you run.
u Your system configuration: the type of machine you use, its total memory,
and its total free memory at assembler execution time, as well as any other
information you think might be useful.
u The command line you used for the assembler, linker, or other MASM tool
that was running when the problem occurred.
u Any object files or libraries you linked with if the problem occurred at link
time.
If your program is very large, reduce it to the smallest possible program that still
produces the problem.
Note the circumstances of the error and notify Microsoft Corporation by
following the instructions in the section “Microsoft Support Services” in the
introduction to Environment and Tools. If you have comments or suggestions
regarding any of the books accompanying this product, please indicate them on
the Document Feedback page at the back of this book and send it to Microsoft.
If you have not yet registered your copy of the Macro Assembler, you should
fill out and return the Registration Card. This enables Microsoft to keep you
informed of updates and other information about the assembler.

1
C H A P T E R 1
Understanding Global Concepts
With the development of the Microsoft Macro Assembler (MASM) version 6.1,
you now have more options available to you for approaching a programming
task. This chapter explains the general concepts of programming in assembly
language, beginning with the environment and a review of the components you
need to work in the assembler environment. Even if you are familiar with
previous versions of MASM, you should examine this chapter for information
on new terms and features.
The first section of this chapter reviews available processors and operating
systems and how they work together. The section also discusses segmented
architecture and how it affects a protected-mode operating environment such as
Windows.
The second section describes some of the language components of MASM that
are common to most programs, such as reserved words, constant expressions,
operators, and registers. The remainder of this book was written with the
assumption that you understand the information presented in this section.
The last section summarizes the assembly process, from assembling a program
through running it. You can affect this process by the way you develop your
code. Finally, this section explores how you can change the assembly process
with the OPTION directive and conditional assembly.
The Processing Environment

The processing environment for MASM 6.1 includes the processor on which
your programs run, the operating system your programs use, and the aspects of
the segmented architecture that influence the choice of programming models.
This section summarizes these elements of the environment and how they affect
your programming choices.
Filename: LMAPGC01.DOC Project:

2 Programmer’s Guide
8086-Based Processors
The 8086 “family” of processors uses segments to control data and code. The
later 8086-based processors have larger instruction sets and more memory
capacity, but they still support the same segmented architecture. Knowing the
differences between the various 8086-based processors can help you select the
appropriate target processor for your programs.
The instruction set of the 8086 processor is upwardly compatible with its
successors. To write code that runs on the widest number of machines, select
the 8086 instruction set. By using the instruction set of a more advanced
processor, you increase the capabilities and efficiency of your program, but you
also reduce the number of systems on which the program can run.
Table 1.1 lists modes, memory, and segment size of processors on which your
application may need to run. Each processor is discussed in more detail
following.
Table 1.1 8086 Family of Processors
Available Addressable Segment
Processor Modes Memory Size
8086/8088 Real 1 megabyte 16 bits
80186/80188 Real 1 megabyte 16 bits
80286 Real and Protected 16 megabytes 16 bits
80386 Real and Protected 4 gigabytes 16 or 32 bits
80486 Real and Protected 4 gigabytes 16 or 32 bits
Processor Modes
Real mode allows only one process to run at a time. The mode gets its name
from the fact that addresses in real mode always correspond to real locations in
memory. The MS-DOS operating system runs in real mode.
Windows 3.1 operates only in protected mode, but runs MS-DOS programs in
real mode or in a simulation of real mode called virtual-86 mode. In protected
mode, more than one process can be active at any one time. The operating
system protects memory belonging to one process from access by another
process; hence the name protected mode.
Protected-mode addresses do not correspond directly to physical memory.
Under protected-mode operating systems, the processor allocates and manages
memory dynamically. Additional privileged instructions initialize protected mode
and control multiple processes. For more information, see “Operating Systems,”
following.

Chapter 1 Understanding Global Concepts 3
8086 and 8088

The 8086 is faster than the 8088 because of its 16-bit data bus; the 8088 has
only an 8-bit data bus. The 16-bit data bus allows you to use EVEN and
ALIGN on an 8086 processor to word-align data and thus improve data-
handling efficiency. Memory addresses on the 8086 and 8088 refer to actual
physical addresses.
80186 and 80188

These two processors are identical to the 8086 and 8088 except that new
instructions have been added and several old instructions have been optimized.
These processors run significantly faster than the 8086.
80286
The 80286 processor adds some instructions to control protected mode, and it
runs faster. It also provides protected mode services, allowing the operating
system to run multiple processes at the same time. The 80286 is the minimum
for running Windows 3.1 and 16-bit versions of OS/2 ®.
80386
Unlike its predecessors, the 80386 processor can handle both 16-bit and 32-bit
data. It supports the entire instruction set of the 80286, and adds several new
instructions as well. Software written for the 80286 runs unchanged on the
80386, but is faster because the chip operates at higher speeds.
The 80386 implements many new hardware-level features, including paged
memory, multiple virtual 8086 processes, addressing of up to 4 gigabytes of
memory, and specialized debugging registers. Thirty-two–bit operating systems
such as Windows NT and OS/2 2.0 can run only on an 80386 or higher
processor.
80486
The 80486 processor is an enhanced version of the 80386, with instruction
“pipelining” that executes many instructions two to three times faster. The chip
incorporates both a math coprocessor and an 8K (kilobyte) memory cache. (The
math coprocessor is disabled on a variation of the chip called the 80486SX.)
The 80486 includes new instructions and is fully compatible with 80386
software.
8087, 80287, and 80387

These math coprocessors work concurrently with the 8086 family of processors.
Performing floating-point calculations with math coprocessors is up to 100 times
faster than emulating the calculations with integer instructions. Although there
are technical and performance differences among the three coprocessors, the
main difference to the applications programmer is that the 80287 and 80387 can

operate in protected mode. The 80387 also has several new instructions. The
80486 does not use any of these coprocessors; its floating-point processor is
built in and is functionally equivalent to the 80387.
Operating Systems
With MASM, you can create programs that run under MS-DOS, Windows, or
Windows NT — or all three, in some cases. For example, ML.EXE can produce
executable files that run in any of the target environments, regardless of the
programmer’s environment. For information on building programs for different
environments, see “Building and Running Programs” in Help for PWB.
MS-DOS and Windows 3.1 provide different processing modes. MS-DOS runs
in the single-process real mode. Windows 3.1 operates in protected mode,
allowing multiple processes to run simultaneously.
Although Windows requires another operating system for loading and file
services, it provides many functions normally associated with an operating
system. When an application requests an MS-DOS service, Windows often
provides the service without invoking MS-DOS. For consistency, this book
refers to Windows as an operating system.
MS-DOS and Windows (in protected mode) differ primarily in system access
methods, size of addressable memory, and segment selection. Table 1.2
summarizes these differences.
Table 1.2 The MS-DOS and Windows Operating Systems Compared
Available Contents of
Operating System Active Addressable Segment Word
System Access Processes Memory Register Length
MS-DOS and Direct to One 1 megabyte Actual 16 bits
Windows real hardware address
mode and OS call
Windows Operating Multiple 1 megabyte Segment 16 bits
virtual-86 mode system call selectors
Windows Operating Multiple 16 megabytes Segment 16 bits
protected mode system call selectors
Windows NT Operating Multiple 512 Segment 32 bits
system call megabytes selectors
MS-DOS
In real-mode programming, you can access system functions by calling MS-
DOS, calling the basic input/output system (BIOS), or directly addressing
hardware. Access is through MS-DOS Interrupt 21h.

Windows
As you can see in Table 1.2, protected mode allows for much larger data
structures than real mode, since addressable memory extends to 16 megabytes.
In protected mode, segment registers contain selector values rather than actual
segment addresses. These selectors cannot be calculated by the program; they
must be obtained by calling the operating system. Programs that attempt to
calculate segment values or to address memory directly do not work in protected
mode.
Protected mode uses privilege levels to maintain system integrity and security.
Programs cannot access data or code that is in a higher privilege level. Some
instructions that directly access ports or affect interrupts (such as CLI, STI, IN,
and OUT) are available at privilege levels normally used only by systems
programmers.
Windows protected mode provides each application with up to 16 megabytes of
“virtual memory,” even on computers that have less physical memory. The term
virtual memory refers to the operating system’s ability to use a swap area on the
hard disk as an extension of real memory. When a Windows application requires
more memory than is available, Windows writes sections of occupied memory
to the swap area, thus freeing those sections for other use. It then provides the
memory to the application that made the memory request. When the owner of
the swapped data regains control, Windows restores the data from disk to
memory, swapping out other memory if required.
Windows NT
Windows NT uses the so-called “flat model” of 80386/486 processors. This
model places the processor’s entire address space within one 32-bit segment.
The section “Defining Basic Attributes with .MODEL” in Chapter 2 explains
how to use the flat model. In flat model, your program can (in theory) access up
to 4 gigabytes of virtual memory. Since code, data, and stack reside in the same
segment, each segment register can hold the same value, which need never
change.
Segmented Architecture
The 8086 family of processors employs a segmented architecture — that is, each
address is represented as a segment and an offset. Segmented addresses affect
many aspects of assembly-language programming, especially addresses and
pointers.
Segmented architecture was originally designed to enable a 16-bit processor to
access an address space larger than 64K. (The section “Segmented Addressing,”
later in this chapter, explains how the processor uses both the segment and
offset to create addresses larger than 64K.) MS-DOS is an example of an
operating system that uses segmented architecture on a 16-bit processor.

With the advent of protected-mode processors such as the 80286, segmented

architecture gained a second purpose. Segments can separate different blocks of
code and data to protect them from undesirable interactions. Windows takes
advantage of the protection features of the 16-bit segments on the 80286.
Segmented architecture went through another significant change with the release
of 32-bit processors, starting with the 80386. These processors are compatible
with the older 16-bit processors, but allow flat model 32-bit offset values up to 4
gigabytes. Offset values of this magnitude remove the memory limitations of
segmented architecture. The Windows NT operating system uses 32-bit
addressing.
Segment Protection
Segmented architecture is an important part of the Windows memory-protection
scheme. In a “multitasking” operating system in which numerous programs can
run simultaneously, programs cannot access the code and data of another
process without permission.
In MS-DOS, the data and code segments are usually allocated adjacent to each
other, as shown in Figure 1.1. In Windows, the data and code segments can be
anywhere in memory. The programmer knows nothing about, and has no
control over, their location. The operating system can even move the segments
to a new memory location or to disk while the program is running.
Figure 1.1 Segment Allocation
Segment protection makes software development easier and more reliable in

Windows than in MS-DOS, because Windows immediately detects illegal

memory accesses. The operating system intercepts illegal memory accesses,

terminates the program, and displays a message. This makes it easier for you to
track down and fix the bug.
Because it runs in real mode, MS-DOS contains no mechanism for detecting an
improper memory access. A program that overwrites data not belonging to it
may continue to run and even terminate correctly. The error may not surface
until later, when MS-DOS or another program reads the corrupted memory.
Segmented Addressing
Segmented addressing refers to the internal mechanism that combines a segment
value and an offset value to form a complete memory address. The two parts of
an address are represented as
segment:offset
The segment portion always consists of a 16-bit value. The offset portion is a
16-bit value in 16-bit mode or a 32-bit value in 32-bit mode.
In real mode, the segment value is a physical address that has an arithmetic
relationship to the offset value. The segment and offset together create a 20-bit
physical address (explained in the next section). Although 20-bit addresses can
access up to 1 megabyte of memory, the BIOS and operating system on
International Standard Architecture (IBM PC/AT and compatible) computers
use part of this memory, leaving the remainder available for programs.
Segment Arithmetic
Manipulating segment and offset addresses directly in real-mode programming is
called “segment arithmetic.” Programs that perform segment arithmetic are not
portable to protected-mode operating systems, in which addresses do not
correspond to a known segment and offset.
To perform segment arithmetic successfully, it helps to understand how the
processor combines a 16-bit segment and a 16-bit offset to form a 20-bit linear
address. In effect, the segment selects a 64K region of memory, and the offset
selects the byte within that region. Here’s how it works:
1. The processor shifts the segment address to the left by four binary places,
producing a 20-bit address ending in four zeros. This operation has the effect
of multiplying the segment address by 16.
2. The processor adds this 20-bit segment address to the 16-bit offset address.
The offset address is not shifted.
3. The processor uses the resulting 20-bit address, called the “physical
address,” to access an actual location in the 1-megabyte address space.

Figure 1.2 illustrates this process.
Figure 1.2 Calculating Physical Addresses
A 20-bit physical address may actually be specified by 4,096 equivalent

segment:offset addresses. For example, the addresses 0000:F800, 0F00:0800,
and 0F80:0000 all refer to the same physical address 0F800.
Language Components of MASM

Programming with MASM requires that you understand the MASM concepts of
reserved words, identifiers, predefined symbols, constants, expressions,
operators, data types, registers, and statements. This section defines important
terms and provides lists that summarize these topics. For detailed information,
see Help or the Reference.
Reserved Words
A reserved word has a special meaning fixed by the language. You can use it
only under certain conditions. Reserved words in MASM include:
u Instructions, which correspond to operations the processor can execute.
u Directives, which give commands to the assembler.
u Attributes, which provide a value for a field, such as segment alignment.
u Operators, which are used in expressions.

u Predefined symbols, which return information to your program.
MASM reserved words are not case sensitive except for predefined symbols
(see “Predefined Symbols,” later in this chapter).
The assembler generates an error if you use a reserved word as a variable, code
label, or other identifier within your source code. However, if you need to use a
reserved word for another purpose, the OPTION NOKEYWORD directive can
selectively disable a word’s status as a reserved word.
For example, to remove the STR instruction, the MASK operator, and the
NAME directive from the set of words MASM recognizes as reserved, use this
statement in the code segment of your program before the first reference to
STR, MASK, or NAME:
OPTION NOKEYWORD:<STR MASK NAME>
The section “Using the OPTION Directive,” later in this chapter, discusses the
OPTION directive. Appendix D provides a complete list of MASM reserved
words.
With the /Zm command-line option or OPTION M510 in effect, MASM does
not reserve any operators or instructions that do not apply to the current CPU
mode. For example, you can use the symbol ENTER when assembling under the
default CPU mode but not under .286 mode, since the 80186/486 processors
recognize ENTER as an instruction. The USE32, FLAT, FAR32, and NEAR32
segment types and the 80386/486 register names are not keywords with
processors other than the 80386/486.
Identifiers
An identifier is a name that you invent and attach to a definition. Identifiers can
be symbols representing variables, constants, procedure names, code labels,
segment names, and user-defined data types such as structures, unions, records,
and types defined with TYPEDEF. Identifiers longer than 247 characters
generate an error.
Certain restrictions limit the names you can use for identifiers. Follow these
rules to define a name for an identifier:
u The first character of the identifier can be an alphabetic character (A–Z) or
any of these four characters: @ _ $ ?
u The other characters in the identifier can be any of the characters listed
above or a decimal digit (0–9).
Avoid starting an identifier with the at sign (@), because MASM 6.1 predefines
some special symbols starting with @ (see “Predefined Symbols,” following).

Beginning an identifier with @ may also cause conflicts with future versions of
the Macro Assembler.
The symbol — and thus the identifier — is visible as long as it remains within
scope. (For more information about visibility and scope, see “Sharing Symbols
with Include Files” in Chapter 8.)
Predefined Symbols
The assembler includes a number of predefined symbols (also called predefined
equates). You can use these symbol names at any point in your code to
represent the equate value. For example, the predefined equate @FileName
represents the base name of the current file. If the current source file is
TASK.ASM, the value of @FileName is TASK. The MASM predefined
symbols are listed according to the kinds of information they provide. Case is
important only if the /Cp option is used. (For additional details, see Help on ML
command-line options.)
The predefined symbols for segment information include:
Symbol Description
@code Returns the name of the code segment.
@CodeSize Returns an integer representing the default code distance.
@CurSeg Returns the name of the current segment.
@data Expands to DGROUP.
@DataSize Returns an integer representing the default data distance.
@fardata Returns the name of the segment defined by the .FARDATA directive.
@fardata? Returns the name of the segment defined by the .FARDATA? directive.
@Model Returns the selected memory model.
@stack Expands to DGROUP for near stacks or STACK for far stacks. (See
“Creating a Stack” in Chapter 2.)
@WordSize Provides the size attribute of the current segment.
The predefined symbols for environment information include:

Symbol Description
@Cpu Contains a bit mask specifying the processor mode.
@Environ Returns values of environment variables during assembly.
@Interface Contains information about the language parameters.
@Version Represents the text equivalent of the MASM version number. In MASM 6.1,
this expands to 610.

The predefined symbols for date and time information include:

Symbol Description
@Date Supplies the current system date during assembly.
@Time Supplies the current system time during assembly.
The predefined symbols for file information include:

Symbol Description
@FileCur Names the current file (base and suffix).
@FileName Names the base name of the main file being assembled as it appears on the
command line.
@Line Gives the source line number in the current file.
The predefined symbols for macro string manipulation include:

Symbol Description
@CatStr Returns concatenation of two strings.
@InStr Returns the starting position of a string within another string.
@SizeStr Returns the length of a given string.
@SubStr Returns substring from a given string.
Integer Constants and Constant Expressions

An integer constant is a series of one or more numerals followed by an optional
radix specifier. For example, in these statements
mov ax, 25
mov bx, 0B3h
the numbers 25 and 0B3h are integer constants. The h appended to 0B3 is a
radix specifier. The specifiers are:
u y for binary (or b if the default radix is not hexadecimal)
u o or q for octal
u t for decimal (or d if the default radix is not hexadecimal)
u h for hexadecimal
Radix specifiers can be either uppercase or lowercase letters; sample code in this
book is in lowercase. If you do not specify a radix, the assembler interprets the
integer according to the current radix. The default radix is decimal, but you can
change the default with the .RADIX directive.

Hexadecimal numbers must always start with a decimal digit (0–9). If necessary,
add a leading zero to distinguish between symbols and hexadecimal numbers
that start with a letter. For example, MASM interprets ABCh as an identifier.
The hexadecimal digits A through F can be either uppercase or lowercase letters.
Sample code in this book is in uppercase letters.
Constant expressions contain integer constants and (optionally) operators such
as shift, logical, and arithmetic operators. The assembler evaluates constant
expressions at assembly time. (In addition to constants, expressions can contain
labels, types, registers, and their attributes.) Constant expressions do not change
value during program execution.
Symbolic Integer Constants

You can define symbolic integer constants with either of the data assignment
directives, EQU or the equal sign (=). These directives assign values to symbols
during assembly, not during program execution. Symbolic constants are used to
assign names to constant values. You can use a symbol with an assigned value
in place of an immediate operand. For example, instead of referring in your code
to keyboard scan codes with numbers such as 30 or 48, you can create more
recognizable symbols:
SCAN_A EQU 30
SCAN_B EQU 48
then use the appropriate symbol in your program rather than the number. Using
symbolic constants instead of undescriptive numbers makes your code more
readable and easier to maintain. The assembler does not allocate data storage
when you use either EQU or =. It simply replaces each occurrence of the
symbol with the value of the expression.
The directives EQU and = have slightly different purposes. Integers defined
with the = directive can be redefined with another value in your source code,
but those defined with EQU cannot. Once you’ve defined a symbolic constant
with the EQU directive, attempting to redefine it generates an error. The syntax
is:
symbol EQU expression
The symbol is a unique name of your choice, except for words reserved by
MASM. The expression can be an integer, a constant expression, a one- or two-
character string constant (four-character on the 80386/486), or an expression
that evaluates to an address. Symbolic constants let you change a constant value
used throughout your source code by merely altering expression in the
definition. This removes the potential for error and saves you the inconvenience
of having to find and replace each occurrence of the constant in your program.

The following example shows the correct use of EQU to define symbolic
integers.
column EQU 80 ; Constant - 80
row EQU 25 ; Constant - 25
screen EQU column * row ; Constant - 2000
line EQU row ; Constant - 25
.DATA
.CODE
.
.
.
mov cx, column
mov bx, line
The value of a symbol defined with the = directive can be different at different
places in the source code. However, a constant value is assigned during
assembly for each use, and that value does not change at run time.
The syntax for the = directive is:
symbol = expression
Size of Constants
The default word size for MASM 6.1 expressions is 32 bits. This behavior can
be modified using OPTION EXPR16 or OPTION M510. Both of these options
set the expression word size to 16 bits, but OPTION M510 affects other
assembler behavior as well (see Appendix A).
It is illegal to change the expression word size once it has been set with
OPTION M510, OPTION EXPR16, or OPTION EXPR32. However, you
can repeat the same directive in your source code as often as you wish. You can
place the same directive in every include file, for example.
Operators
Operators are used in expressions. The value of the expression is determined at
assembly time and does not change when the program runs.
Operators should not be confused with processor instructions. The reserved
word ADD is an instruction; the plus sign (+) is an operator. For example,
Amount+2 illustrates a valid use of the plus operator (+). It tells the assembler
to add 2 to the constant value Amount, which might be a value or an address.
Contrast this operation, which occurs at assembly time, with the processor’s
ADD instruction. ADD tells the processor at run time to add two numbers and
store the result.

The assembler evaluates expressions that contain more than one operator
according to the following rules:
u Operations in parentheses are performed before adjacent operations.
u Binary operations of highest precedence are performed first.
u Operations of equal precedence are performed from left to right.
u Unary operations of equal precedence are performed right to left.
Table 1.3 lists the order of precedence for all operators. Operators on the same
line have equal precedence.
Table 1.3 Operator Precedence
Precedence Operators
1 ( ), [ ]
2 LENGTH, SIZE, WIDTH, MASK, LENGTHOF, SIZEOF
3 . (structure-field-name operator)
4 : (segment-override operator), PTR
5 LROFFSET, OFFSET, SEG, THIS , TYPE
6 HIGH, HIGHWORD, LOW, LOWWORD
7 + ,– (unary)
8 *, /, MOD, SHL, SHR
9 +, – (binary)
10 EQ, NE, LT, LE, GT, GE
11 NOT
12 AND
13 OR, XOR
14 OPATTR, SHORT, .TYPE
Data Types
A “data type” describes a set of values. A variable of a given type can have any
of a set of values within the range specified for that type.
The intrinsic types for MASM 6.1 are BYTE, SBYTE, WORD, SWORD,
DWORD, SDWORD, FWORD, QWORD, and TBYTE. These types define
integers and binary coded decimals (BCDs), as discussed in Chapter 6. The
signed data types SBYTE, SWORD, and SDWORD work in conjunction with
directives such as INVOKE (for calling procedures) and .IF (introduced in
Chapter 7). The REAL4, REAL8, and REAL10 directives define floating-point
types. (See Chapter 6.)

Versions of MASM prior to 6.0 had separate directives for types and initializers.
For example, BYTE is a type and DB is the corresponding initializer. The
distinction does not apply in MASM 6.1. You can use any type (intrinsic or
user-defined) as an initializer.
MASM does not have specific types for arrays and strings. However, you can
treat a sequence of data units as arrays, and character or byte sequences as
strings. (See “Arrays and Strings” in Chapter 5.)
Types can also have attributes such as langtype and distance (NEAR and
FAR). For information on these attributes, see “Declaring Parameters with the
PROC Directive” in Chapter 7.
You can also define your own types with STRUCT, UNION, and RECORD.
The types have fields that contain string or numeric data, or records that contain
bits. These data types are similar to the user-defined data types in high-level
languages such as C, Pascal, and FORTRAN. (See Chapter 5, “Defining and
Using Complex Data Types.”)
You can define new types, including pointer types, with the TYPEDEF
directive. TYPEDEF assigns a qualifiedtype (explained in the following) to a
typename of your choice. This lets you build new types with descriptive names
of your choosing, making your programs more readable. For example, the
following statement makes the symbol CHAR a synonym for the intrinsic type
BYTE:
CHAR TYPEDEF BYTE
The qualifiedtype is any type or pointer to a type of the form:

[[distance]] PTR [[qualifiedtype]]
where distance is NEAR, FAR, or any distance modifier. (For more
information on distance, see “Declaring Parameters with the PROC Directive”
in Chapter 7.)
The qualifiedtype can also be any type previously defined with TYPEDEF. For
example, if you use TYPEDEF to create an alias for BYTE — say, CHAR as in
the preceding example — you can use CHAR as a qualifiedtype when defining the
pointer type PCHAR, like this:
CHAR TYPEDEF BYTE
PCHAR TYPEDEF PTR CHAR
The typename CHAR in the first line becomes a qualifiedtype in the second line.
Use of the TYPEDEF directive to define pointers is explained in “Accessing
Data with Pointers and Addresses” in Chapter 3.

Since distance and qualifiedtype are optional syntax elements, you can use
variables of type PTR or FAR PTR. You can also define procedure prototypes
with qualifiedtype. For more information about procedure prototypes, see
“Declaring Procedure Prototypes” in Chapter 7.
These rules govern the use of qualifiedtype:
u The only component of a qualifiedtype definition that can be forward-
referenced is a structure or union type identifier.
u If you do not specify distance, the assembler assumes a distance that
corresponds to the memory model. The assumed distance is NEAR for tiny,
small, and medium models, and FAR for other models.
u If you do not specify a memory model with .MODEL, the assembler
assumes SMALL model (and therefore NEAR pointers).
You can use a qualifiedtype in seven places:

Use Example
In procedure arguments proc1 PROC pMsg:PTR BYTE
In prototype arguments proc2 PROTO pMsg:FAR PTR WORD
With local variables declared inside LOCAL pMsg:PTR
procedures
With the LABEL directive TempMsg LABEL PTR WORD
With the EXTERN and EXTERNDEF EXTERN pMsg:FAR PTR BYTE
directives EXTERNDEF MyProc:PROTO
With the COMM directive COMM var1:WORD:3
With the TYPEDEF directive PBYTE TYPEDEF PTR BYTE
PFUNC TYPEDEF PROTO MyProc
“Defining Pointer Types with TYPEDEF” in Chapter 3 shows ways to write a

TYPEDEF type for a qualifiedtype. Attributes such as NEAR and FAR can
also apply to a qualifiedtype.
You can determine an accurate definition for TYPEDEF and qualifiedtype from
the BNF grammar definitions given in Appendix B. The BNF grammar defines
each component of the syntax for any directive, showing the recursive
properties of components such as qualifiedtype.
Registers
The 8086 family of processors have the same base set of 16-bit registers. Each
processor can treat certain registers as two separate 8-bit registers. The
80386/486 processors have extended 32-bit registers. To maintain compatibility

with their predecessors, 80386/486 processors can access their registers as 16-
bit or, where appropriate, as 8-bit values.
Figure 1.3 shows the registers common to all the 8086-based processors. Each
register has its own special uses and limitations.
Figure 1.3 Registers for 8088 – 80286 Processors
80386/486 Only
The 80386/486 processors use the same 8-bit and 16-bit registers used by the
rest of the 8086 family. All of these registers can be further extended to 32 bits,
except segment registers, which always occupy 16 bits. The extended register
names begin with the letter “E.” For example, the 32-bit extension of AX is
EAX. The 80386/486 processors have two additional segment registers, FS and
GS. Figure 1.4 shows the extended registers of the 80386/486.

Figure 1.4 Extended Registers for the 80386/486 Processors
Segment Registers
At run time, all addresses are relative to one of four segment registers: CS, DS,
SS, or ES. (The 80386/486 processors add two more: FS and GS.) These
registers, their segments, and their purposes include:
Register and Segment Purpose
CS (Code Segment) Contains processor instructions and their immediate operands.
DS (Data Segment) Normally contains data allocated by the program.
SS (Stack Segment) Contains the program stack for use by PUSH , POP, CALL,
and RET.

Register and Segment Purpose

ES (Extra Segment) References secondary data segment. Used by string
instructions.
FS, GS Provides extra segments on the 80386/486.
General-Purpose Registers
The AX, DX, CX, BX, BP, DI, and SI registers are 16-bit general-purpose
registers, used for temporary data storage. Since the processor accesses registers
more quickly than it accesses memory, you can make your programs run faster
by keeping the most-frequently used data in registers.
The 8086-based processors do not perform memory-to-memory operations. For
example, the processor cannot directly copy a variable from one location in
memory to another. You must first copy from memory to a register, then from
the register to the new memory location. Similarly, to add two variables in
memory, you must first copy one variable to a register, then add the contents of
the register to the other variable in memory.
The processor can access four of the general registers — AX, DX, CX, and BX
— either as two 8-bit registers or as a single 16-bit register. The AH, DH, CH,
and BH registers represent the high-order 8 bits of the corresponding registers.
Similarly, AL, DL, CL, and BL represent the low-order 8 bits of the registers.
The 80386/486 processors can extend all the general registers to 32 bits, though
as Figure 1.4 shows, you cannot treat the upper 16 bits as a separate register as
you can the lower 16 bits. To use EAX as an example, you can directly
reference the low byte as AL, the next lowest byte as AH, and the low word as
AX. To access the high word of EAX, however, you must first shift the upper
16 bits into the lower 16 bits.
Special-Purpose Registers
The 8086 family of processors has two additional registers, SP and IP, whose
values are changed automatically by the processor.
SP (Stack Pointer)
The SP register points to the current location within the stack segment. Pushing
a value onto the stack decreases the value of SP by two; popping from the stack
increases the value of SP by two. Thirty-two–bit operands on 80386/486
processors increase or decrease SP by four instead of two. The CALL and INT
instructions store the return address on the stack and reduce SP accordingly.
Return instructions retrieve the stored address from the stack and reset SP to its
value before the call. SP can also be adjusted with instructions such as ADD.
The program stack is described in detail in Chapter 3.

IP (Instruction Pointer)
The IP register always contains the address of the next instruction to be
executed. You cannot directly access or change the instruction pointer.
However, instructions that control program flow (such as calls, jumps, loops,
and interrupts) automatically change the instruction pointer.
Flags Register
The 16 bits in the flags register control the execution of certain instructions and
reflect the current status of the processor. In 80386/486 processors, the flags
register is extended to 32 bits. Some bits are undefined, so there are actually 9
flags for real mode, 11 flags (including a 2-bit flag) for 80286 protected mode,
13 for the 80386, and 14 for the 80486. The extended flags register of the
80386/486 is sometimes called “Eflags.”
Figure 1.5 shows the bits of the 32-bit flags register for the 80386/486. Earlier
8086-family processors use only the lower word. The unmarked bits are
reserved for processor use, and should not be modified.
Figure 1.5 Flags for 8088-80486 Processors
In the following descriptions and throughout this book, “set” means a bit value
of 1, and “cleared” means the bit value is 0. The nine flags common to all 8086-
family processors, starting with the low-order flags, include:

Flag Description
Carry Set if an operation generates a carry to or a borrow from a destination
operand.
Parity Set if the low-order bits of the result of an operation contain an even
number of set bits.
Auxiliary Carry Set if an operation generates a carry to or a borrow from the low-order 4
bits of an operand. This flag is used for binary coded decimal (BCD)
arithmetic.
Zero Set if the result of an operation is 0.
Sign Equal to the high-order bit of the result of an operation (0 is positive, 1 is
negative).
Trap If set, the processor generates a single-step interrupt after each
instruction. A debugging program can use this feature to execute a
program one instruction at a time.
Interrupt Enable If set, interrupts are recognized and acted on as they are received. The bit
can be cleared to turn off interrupt processing temporarily.
Direction If set, string operations process down from high addresses to low
addresses. If cleared, string operations process up from low addresses to
high addresses.
Overflow Set if the result of an operation is too large or small to fit in the destination
operand.
Although all flags serve a purpose, most programs require only the carry, zero,
sign, and direction flags.
Statements
Statements are the line-by-line components of source files. Each MASM
statement specifies an instruction or directive for the assembler. Statements have
up to four fields, as shown here:
[[name:]] [[operation]] [[operands]] [[;comment]]
The following list explains each field:
Field Purpose
name Labels the statement, so that instructions elsewhere in the program can refer to
the statement by name. The name field can label a variable, type, segment, or
code location.
operation Defines the action of the statement. This field contains either an instruction or an
assembler directive.
operands Lists one or more items on which the instruction or directive operates.
comment Provides a comment for the programmer. Comments are for documentation
only; they are ignored by the assembler.

The following line contains all four fields:

mainlp: mov ax, 7 ; Load AX with the value 7
Here, mainlp is the label, mov is the operation, and ax and 7 are the operands,
separated by a comma. The comment follows the semicolon.
All fields are optional, although certain directives and instructions require an
entry in the name or operand field. Some instructions and directives place
restrictions on the choice of operands. By default, MASM is not case sensitive.
Each field (except the comment field) must be separated from other fields by
white-space characters (spaces or tabs). MASM also requires code labels to be
followed by a colon, operands to be separated by commas, and comments to be
preceded by a semicolon.
A logical line can contain up to 512 characters and occupy one or more physical
lines. To extend a logical line into two or more physical lines, put the backslash
character (\) as the last non-whitespace character before the comment or end of
the line. You can place a comment after the backslash as shown in this example:
.IF (x > 0) \ ; X must be positive
&& (ax > x) \ ; Result from function must be > x
&& (cx == 0) ; Check loop counter, too
mov dx, 20h
.ENDIF
Multiline comments can also be specified with the COMMENT directive. The
assembler ignores all text and code between the delimiters or on the same line as
the delimiters. This example illustrates the use of COMMENT.
COMMENT ^ The assembler
ignores this text
^ mov ax, 1 and this code
The Assembly Process

Creating and running an executable file involves four steps:
1. Assembling the source code into an object file
2. Linking the object file with other modules or libraries into an executable
program
3. Loading the program into memory
4. Running the program

Once you have written your assembly-language program, MASM provides

several options for assembling it. The OPTION directive has several different
arguments that let you control the way MASM assembles your programs.
Conditional assembly allows you to create one source file that can generate a
variety of programs, depending on the status of various conditional-assembly
statements.
Generating and Running Executable Programs

This section briefly lists all the actions that take place during each of the
assembly steps. You can change the behavior of some of these actions in
various ways, such as using macros instead of procedures, or using the
OPTION directive or conditional assembly. The other chapters in this book
include specific programming methods; this section simply gives you an
overview.
Assembling
The ML.EXE program does two things to create an executable program. First, it
assembles the source code into an intermediate object file. Second, it calls the
linker, LINK.EXE, which links the object files and libraries into an executable
program.
At assembly time, the assembler:
u Evaluates conditional-assembly directives, assembling if the conditions are
true.
u Expands macros and macro functions.
u Evaluates constant expressions such as MYFLAG AND 80H, substituting the
calculated value for the expression.
u Encodes instructions and nonaddress operands. For example, mov cx, 13
can be encoded at assembly time because the instruction does not access
memory.
u Saves memory offsets as offsets from their segments.
u Places segments and segment attributes in the object file.
u Saves placeholders for offsets and segments (relocatable addresses).
u Outputs a listing if requested.
u Passes messages (such as INCLUDELIB and .DOSSEG) directly to the
linker.
For information about conditional assembly, see “Conditional Directives” in this

chapter; for macros, see Chapter 9. Further details about segments and offsets
are included in Chapters 2 and 3. Assembly listings are explained in Appendix C.

Linking
Once your source code is assembled, the resulting object file is passed to the
linker. At this point, the linker may combine several object files into an
executable program. The linker:
u Combines segments according to the instructions in the object files,
rearranging the positions of segments that share the same class or group.
u Fills in placeholders for offsets (relocatable addresses).
u Writes relocations for segments into the header of .EXE files (but not .COM
files).
u Writes the result as an executable program file.
Classes and groups are defined in “Defining Segment Groups” in Chapter 2.

Segments and offsets are explained in Chapter 3, “Using Addresses and
Pointers.”
Loading
After loading the executable file into memory, the operating system:
u Creates the program segment prefix (PSP) header in memory.
u Allocates memory for the program, based on the values in the PSP.
u Loads the program.
u Calculates the correct values for absolute addresses from the relocation table.
u Loads the segment registers SS, CS, DS, and ES with values that point to the
proper areas of memory.
For information about segment registers, the instruction pointer (IP), and the
stack pointer (SP), see “Registers” earlier in this chapter. For more information
on the PSP see Help or an MS-DOS reference.
Running
To run your program, MS-DOS jumps to the program’s first instruction. Some
program operations, such as resolving indirect memory operands, cannot be
handled until the program runs. For a description of indirect references, see
“Indirect Operands” in Chapter 7.
Using the OPTION Directive

The OPTION directive lets you modify global aspects of the assembly process.
With OPTION, you can change command-line options and default arguments.
These changes affect only statements that follow the OPTION keyword.

For example, you may have MASM code in which the first character of a
variable, macro, structure, or field name is a dot (.). Since a leading dot causes
MASM 6.1 to generate an error, you can use this statement in your program:
OPTION DOTNAME
This enables the use of the dot for the first character.
Changes made with OPTION override any corresponding command-line option.
For example, suppose you compile a module with this command line (which
enables M510 compatibility):
ML /Zm TEST.ASM
The assembler disables M510 compatibility options for all code following this
statement:
OPTION NOM510
The following lists explain each of the arguments for the OPTION directive.
Where appropriate, an underline identifies the default argument. If you wish to
place more than one OPTION statement on a line, separate them by commas.
Options for M510 compatibility include:
Argument Description
CASEMAP: maptype CASEMAP:NONE (or /Cx) causes internal
symbol recognition to be case sensitive and causes
the case of identifiers in the .OBJ file to be the same
as specified in the EXTERNDEF, PUBLIC, or
COMM statement. The default is
CASEMAP:NOTPUBLIC (or /Cp). It specifies
case insensitivity for internal symbol recognition and
the same behavior as CASEMAP:NONE for case
of identifiers in .OBJ files. CASEMAP:ALL (/Cu)
specifies case insensitivity for identifiers and
converts all identifier names to uppercase.
DOTNAME | NODOTNAME Enables the use of the dot (.) as the leading
character in variable, macro, structure, union, and
member names.
M510 | NOM510 Sets all features to be compatible with MASM
version 5.1, disabling the SCOPED argument and
enabling OLDMACROS, DOTNAME, and,
OLDSTRUCTS. OPTION M510 conditionally
sets other arguments for the OPTION directive.
For more information on using OPTION M510,
see Appendix A.

OLDMACROS | NOOLDMACROS Enables the version 5.1 treatment of macros.
MASM 6.1 treats macros differently.
OLDSTRUCTS | NOOLDSTRUCTS Enables compatibility with MASM 5.1 for treatment
of structure members. See Chapter 5 for
information on structures.
SCOPED | NOSCOPED Guarantees that all labels inside procedures are
local to the procedure when SCOPED (the default)
is enabled.
SETIF2: TRUE | FALSE If TRUE, .ERR2 statements and IF2 and
ELSEIF2 conditional blocks are evaluated on
every pass. If FALSE, they are not evaluated. If
SETIF2 is not specified (or implied), .ERR2, IF2 ,
and ELSEIF2 expressions cause an error. Both the
/Zm command-line argument and OPTION M510
imply SETIF2:TRUE.
Options for procedure use include:

LANGUAGE : langtype Specifies the default language type (C, PASCAL,
FORTRAN, BASIC, SYSCALL, or STDCALL)
to be used with PROC, EXTERN, and PUBLIC.
This use of the OPTION directive overrides the
.MODEL directive but is normally used when
.MODEL is not given.
EPILOGUE: macroname Instructs the assembler to call the macroname to
generate a user-defined epilogue instead of the
standard epilogue code when a RET instruction is
encountered. See Chapter 7.
PROLOGUE: macroname Instructs the assembler to call macroname to
generate a user-defined prologue instead of
generating the standard prologue code. See
Chapter 7.
PROC: visibility Lets you explicitly set the default visibility as
PUBLIC, EXPORT, or PRIVATE.
Other options include:

EXPR16 | EXPR32 Sets the expression word size to 16 or 32 bits. The
default is 32 bits. The M510 argument to the
OPTION directive sets the word size to 16 bits.
Once set with the OPTION directive, the
expression word size cannot be changed.

EMULATOR | NOEMULATOR Controls the generation of floating-point
instructions.The NOEMULATOR option
generates the coprocessor instructions directly. The
EMULATOR option generates instructions with
special fixup records for the linker so that the
Microsoft floating-point emulator, supplied with
other Microsoft languages, can be used. It produces
the same result as setting the /Fpi command-line
option. You can set this option only once per
module.
LJMP | NOLJMP Enables automatic conditional-jump lengthening.
For information about conditional-jump lengthening,
see Chapter 7.
NOKEYWORD:<keywordlist> Disables the specified reserved words. For an
example of the syntax for this argument, see
“Reserved Words” in this chapter.
NOSIGNEXTEND Overrides the default sign-extended opcodes for the
AND, OR, and XOR instructions and generates the
larger non-sign-extended forms of these
instructions. Provided for compatibility with NEC
V25 and NEC V35 controllers.
OFFSET: offsettype Determines the result of OFFSET operator fixups.
SEGMENT sets the defaults for fixups to be
segment-relative (compatible with MASM 5.1).
GROUP, the default, generates fixups relative to the
group (if the label is in a group). FLAT causes
fixups to be relative to a flat frame. (The .386 mode
must be enabled to use FLAT.) See Appendix A.
READONLY | NOREADONLY Enables checking for instructions that modify code
segments, thereby guaranteeing that read-only code
segments are not modified. Same as the /p
command-line option of MASM 5.1, except that it
affects only segments with at least one assembly
instruction, not all segments. The argument is useful
for protected mode programs, where code
segments must remain read-only.
SEGMENT: segSize Allows global default segment size to be set. Also
determines the default address size for external
symbols defined outside any segment. The segSize
can be USE16, USE32, or FLAT.

Conditional Directives
MASM 6.1 provides conditional-assembly directives and conditional-error
directives. Conditional-assembly directives let you test for a specified condition
and assemble a block of statements if the condition is true. Conditional-error
directives allow you to test for a specified condition and generate an assembly
error if the condition is true.
Both kinds of conditional directives test assembly-time conditions, not run-time
conditions. You can test only expressions that evaluate to constants during
assembly. For a list of the predefined symbols often used in conditional
assembly, see “Predefined Symbols,” earlier in this chapter.
Conditional-Assembly Directives
The IF and ENDIF directives enclose the conditional statements. The optional
ELSEIF and ELSE blocks follow the IF directive. There are many forms of the
IF and ELSE directives. Help provides a complete list.
The following statements show the syntax for the IF directives. The syntax for
other condition-assembly directives follow the same form.
IF expression1
ifstatements
[[ELSEIF expression2
elseifstatements]]
[[ELSE
elsestatements]]
ENDIF
The statements within an IF block can be any valid instructions, including other
conditional blocks, which in turn can contain any number of ELSEIF blocks.
ENDIF ends the block.
MASM assembles the statements following the IF directive only if the
corresponding condition is true. If the condition is not true and the block
contains an ELSEIF directive, the assembler checks to see if the corresponding
condition is true. If so, it assembles the statements following the ELSEIF
directive. If no IF or ELSEIF conditions are satisfied, the assembler processes
only the statements following the ELSE directive.
For example, you may want to assemble a line of code only if your program
defines a particular variable. In this example,
IFDEF buffer
buff BYTE buffer DUP(?)
ENDIF
the assembler allocates buff only if buffer has been previously defined.

MASM 6.1 provides the directives IF1, IF2, ELSEIF1, and ELSIF2 to grant
assembly only on pass one or pass two. To use these directives, you must either
enable 5.1 compatibility (with the /Zm command-line switch or OPTION
M510) or set OPTION SETIF2:TRUE, as described in the previous section.
The following list summarizes the conditional-assembly directives:
The Directive Grants Assembly If
IF expression expression is true (nonzero)
IFE expression expression is false (zero)
IFDEF name name has been previously defined
IFNDEF name name has not been previously defined
IFB argument* argument is blank
IFNB argument* argument is not blank
IFIDN[I] arg1, arg2* arg1 equals arg2
IFDIF[I] arg1, arg2* arg1 does not equal arg2
The optional I suffix (IFIDNI and IFDIFI) makes comparisons
insensitive to differences in case.
* Used only in macros.
Conditional-Error Directives
You can use conditional-error directives to debug programs and check for
assembly-time errors. By inserting a conditional-error directive at a key point in
your code, you can test assembly-time conditions at that point. You can also use
conditional-error directives to test for boundary conditions in macros.
Like other severe errors, those generated by conditional-error directives cause
the assembler to return a nonzero exit code. If MASM encounters a severe error
during assembly, it does not generate the object module.
For example, the .ERRNDEF directive produces an error if the program has not
defined a given label. In the following example, .ERRNDEF makes sure a label
called publevel actually exists.
.ERRNDEF publevel
IF publevel LE 2
PUBLIC var1, var2
ELSE
PUBLIC var1, var2, var3
ENDIF
The conditional-error directives use the syntax given in the previous section.
The following list summarizes the conditional-error directives. Note their close
correspondence with the previous list of conditional-assembly directives.

The Directive Generates an Error

.ERR Unconditionally where it occurs in the source file. Usually placed
within a conditional-assembly block.
.ERRE expression If expression is false (zero).
.ERRNZ expression If expression is true (nonzero).
.ERRDEF name If name has been defined.
.ERRNDEF name If name has not been defined.
.ERRB argument* If argument is blank.
.ERRNB argument* If argument is not blank.
.ERRIDN[I] arg1, arg2* If arg1 equals arg2.
.ERRDIF[I] arg1, arg2* If arg1 does not equal arg2.
The optional I suffix (.ERRIDNI and .ERRDIFI) makes
comparisons insensitive to case.
* Used only in macros
Two special conditional-error directives, .ERR1 and .ERR2, generate an error

only on pass one or pass two. To use these directives, you must either enable
5.1 compatibility (with the /Zm command-line switch or OPTION M510) or set
OPTION SETIF2:TRUE, as described in the previous section.

31
C H A P T E R 2
Organizing Segments
Understanding segments is an essential part of programming in assembly

language. In the family of 8086-based processors, the term segment has two
meanings:
u A block of memory of discrete size, called a “physical segment.” The
number of bytes in a physical memory segment is 64K for 16-bit processors
or 4 gigabytes for 32-bit processors.
u A variable-sized block of memory, called a “logical segment,” occupied by a
program’s code or data.
As you read this chapter, the distinction between the two definitions will become
clear. The adjectives “physical” and “logical” are not often used when speaking
of segments. The beginning programmer is left to infer from context which
definition applies. Fortunately, this is not difficult, and a distinction is often not
required.
This chapter begins with a close look at physical memory segments. This lays
the foundation for understanding logical segments, which form the subject of
most of the following sections.
The section “Using Simplified Segment Directives” explains how to begin, end,
and organize segments. It also explains how to access far data and code with
simplified segment directives.
The next section, “Using Full Segment Definitions,” describes how to order,
combine, and divide segments, and how to use the SEGMENT directive to
define full segments. It also explains how to create a segment group so that you
can use one segment address to access all the data.
Most of the information in this chapter also applies to writing modules to be
called from other programs. Exceptions are noted when they apply. For more
information about multiple-module programming, see Chapter 8, “Sharing Data
and Procedures Among Modules and Libraries.”

Physical Memory Segments

As explained in Chapter 1, a physical segment can begin only at memory
locations evenly divisible by 16, including address 0. Intel calls such locations
“paragraphs.” You can easily recognize a paragraph location because its
hexadecimal address always ends with 0, as in 10000h or 2EA70h. The
8086/286 processors allow segments 64K in size, the largest number 16 bits can
represent. The 80386/486 processors still adhere to the 64K limit when running
in real mode. In protected mode, however, they use 32-bit registers that can
hold addresses up to 4 gigabytes.
Segmented architecture presents certain hurdles for the assembly-language
programmer. For small programs, the limitations lose importance. Code and data
each occupy less than 64K and reside in individual segments. A simple offset
locates each variable or instruction within a segment.
Larger programs, however, must contend with problems of segmented memory
areas. If data occupies two or more segments, the program must specify both
segment and offset to access a variable. When the data forms a continuous
stream across segments — such as the text in a word processor’s workspace —
the problems become more acute. Whenever it adds or deletes text in the first
segment, the word processor must seamlessly move data back and forth over
the boundaries of each following segment.
The problem of segment boundaries disappears in the so-called flat address
space of 32-bit protected mode. Although segments still exist, they easily hold all
the code and data of the largest programs. Even a very large program becomes
in effect a small application, able to reach all code and data with a single offset
address.
Logical Segments
Logical segments contain the three components of a program: code, data, and
stack. MASM organizes the three parts for you so they occupy physical
segments of memory. The segment registers CS, DS, and SS contain the
addresses of the physical memory segments where the logical segments reside.
You can define segments in two ways: with simplified segment directives and
with full segment definitions. You can also use both kinds of segment definitions
in the same program.
Simplified segment directives hide many of the details of segment definition and
assume the same conventions used by Microsoft high-level languages. (See the
following section, “Using Simplified Segment Directives.”) The simplified
segment directives generate necessary code, specify segment attributes, and
arrange segment order.

Chapter 2 Organizing Segments 33
Full segment definitions require more complex syntax but provide more
complete control over how the assembler generates segments. (See “Using Full
Segment Definitions” later in this chapter.) If you use full segment definitions,
you must write code to handle all the tasks performed automatically by the
simplified segment directives.
Using Simplified Segment Directives

Structuring a MASM program using simplified segments requires use of several
directives to assign standard names, alignment, and attributes to the segments in
your program. These directives define the segments in such a way that linking
with Microsoft high-level languages is easy.
The simplified segment directives are .MODEL, .CODE, .CONST, .DATA,
.DATA?, .FARDATA, .FARDATA?, .STACK, .STARTUP, and .EXIT. The
following sections discuss these directives and the arguments they take.
MASM programs consist of modules made up of segments. Every program
written only in MASM has one main module, where program execution begins.
This main module can contain code, data, or stack segments defined with all of
the simplified segment directives. Any additional modules should contain only
code and data segments. Every module that uses simplified segments must,
however, begin with the .MODEL directive.
The following example shows the structure of a main module using simplified
segment directives. It uses the default processor (8086) and the default stack
distance (NEARSTACK). Additional modules linked to this main program
would use only the .MODEL, .CODE, and .DATA directives and the END
statement.

; This is the structure of a main module

; using simplified segment directives
.MODEL small, c ; This statement is required before you

; can use other simplified segment directives
.STACK ; Use default 1-kilobyte stack
.DATA ; Begin data segment
; Place data declarations here
.CODE ; Begin code segment

.STARTUP ; Generate start-up code
; Place instructions here
.EXIT ; Generate exit code

END
The .DATA and .CODE statements do not require any separate statements to
define the end of a segment. They close the preceding segment and then open a
new segment. The .STACK directive opens and closes the stack segment but
does not close the current segment. The END statement closes the last segment
and marks the end of the source code. It must be at the end of every module.
Defining Basic Attributes with .MODEL

The .MODEL directive defines the attributes that affect the entire module:
memory model, default calling and naming conventions, operating system, and
stack type. This directive enables use of simplified segments and controls the
name of the code segment and the default distance for procedures.
You must place .MODEL in your source file before any other simplified
segment directive. The syntax is:
.MODEL memorymodel [[, modeloptions ]]
The memorymodel field is required and must appear immediately after the
.MODEL directive. The use of modeloptions, which define the other attributes,
is optional. The modeloptions must be separated by commas. You can also use
equates passed from the ML command line to define the modeloptions.
The following list summarizes the memorymodel field and the modeloptions
fields, which specify language and stack distance:

Field Description
Memory model TINY, SMALL, COMPACT, MEDIUM , LARGE , HUGE , or
FLAT. Determines size of code and data pointers. This field is
required.
Language C, BASIC, FORTRAN, PASCAL, SYSCALL, or STDCALL. Sets
calling and naming conventions for procedures and public symbols.
Stack distance NEARSTACK or FARSTACK. Specifying NEARSTACK groups
the stack segment into a single physical segment (DGROUP) along with
data. SS is assumed to equal DS. FARSTACK does not group the
stack with DGROUP; thus SS does not equal DS.

You can use no more than one reserved word from each field. The following
examples show how you can combine various fields:
.MODEL small ; Small memory model
.MODEL large, c, farstack ; Large memory model,
; C conventions,
; separate stack
.MODEL medium, pascal ; Medium memory model,
; Pascal conventions,
; near stack (default)
The next four sections give more detail on each field.
Defining the Memory Model

MASM supports the standard memory models used by Microsoft high-level
languages — tiny, small, medium, compact, large, huge, and flat. You specify the
memory model with attributes of the same name placed after the .MODEL
directive. With the exception of the flat model, which requires instructions
specific to the 80386/486, your choice of a memory model does not limit the
kind of instructions you can write. The memory model does, however, control
segment defaults and determine whether data and code are near or far by
default, as indicated in the following table.
Table 2.1 Attributes of Memory Models
Memory Default Default Operating Data and Code
Model Code Data System Combined
Tiny Near Near MS-DOS Yes
Small Near Near MS-DOS, Windows No
Medium Far Near MS-DOS, Windows No
Compact Near Far MS-DOS, Windows No
Large Far Far MS-DOS, Windows No
Huge Far Far MS-DOS, Windows No
Flat Near Near Windows NT Yes
When writing assembler modules for a high-level language, you should use the
same memory model as the calling language. Choose the smallest memory
model available that can contain your data and code, since near references
operate more efficiently than far references.
The predefined symbol @Model returns the memory model, encoding memory
models as integers 1 through 7. For more information on predefined symbols,
see “Predefined Symbols” in Chapter 1. For an example of how to use them,
see Help.

The seven memory models supported by MASM 6.1 fall into three groups,
described in the following paragraphs.
Small, Medium, Compact, Large, and Huge Models

The traditional memory models recognized by many languages are small,
medium, compact, large, and huge. Small model supports one data segment and
one code segment. All data and code are near by default. Large model supports
multiple code and multiple data segments. All data and code are far by default.
Medium and compact models are in-between. Medium model supports multiple
code and single data segments; compact model supports multiple data segments
and a single code segment.
Huge model implies individual data items larger than a single segment, but the
implementation of huge data items must be coded by the programmer. Since the
assembler provides no direct support for this feature, huge model is essentially
the same as large model.
In each of these models, you can override the default. For example, you can
make large data items far in small model, or internal procedures near in large
model.
Tiny Model
Tiny-model programs run only under MS-DOS. Tiny model places all data and
code in a single segment. Therefore, the total program file size can occupy no
more than 64K. The default is near for code and static data items; you cannot
override this default. However, you can allocate far data dynamically at run time
using MS-DOS memory allocation services.
Tiny model produces MS-DOS .COM files. Specifying .MODEL tiny
automatically sends the /TINY argument to the linker. Therefore, the /AT
argument is not necessary with .MODEL tiny. However, /AT does not insert a
.MODEL directive. It only verifies that there are no base or pointer fixups, and
sends /TINY to the linker.
Flat Model
The flat memory model is a nonsegmented configuration available in 32-bit
operating systems. It is similar to tiny model in that all code and data go in a
single 32-bit segment.
To write a flat model program, specify the .386 or .486 directive before
.MODEL FLAT. All data and code (including system resources) are in a single
32-bit segment. The operating system automatically initializes segment registers
at load time; you need to modify them only when mixing 16-bit and 32-bit
segments in a single application. CS, DS, ES, and SS all occupy the supergroup
FLAT. Addresses and pointers passed to system services are always 32-bit near
addresses and pointers.

Choosing the Language Convention

The language option facilitates compatibility with high-level languages by
determining the internal encoding for external and public symbol names, the
code generated for procedure initialization and cleanup, and the order that
arguments are passed to a procedure with INVOKE. It also facilitates
compatibility with high-level – language modules. The PASCAL, BASIC, and
FORTRAN conventions are identical. C and SYSCALL have the same calling
convention but different naming conventions. Functions in the Windows API
use the Pascal calling convention.
Procedure definitions (PROC) and high-level procedure calls (INVOKE)
automatically generate code consistent with the calling convention of the
specified language. The PROC, INVOKE, PUBLIC, and EXTERN directives
all use the naming convention of the language. These directives follow the
default language conventions from the .MODEL directive unless you
specifically override the default. Use of these directives is explained in
“Controlling Program Flow,” Chapter 7. You can also use the OPTION
directive to set the language type. (See “Using the OPTION Directive” in
Chapter 1.) Not specifying a language type in either the .MODEL, OPTION,
EXTERN, PROC, INVOKE, or PROTO statement causes the assembler to
generate an error.
The predefined symbol @Interface provides information about the language
parameters. For a description of the bit flags, see Help.
For more information on calling and naming conventions, see Chapter 12,
“Mixed-Language Programming.” For information about writing procedures and
prototypes, see Chapter 7, “Controlling Program Flow.” For information on
multiple-module programming, refer to Chapter 8, “Sharing Data and
Procedures Among Modules and Libraries.”
Setting the Stack Distance

The NEARSTACK keyword places the stack segment in the group DGROUP
along with the data segment. The .STARTUP directive then generates code to
adjust SS:SP so that SS (Stack Segment register) holds the same address as DS
(Data Segment register). If you do not use .STARTUP, you must make this
adjustment or your program may fail to run. (For information about startup
code, see “Starting and Ending Code with .STARTUP and .EXIT,” later in this
chapter.) In this case, you can use DS to access stack items (including
parameters and local variables) and SS to access near data. Furthermore, since
stack items share the same segment address as near data, you can reliably pass
near pointers to stack items.
The FARSTACK setting gives the stack a segment of its own. That is, SS does
not equal DS. The default stack type, NEARSTACK, is a convenient setting for

most programs. Use FARSTACK for special cases such as memory-resident

programs

and dynamic-link libraries (discussed in Chapters 10 and 11) when you cannot
assume that the caller’s stack is near. You can use the predefined symbol
@Stack to determine if the stack location is DGROUP (for near stacks) or
STACK (for far stacks).
Specifying a Processor and Coprocessor

MASM supports a set of directives for selecting processors and coprocessors.
Once you select a processor, you must use only the instruction set for that
processor. The default is the 8086 processor. If you always want your code to
run on this processor, you do not need to add any processor directives.
To enable a different processor mode and the additional instructions available on
that processor, use the directives .186, .286, .386, and .486. The instruction
timings on a listing (see Appendix C, “Generating and Reading Assembly
Listings”) correspond to whichever processor directive you select.
The .286P, .386P, and .486P directives enable the instructions available only at
higher privilege levels in addition to the normal instruction set for the given
processor. Generally, you don’t need privileged instructions unless you are
writing operating-systems code or device drivers.
In addition to enabling different instruction sets, the processor directives also
affect the behavior of extended language features. For example, the INVOKE
directive pushes arguments onto the stack. If the .286 directive is in effect,
INVOKE takes advantage of operations possible only on 80286 and later
processors.
Use the directives .8087 (the default), .287, .387, and .NO87 to select a math
coprocessor instruction set. The .NO87 directive turns off assembly of all
coprocessor instructions. Note that .486 also enables assembly of all
coprocessor instructions because the 80486 processor has a complete set of
coprocessor registers and instructions built into the chip. The processor
instructions imply the corresponding coprocessor directive. The coprocessor
directives are provided to override the defaults.
Creating a Stack
The stack is the section of memory used for pushing or popping registers and
storing the return address when a subroutine is called. The stack often holds
temporary and local variables.
If your main module is written in a high-level language, that language handles
the details of creating a stack. Use the .STACK directive only when you write a
main module in assembly language.

The .STACK directive creates a stack segment. By default, the assembler

allocates 1K of memory for the stack. This size is sufficient for most small
programs.

To create a stack of a size other than the default size, give .STACK a single
numeric argument indicating stack size in bytes:
.STACK 2048 ; Use 2K stack
For a description of how stack memory is used with procedure calls and local
variables, see Chapter 7, “Controlling Program Flow.”
Creating Data Segments

Programs can contain both near and far data. In general, you should place
important and frequently used data in the near data area, where data access is
faster. This area can get crowded, however, because in 16-bit operating systems
the total amount of all near data in all modules cannot exceed 64K. Therefore,
you may want to place infrequently used or particularly large data items in a far
data segment.
The .DATA, .DATA?, .CONST, .FARDATA, and .FARDATA? directives
create data segments. You can access the various segments within DGROUP
without reloading segment registers (see “Defining Segment Groups,” later in
this chapter). These five directives also prevent instructions from appearing in
data segments by assuming CS to ERROR.
Near Data Segments

The .DATA directive creates a near data segment. This segment contains the
frequently used data for your program. It can occupy up to 64K in MS-DOS or
512 megabytes under flat model in Windows NT. It is placed in a special group
identified as DGROUP, which is also limited to 64K.
When you use .MODEL, the assembler automatically defines DGROUP for
your near data segment. The segments in DGROUP form near data, which can
normally be accessed directly through DS or SS.
You can also define the .DATA? and .CONST segments that go into DGROUP
unless you are using flat model. Although all of these segments (along with the
stack) are eventually grouped together and handled as data segments, .DATA?
and .CONST enhance compatibility with Microsoft high-level languages. In
Microsoft languages, .CONST is used to define constant data such as strings
and floating-point numbers that must be stored in memory. The .DATA?
segment is used for storing uninitialized variables. You can follow this
convention if you want. If you use C startup code, .DATA? is initialized to 0.
You can use @data to determine the group of the data segment and
@DataSize to determine the size of the memory model set by the .MODEL
directive. The predefined symbols @WordSize and @CurSeg return the size
attribute and name of the current segment, respectively. See “Predefined
Symbols” in Chapter 1.

Far Data Segments

The compact, large, and huge memory models use far data addresses by default.
With these memory models, however, you can still construct data segments
using .DATA, .DATA?, and .CONST. The effect of these directives does not
change from one memory model to the next. They always contribute segments
to the default data area, DGROUP, which has a total limit of 64K.
When you use .FARDATA or .FARDATA? in the small and medium memory
models, the assembler creates far data segments FAR_DATA and FAR_BSS,
respectively. You can access variables with:
mov ax, SEG farvar2
mov ds, ax
For more information on far data, see “Near and Far Addresses” in Chapter 3.
Creating Code Segments

Whether you are writing a main module or a module to be called from another
module, you can have both near and far code segments. This section explains
how to use near and far code segments and how to use the directives and
predefined equates that relate to code segments.
Near Code Segments

The small memory model is often the best choice for assembly programs that
are not linked to modules in other languages, especially if you do not need more
than 64K of code. This memory model defaults to near (two-byte) addresses for
code and data, which makes the program run faster and use less memory.
When you use .MODEL and simplified segment directives, the .CODE
directive in your program instructs the assembler to start a code segment. The
next segment directive closes the previous segment; the END directive at the
end of your program closes remaining segments. The example at the beginning
of “Using Simplified Segment Directives,” earlier in this chapter, shows how to
do this.
You can use the predefined symbol @CodeSize to determine whether code
pointers default to NEAR or FAR.
Far Code Segments

When you need more than 64K of code, use the medium, large, or huge
memory model to create far segments.
The medium, large, and huge memory models use far code addresses by
default. In the larger memory models, the assembler creates a different code
segment for each module. If you use multiple code segments in the small,

compact, or tiny model, the linker combines the .CODE segments for all
modules into one segment.
For far code segments, the assembler names each code segment
MODNAME_TEXT, in which MODNAME is the name of the module. With
near code, the assembler names every code segment _TEXT, causing the linker
to concatenate these segments into one. You can override the default name by
providing an argument after .CODE. (For a complete list of segment names
generated by MASM, see Appendix E, “Default Segment Names.”)
With far code, a single module can contain multiple code segments. The .CODE
directive takes an optional text argument that names the segment. For instance,
the following example creates two distinct code segments, FIRST_TEXT and
SECOND_TEXT.
.CODE FIRST
.
. ; First set of instructions here
.
.CODE SECOND
.
. ; Second set of instructions here
.
Whenever the processor executes a far call or jump, it loads CS with the new
segment address. No special action is necessary other than making sure that you
use far calls and jumps. See “Near and Far Addresses” in Chapter 3.
Note The assembler always assumes that the CS register contains the address
of the current code segment or group.
Starting and Ending Code with .STARTUP and .EXIT

The easiest way to begin and end an MS-DOS program is to use the
.STARTUP and .EXIT directives in the main module. The main module
contains the starting point and usually the termination point. You do not need
these directives in a module called by another module.
These directives make MS-DOS programs easy to maintain. They automatically
generate code appropriate to the stack distance specified with .MODEL.
However, they do not apply to flat-model programs written for 32-bit operating
systems. Thus, you should not use .STARTUP or .EXIT in programs written
for Windows NT.

To start a program, place the .STARTUP directive where you want execution to
begin. Usually, this location immediately follows the .CODE directive:
.CODE
.STARTUP
.
. ; Place executable code here
.
.EXIT
END
Note that .EXIT generates executable code, while END does not. The END
directive informs the assembler that it has reached the end of the module. All
modules must end with the END directive whether you use simplified or full
segments.
If you do not use .STARTUP, you must give the starting address as an
argument to the END directive. For example, the following fragment shows how
to identify a program’s starting instruction with the label start:
.CODE
start:
.
. ; Place executable code here
.
END start
Only the END directive for the module with the starting instruction should have
an argument. When .STARTUP is present, the assembler ignores any argument
to END.
For the default NEARSTACK attribute, .STARTUP points DS to DGROUP
and sets SS:SP relative to DGROUP, generating the following code:

@Startup:
mov dx, DGROUP
mov ds, dx
mov bx, ss
sub bx, dx
shl bx, 1 ; If .286 or higher, this is
shl bx, 1 ; shortened to shl bx, 4
shl bx, 1
shl bx, 1
cli ; Not necessary in .286 or higher
mov ss, dx
add sp, bx
sti ; Not necessary in .286 or higher
.
.
.
END @Startup
An MS-DOS program with the FARSTACK attribute does not need to adjust
SS:SP, so .STARTUP just initializes DS, like this:
@Startup:
mov dx, DGROUP
mov ds, dx
.
.
.
END @Startup
When the program terminates, you can return an exit code to the operating
system. Applications that check exit codes usually assume that an exit code of 0
means no problem occurred, and that an exit code of 1 means an error
terminated the program. The .EXIT directive accepts a 1-byte exit code as its
optional argument:
.EXIT 1 ; Return exit code 1
.EXIT generates the following code that returns control to MS-DOS, thus
terminating the program. The return value, which can be a constant, memory
reference, or 1-byte register, goes into AL:
mov al, value
mov ah, 04Ch
int 21h
If your program does not specify a return value, .EXIT returns whatever value
happens to be in AL.

Using Full Segment Definitions

If you need complete control over segments, you can fully define the segments
in your program. This section explains segment definitions, including how to
order segments and how to define the segment types.
If you write a program under MS-DOS without .MODEL and .STARTUP, you
must initialize registers yourself and use the END directive to indicate the
starting address. The Windows operating system does not require you to
initialize registers, as described in Chapter 3. For a description of typical startup
code, see “Controlling the Segment Order,” later in this chapter.
Defining Segments with the SEGMENT Directive

A defined segment begins with the SEGMENT directive and ends with the
ENDS directive:
name SEGMENT [[align]] [[READONLY]] [[combine]] [[use]] [[’class’]]
statements
name ENDS
The name defines the name of the segment. Within a module, all segment
definitions with the same name are treated as though they reference the same
segment. The linker also combines identically named segments from different
modules unless the combine type is PRIVATE. In addition, segments can be
nested.
The optional types that follow the SEGMENT directive give the linker and the
assembler instructions on how to set up and combine segments. The optional
types, which are explained in detail in the following sections, include:
Type Description
align Defines the memory boundary on which a new segment begins.
READONLY Tells the assembler to report an error if it detects an instruction
modifying any item in a READONLY segment.
combine Determines how the linker combines segments from different modules
when building executable files.
use (80386/486 only) Determines the size of a segment. USE16 indicates that offsets in the
segment are 16 bits wide. USE32 indicates 32-bit offsets.
class Provides a class name for the segment. The linker automatically
groups segments of the same class in memory.
Types can be specified in any order. You can specify only one attribute from
each of these fields; for example, you cannot have two different align types.

You can close a segment and reopen it later with another SEGMENT directive.
When you reopen a segment, you need only give the segment name. You
cannot change the attributes of a segment once you have defined it.
Note The PAGE align type and the PUBLIC combine type are distinct from
the PAGE and PUBLIC directives. The assembler distinguishes them by means
of context.
Aligning Segments
The optional align type in the SEGMENT directive defines the range of
memory addresses from which a starting address for the segment can be
selected. The align type can be any of the following:
Align Type Starting Address
BYTE Next available byte address.
WORD Next available word address.
DWORD Next available doubleword address.
PARA Next available paragraph address (16 bytes per paragraph). Default.
PAGE Next available page address (256 bytes per page).
The linker uses the alignment information to determine the relative starting
address for each segment. The operating system calculates the actual starting
address when the program is loaded.
Making Segments Read-Only

The optional READONLY attribute is helpful when creating read-only code
segments for protected mode, or when writing code to be placed in read-only
memory (ROM). It protects against illegal self-modifying code.
The READONLY attribute causes the assembler to check for instructions that
modify the segment and to generate an error if it finds any. The assembler
generates an error if you attempt to write directly to a read-only segment.
Combining Segments
The optional combine type in the SEGMENT directive defines how the linker
combines segments having the same name but appearing in different modules.

The combine type controls linker behavior, not assembler behavior. The
combine types, which are described in full detail in Help, include:
Combine Type Linker Action
PRIVATE Does not combine the segment with segments from other modules,
even if they have the same name. Default.
PUBLIC Concatenates all segments having the same name to form a single,
contiguous segment.
STACK Concatenates all segments having the same name and causes the
operating system to set SS:00 to the bottom and SS:SP to the top
of the resulting segment. Data initialization is unreliable, as
discussed following.
COMMON Overlaps segments. The length of the resulting area is the length of
the largest of the combined segments. Data initialization is
unreliable, as discussed following.
MEMORY Used as a synonym for the PUBLIC combine type.
AT address Assumes address as the segment location. An AT segment cannot
contain any code or initialized data, but is useful for defining
structures or variables that correspond to specific far memory
locations, such as a screen buffer or low memory.
You cannot use the AT combine type in protected-mode
programs.
Do not place initialized data in STACK or COMMON segments. With these

combine types, the linker overlays initialized data for each module at the
beginning of the segment. The last module containing initialized data writes over
any data from other modules.
Note Normally, you should provide at least one stack segment (having STACK
combine type) in a program. If no stack segment is declared, LINK displays a
warning message. You can ignore this message if you have a specific reason for
not declaring a stack segment. For example, you would not have a separate
stack segment in a MS-DOS tiny model (.COM) program, nor would you need
a separate stack in a DLL that uses the caller’s stack.
Setting Segment Word Sizes (80386/486 Only)

The use type in the SEGMENT directive specifies the segment word size on the
80386/486 processors. Segment word size determines the default operand and
address size of all items in a segment.
The size attribute can be USE16, USE32, or FLAT. If you specify the .386 or
.486 directive before the .MODEL directive, USE32 is the default. This
attribute specifies that items in the segment are addressed with a 32-bit offset
rather than a

16-bit offset. If .MODEL precedes the .386 or .486 directive, USE16 is the
default. To make USE32 the default, put .386 or .486 before .MODEL. You
can override the USE32 default with the USE16 attribute, or vice versa.
Note Programs written for MS-DOS must not specify USE32. Mixing 16-bit
and 32-bit segments in the same program is possible but usually applies only to
systems programming.
Setting Segment Order with Class Type

The optional class type in the SEGMENT directive helps control segment
ordering. Two segments with the same name are not combined if their class is
different. The linker arranges segments so that all segments identified with a
given class type are next to each other in the executable file. However, within a
particular class, the linker arranges segments in the order encountered. The
.ALPHA, .SEQ, or .DOSSEG directive determines this order in each .OBJ file.
The most common method for specifying a class type is to place all code
segments first in the executable file.
Controlling the Segment Order

The assembler normally positions segments in the object file in the order in
which they appear in source code. The linker, in turn, processes object files in
the order in which they appear on the command line. Within each object file, the
linker outputs segments in the order they appear, subject to any group, class,
and .DOSSEG requirements.
You can usually ignore segment ordering. However, it is important whenever
you want certain segments to appear at the beginning or end of a program or
when you make assumptions about which segments are next to each other in
memory. For tiny model (.COM) programs, code segments must appear first in
the executable file, because execution must start at the address 100h.
Segment Order Directives

You can control the order in which segments appear in the executable program
with three directives. The default, .SEQ, arranges segments in the order in
which you declare them.
The .ALPHA directive specifies alphabetical segment ordering within a module.
.ALPHA is provided for compatibility with early versions of the IBM
assembler. If you have trouble running code from older books on assembly
language, try using .ALPHA.
The .DOSSEG directive specifies the MS-DOS segment-ordering convention. It
places segments in the standard order required by Microsoft languages. Do not
use .DOSSEG in a module to be called from another module.

The .DOSSEG directive orders segments as follows:

1. Code segments
2. Data segments, in this order:
a. Segments not in class BSS or STACK
b. Class BSS segments
c. Class STACK segments
When you declare two or more segments to be in the same class, the linker
automatically makes them contiguous. This rule overrides the segment-ordering
directives. (For more about segment classes, see “Setting Segment Order with
Class Type” in the previous section.)
Linker Control
Most of the segment-ordering techniques (class names, .ALPHA, and .SEQ)
control the order in which the assembler outputs segments. Usually, you are
more interested in the order in which segments appear in the executable file. The
linker controls this order.
The linker processes object files in the order in which they appear on the
command line. Within each module, it then outputs segments in the order given
in the object file. If the first module defines segments DSEG and STACK and
the second module defines CSEG, then CSEG is output last. If you want to
place CSEG first, there are two ways to do so.
The simpler method is to use .DOSSEG. This directive is output as a special
record to the object file linker, and it tells the linker to use the Microsoft
segment-ordering convention. This convention overrides command-line order of
object files, and it places all segments of class 'CODE' first. (See “Defining
Segments with the SEGMENT Directive,” previous.)
The other method is to define all the segments as early as possible (in an include
file, for example, or in the first module). These definitions can be “dummy
segments” — that is, segments with no content. The linker observes the segment
ordering given, then later combines the empty segments with segments in other
modules that have the same name.

For example, you might include the following at the start of the first module of
your program or in an include file:
_TEXT SEGMENT WORD PUBLIC 'CODE'
_TEXT ENDS
_DATA SEGMENT WORD PUBLIC 'DATA'
_DATA ENDS
CONST SEGMENT WORD PUBLIC 'CONST'
CONST ENDS
STACK SEGMENT PARA STACK 'STACK'
STACK ENDS
Later in the program, the order in which you write _TEXT, _DATA, or other
segments does not matter because the ultimate order is controlled by the
segment order defined in the include file.
Setting the ASSUME Directive for Segment Registers

Many of the assembler instructions assume a default segment. For example,
JMP assumes the segment associated with the CS register, PUSH and POP
assume the segment associated with the SS register, and MOV instructions
assume the segment associated with the DS register.
When the assembler needs to reference an address, it must know what segment
contains the address. It finds this by using the default segment or group
addresses assigned with the ASSUME directive. The syntax is:
ASSUME segregister : seglocation [, segregister : seglocation] ]
ASSUME dataregister : qualifiedtype [, dataregister : qualifiedtype]
ASSUME register : ERROR [, register : ERROR]
ASSUME [register :] NOTHING [, register : NOTHING]
ASSUME register : FLAT [, register : FLAT]
The seglocation must be the name of the segment or group that is to be
associated with segregister. Subsequent instructions that assume a default
register for referencing labels or variables automatically assume that if the
default segment is segregister, the label or variable is in the seglocation. MASM
6.1 automatically gives CS the address of the current code segment. Therefore,
you do not need to include
ASSUME CS : MY_CODE
at the beginning of your program if you want the current segment associated
with CS.

Note Using the ASSUME directive to tell the assembler which segment to
associate with a segment register is not the same as telling the processor. The
ASSUME directive affects only assembly-time assumptions. You may need to
use instructions to change run-time conditions. Initializing segment registers at
run time is discussed in “Informing the Assembler About Segment Values,”
Chapter 3.
The ASSUME directive can define a segment for each of the segment registers.
The segregister can be CS, DS, ES, or SS (and FS and GS on the 80386/486).
The seglocation must be one of the following:
u The name of a segment defined in the source file with the SEGMENT
directive.
u The name of a group defined in the source file with the GROUP directive.
u The keyword NOTHING, ERROR, or FLAT.
u A SEG expression (see “Immediate Operands” in Chapter 3).
u A string equate (text macro) that evaluates to a segment or group name (but
not a string equate that evaluates to a SEG expression).
It is legal to combine assumes to FLAT with assumes to specific segments.

Combinations might be necessary in operating-system code that handles both
16- and 32-bit segments.
The keyword NOTHING cancels the current segment assumptions. For
example, the statement ASSUME NOTHING cancels all register assumptions
made by previous ASSUME statements.
Usually, a single ASSUME statement defines all four segment registers at the
start of the source file. However, you can use the ASSUME directive at any
point to change segment assumptions.
Using the ASSUME directive to change segment assumptions is often equivalent
to changing assumptions with the segment-override operator (:). See “Direct
Memory Operands” in Chapter 3. The segment-override operator is more
convenient for one-time overrides. The ASSUME directive may be more
convenient if previous assumptions must be overridden for a sequence of
instructions.
However, in either case, your program must explicitly load a segment register
with a segment address before accessing data within the segment. ASSUME
only tells the assembler to assume that the register is correctly initialized; it does
not by itself generate any code to load the register.

You can also prevent the use of a register with:

ASSUME SegRegister : ERROR
The assembler generates an ASSUME CS:ERROR when you use simplified

directives to create data segments, effectively preventing instructions or code
labels from appearing in a data segment.
For more information about ASSUME, refer to “Defining Register Types with
ASSUME” in Chapter 3.
Defining Segment Groups

A group is a collection of segments totalling not more than 64K in 16-bit mode.
A program addresses a code or data item in the group relative to the beginning
of the group.
A group lets you develop separate logical segments for different kinds of data
and then combine these into one segment (a group) for all the data. Using a
group can save you from having to continually reload segment registers to
access different segments. As a result, the program uses fewer instructions and
runs faster.
The most common example of a group is the specially named group for near
data, DGROUP. In the Microsoft segment model, several segments (_DATA,
_BSS, CONST, and STACK) are combined into a single group called
DGROUP. Microsoft high-level languages place all near data segments in this
group. (By default, the stack is placed here, too.) The .MODEL directive
automatically defines DGROUP. The DS register normally points to the
beginning of the group, giving you relatively fast access to all data in DGROUP.
The syntax of the group directive is:
name GROUP segment [[, segment]]...
The name labels the group. It can refer to a group that was previously defined.
This feature lets you add segments to a group one at a time. For example, if
MYGROUP was previously defined to include ASEG and BSEG, then the
statement
MYGROUP GROUP CSEG
is perfectly legal. It simply adds CSEG to the group MYGROUP; ASEG and BSEG
are not removed.
Each segment can be any valid segment name (including a segment defined later
in source code), with one restriction: a segment cannot belong to more than one
group.

The GROUP directive does not affect the order in which segments of a group
are loaded. You can place any number of 16-bit segments in a group as long as
the total size does not exceed 65,536 bytes. If the processor is in 32-bit mode,
the maximum size is 4 gigabytes. You need to make sure that non-grouped
segments do not get placed between grouped segments in such a way that the
size of the group exceeds 64K or 4 gigabytes. Neither can you place a 16-bit and
a 32-bit segment in the same group.

53
C H A P T E R 3
Using Addresses and Pointers
MASM applications running in real mode require segmented addresses to access

code and data. The address of the code or data in a segment is relative to a
segment address in a segment register. You can also use pointers to access data
in assembly language programs. (A pointer is a variable that contains an address
as its value.)
The first section of this chapter describes how to initialize default segment
registers to access near and far addresses. The next section describes how to
access code and data. It also describes related operators, syntax, and
displacements. The discussion of memory operands lays the foundation for the
third section, which describes the stack.
The fourth section of this chapter explains how to use the TYPEDEF directive
to declare pointers and the ASSUME directive to give the assembler information
about registers containing pointers. This section also shows you how to do
typical pointer operations and how to write code that works for pointer variables
in any memory model.
Programming Segmented Addresses

Before you use segmented addresses in your programs, you need to initialize the
segment registers. The initialization process depends on the registers used and
on your choice of simplified segment directives or full segment definitions. The
simplified segment directives (introduced in Chapter 2) handle most of the
initialization process for you. This section explains how to inform the assembler
and the processor of segment addresses, and how to access the near and far
code and data in those segments.
Initializing Default Segment Registers

The segmented architecture of the 8086-family of processors does not require
that you specify two addresses every time you access memory. As explained in
Chapter 2, “Organizing Segments,” the 8086 family of processors uses a system

of default segment registers to simplify access to the most commonly used data
and code.
The segment registers DS, SS, and CS are normally initialized to default
segments at the beginning of a program. If you write the main module in a high-
level language, the compiler initializes the segment registers. If you write the
main module in assembly language, you must initialize the segment registers
yourself. Follow these steps to initialize segments:
1. Tell the assembler which segment is associated with a register. The
assembler must know the default segments at assembly time.
2. Tell the processor which segment is associated with a register by writing the
necessary code to load the correct segment value into the segment register on
the processor.
These steps are discussed separately in the following sections.
Informing the Assembler About Segment Values

The first step in initializing segments is to tell the assembler which segment to
associate with a register. You do this with the ASSUME directive. If you use
simplified segment directives, the assembler automatically generates the
appropriate ASSUME statements. If you use full segment definitions, you must
code the ASSUME statements for registers other than CS yourself. (ASSUME
can also be used on general-purpose registers, as explained in “Defining Register
Types with ASSUME” later in this chapter.)
The .STARTUP directive generates startup code that sets DS equal to SS
(unless you specify FARSTACK), allowing default data to be accessed through
either SS or DS. This can improve efficiency in the code generated by
compilers. The “DS equals SS” convention may not work with certain
applications, such as memory-resident programs in MS-DOS and Windows
dynamic-link libraries (see Chapter 10). The code generated for .STARTUP is
shown in “Starting and Ending Code with .STARTUP and .EXIT” in Chapter 2.
You can use similar code to set DS equal to SS in programs using full segment
definitions.
Here is an example of ASSUME using full segment definitions:
ASSUME cs:_TEXT, ds:DGROUP, ss:DGROUP
This example is equivalent to the ASSUME statement generated with simplified

segment directives in small model with NEARSTACK. Note that DS and SS
are part of the same segment group. It is also possible to have different
segments for data and code, and to use ASSUME to set ES, as shown here:

Chapter 3 Using Addresses and Pointers 55
ASSUME cs:MYCODE, ds:MYDATA, ss:MYSTACK, es:OTHER
Correct use of the ASSUME statement can help find addressing errors. With
.CODE, the assembler assumes CS is the current segment. When you use the
simplified segment directives .DATA, .DATA?, .CONST, .FARDATA, or
.FARDATA?, the assembler automatically assumes CS is the ERROR
segment. This prevents instructions from appearing in these segments. If you
use full segment definitions, you can accomplish the same by placing ASSUME
CS:ERROR in a data segment.
With simple or full segments, you can cancel the control of an ASSUME
statement by assuming NOTHING. You can cancel the previous assumption for
ES with the following statement:
ASSUME es:NOTHING
Prior to the .MODEL statement (or in its absence), the assembler sets the
ASSUME statement for DS, ES, and SS to the current segment.
Informing the Processor About Segment Values

The second and final step in initializing segments is to inform the processor of
segment values at run time. How segment values are initialized at run time
differs for each segment register and depends on the operating system and on
your use of simplified segment directives or full segment definitions.
Specifying a Starting Address

A program’s starting address determines where execution begins. After the
operating system loads a program, it simply jumps to the starting address, giving
processor control to the program. The true starting address is known only to the
loader; the linker determines only the offset of the address within an
undetermined code segment. That’s why a normal application is often referred
to as “relocatable code,” because it runs regardless of where the loader places it
in memory.
The offset of the starting address depends on the program type. Programs with
an .EXE extension contain a header from which the loader reads the offset and
combines it with a segment to form the starting address. Programs with a .COM
extension (tiny model) have no such header, so by convention the loader jumps
to the first byte of the program.
In either case, the .STARTUP directive identifies where execution begins,
provided you use simplified segment directives. For an .EXE program, place
.STARTUP immediately before the instruction where you want execution to
start. In a .COM program, place .STARTUP before the first assembly
instruction in your source code.

If you use full segment directives or prefer not to use .STARTUP, you must
identify the starting instruction in two steps:
1. Label the starting instruction.
2. Provide the same label in the END directive.
These steps tell the linker where execution begins in the program. The following
example illustrates the two steps for a tiny model program:
_TEXT SEGMENT WORD PUBLIC 'CODE'
ORG 100h ; Use this declaration for .COM files only
start: . ; First instruction here
.
.
_TEXT ENDS
END start ; Name of starting label
Notice the ORG statement in this example. This statement is mandatory in a

tiny model program without the .STARTUP directive. It places the first
instruction at offset 100h in the code segment to create space for a 256-byte
(100h) data area called the Program Segment Prefix (PSP). The operating
system takes care of initializing the PSP, so you need only make sure the area
exists. (For a description of what data resides in the PSP, refer to the “Tables”
chapter in the Reference.)
Initializing DS
The DS register is automatically initialized to the correct value (DGROUP) if
you use .STARTUP or if you are writing a program for Windows. If you do not
use .STARTUP with MS-DOS, you must initialize DS using the following
instructions:
mov ax, DGROUP
mov ds, ax
The initialization requires two instructions because the segment name is a

constant and the assembler does not allow a constant to be loaded directly to a
segment register. The previous example loads DGROUP, but you can load any
valid segment or group.
Initializing SS and SP
The SS and SP registers are initialized automatically if you use the .STACK
directive with simplified segments or if you define a segment that has the
STACK combine type with full segment definitions. Using the STACK directive
initializes SS to the stack segment. If you want SS to be equal to DS, use
.STARTUP or its equivalent. (See “Combining Segments,” page 45.) For an
.EXE file, the stack address is encoded into the executable header and resolved

at load time. For a .COM file, the loader sets SS equal to CS and initializes SP
to 0FFFEh.
If your program does not access far data, you do not need to initialize the ES
register. If you choose to initialize, use the same technique as for the DS
register. You can initialize SS to a far stack in the same way.
Near and Far Addresses

Addresses that have an implied segment name or segment registers associated
with them are called “near addresses.” Addresses that have an explicit segment
associated with them are called “far addresses.” The assembler handles near and
far code automatically, as described in the following sections. You must specify
how to handle far data.
The Microsoft segment model puts all near data and the stack in a group called
DGROUP. Near code is put in a segment called _TEXT. Each module’s far
code or far data is placed in a separate segment. This convention is described in
“Controlling the Segment Order” in Chapter 2.
The assembler cannot determine the address for some program components;
these are said to be relocatable. The assembler generates a fixup record and the
linker provides the address once it has determined the location of all segments.
Usually a relocatable operand references a label, but there are exceptions.
Examples in the next two sections include information about relocating near and
far data.
Near Code
Control transfers within near code do not require changes to segment registers.
The processor automatically handles changes to the offset in the IP register
when control-flow instructions such as JMP, CALL, and RET are used. The
statement
call nearproc ; Change code offset
changes the IP register to the new address but leaves the segment unchanged.
When the procedure returns, the processor resets IP to the offset of the next
instruction after the CALL instruction.
Far Code
The processor automatically handles segment register changes when dealing with
far code. The statement
call farproc ; Change code segment and offset
automatically moves the segment and offset of the farproc procedure to the
CS and IP registers. When the procedure returns, the processor sets CS to the

original code segment and sets IP to the offset of the next instruction after the
call.
Near Data
A program can access near data directly, because a segment register already
holds the correct segment for the data item. The term “near data” is often used
to refer to the data in the DGROUP group.
After the first initialization of the DS and SS registers, these registers normally
point into DGROUP. If you modify the contents of either of these registers
during the execution of the program, you must reload the register with
DGROUP’s address before referencing any DGROUP data.
The processor assumes all memory references are relative to the segment in the
DS register, with the exception of references using BP or SP. The processor
associates these registers with the SS register. (You can override these
assumptions with the segment override operator, described in “Direct Memory
Operands,” on page 62.)
The following lines illustrate how the processor accesses either the DS or SS
segments, depending on whether the pointer operand contains BP or SP. Note
the distinction loses significance when DS and SS are equal.
nearvar WORD 0
.
.
.
mov ax, nearvar ; Reads from DS:[nearvar]
mov di, [bx] ; Reads from DS:[bx]
mov [di], cx ; Writes to DS:[di]
mov [bp+6], ax ; Writes to SS:[bp+6]
mov bx, [bp] ; Reads from SS:[bp]
Far Data
To read or modify a far address, a segment register must point to the segment of
the data. This requires two steps. First load the segment (normally either ES or
DS) with the correct value, and then (optionally) set an assume of the segment
register to the segment of the address.
Note Flat model does not require far addresses. By default, all addressing is
relative to the initial values of the segment registers. Therefore, this section on
far addressing does not apply to flat model programs.
One method commonly used to access far data is to initialize the ES segment
register. This example shows two ways to do this:

; First method
mov ax, SEG farvar ; Load segment of the
mov es, ax , far address into ES
mov ax, es:farvar ; Provide an explicit segment
; override on the addressing

; Second method
mov ax, SEG farvar2 ; Load the segment of the
mov es, ax ; far address into ES
ASSUME ES:SEG farvar2 ; Tell the assembler that ES points
; to the segment containing farvar2
mov ax, farvar2 ; The assembler provides the ES
; override since it knows that
; the label is addressable
After loading the segment of the address into the ES segment register, you can
explicitly override the segment register so that the addressing is correct (method
1) or allow the assembler to insert the override for you (method 2). The
assembler uses ASSUME statements to determine which segment register can
be used to address a segment of memory. To use the segment override operator,
the left operand must be a segment register, not a segment name. (For more
information on segment overrides, see “Direct Memory Operands” on page 62.)
If an instruction needs a segment override, the resulting code is slightly larger
and slower, since the override must be encoded into the instruction. However,
the resulting code may still be smaller than the code for multiple loads of the
default segment register for the instruction.
The DS, SS, FS, and GS segment registers (FS and GS are available only on the
80386/486 processors) may also be used for addressing through other segments.
If a program uses ES to access far data, it need not restore ES when finished
(unless the program uses flat model). However, some compilers require that you
restore ES before returning to a module written in a high-level language.
To access far data, first set DS to the far segment and then restore the original
DS when finished. Use the ASSUME directive to let the assembler know that
DS no longer points to the default data segment, as shown here:
push ds ; Save original segment
mov ax, SEG fararray ; Move segment into data register
mov ds, ax ; Initialize segment register
ASSUME ds:SEG fararray ; Tell assembler where data is
mov ax, fararray[0] ; Set DX:AX = dword variable
mov dx, fararray[2] ; fararray
.
.
.
pop ds ; Restore segment
ASSUME ds:@DATA ; and default assumption

“Direct Memory Operands,”on page 62, describes an alternative method for

accessing far data. The technique of resetting DS as shown in the previous
example is best for a lengthy series of far data references. The segment override
method described in “Direct Memory Operands” serves best when accessing
only one or two far variables.
If your program changes DS to access far data, it should restore DS when
finished. This allows procedures to assume that DS is the segment for near data.
Many compilers, including Microsoft compilers, use this convention.
Operands
With few exceptions, assembly language instructions work on sources of data
called operands. In a listing of assembly code (such as the examples in this
book), operands appear in the operand field immediately to the right of the
instructions.
This section describes the four kinds of instruction operands: register,
immediate, direct memory, and indirect memory. Some instructions, such as
POPF and STI, have implied operands which do not appear in the operand
field. Otherwise, an implied operand is just as real as one stated explicitly.
Certain other instructions such as NOP and WAIT deserve special mention.
These instructions affect only processor control and do not require an operand.
The following four types of operands are described in the rest of this section:
Operand Type Addressing Mode
Register An 8-bit or 16-bit register on the 8086–80486; can also be 32-bit on the
80386/486.
Immediate A constant value contained in the instruction itself.
Direct memory A fixed location in memory.
Indirect memory A memory location determined at run time by using the address stored in
one or two registers.
Instructions that take two or more operands always work right to left. The right
operand is the source operand. It specifies data that will be read, but not
changed, in the operation. The left operand is the destination operand. It
specifies the data that will be acted on and possibly changed by the instruction.

Register Operands
Register operands refer to data stored in registers. The following examples show
typical register operands:
mov bx, 10 ; Load constant to BX
add ax, bx ; Add BX to AX
jmp di ; Jump to the address in DI
An offset stored in a base or index register often serves as a pointer into

memory. You can store an offset in one of the base or index registers, then use
the register as an indirect memory operand. (See “Indirect Memory Operands,”
following.) For example:
mov [bx], dl ; Store DL in indirect memory operand
inc bx ; Increment register operand
mov [bx], dl ; Store DL in new indirect memory operand
This example moves the value in DL to 2 consecutive bytes of a memory

location pointed to by BX. Any instruction that changes the register value also
changes the data item pointed to by the register.
Immediate Operands
An immediate operand is a constant or the result of a constant expression. The
assembler encodes immediate values into the instruction at assembly time. Here
are some typical examples showing immediate operands:
mov cx, 20 ; Load constant to register
add var, 1Fh ; Add hex constant to variable
sub bx, 25 * 80 ; Subtract constant expression
Immediate data is never permitted in the destination operand. If the source

operand is immediate, the destination operand must be either a register or direct
memory to provide a place to store the result of the operation.
Immediate expressions often involve the useful OFFSET and SEG operators,
described in the following paragraphs.
The OFFSET Operator

An address constant is a special type of immediate operand that consists of an
offset or segment value. The OFFSET operator returns the offset of a memory
location, as shown here:
mov bx, OFFSET var ; Load offset address
For information on differences between MASM 5.1 behavior and MASM 6.1
behavior related to OFFSET, see Appendix A.

Since data in different modules may belong to a single segment, the assembler
cannot know for each module the true offsets within a segment. Thus, the offset
for var, although an immediate value, is not determined until link time.
The SEG Operator

The SEG operator returns the segment of a memory location:
mov ax, SEG farvar ; Load segment address
mov es, ax
The actual value of a particular segment is not known until the program is
loaded into memory. For .EXE programs, the linker makes a list in the
program’s header of all locations in which the SEG operator appears. The
loader reads this list and fills in the required segment address at each location.
Since .COM programs have no header, the assembler does not allow relocatable
segment expressions in tiny model programs.
The SEG operator returns a variable’s “frame” if it appears in the instruction.
The frame is the value of the segment, group, or segment override of a
nonexternal variable. For example, the instruction
mov ax, SEG DGROUP:var
places in AX the value of DGROUP, where var is located. If you do not

include a frame, SEG returns the value of the variable’s group if one exists. If
the variable is not defined in a group, SEG returns the variable’s segment
address.
This behavior can be changed with the /Zm command-line option or with the
OPTION OFFSET:SEGMENT statement. (See Appendix A, “Differences
between MASM 6.1 and 5.1.”) “Using the OPTION Directive” in Chapter 1
introduces the OPTION directive.
Direct Memory Operands

A direct memory operand specifies the data at a given address. The instruction
acts on the contents of the address, not the address itself. Except when size is
implied by another operand, you must specify the size of a direct memory
operand so the instruction accesses the correct amount of memory. The
following example shows how to explicitly specify data size with the BYTE
directive:

.DATA? ; Segment for uninitialized data

var BYTE ? ; Reserve one byte, labeled "var"
.CODE
.
.
.
mov var, al ; Copy AL to byte at var
Any location in memory can be a direct memory operand as long as a size is

specified (or implied) and the location is fixed. The data at the address can
change, but the address cannot. By default, instructions that use direct memory
addressing use the DS register. You can create an expression that points to a
memory location using any of the following operators:
Operator Name Symbol
Plus +
Minus –
Index []
Structure member .
Segment override :
These operators are discussed in more detail in the following section.
Plus, Minus, and Index

The plus and index operators perform in exactly the same way when applied to
direct memory operands. For example, both the following statements move the
second word value from an array into the AX register:
mov ax, array[2]
mov ax, array+2
The index operator can contain any direct memory operand. The following
statements are equivalent:
mov ax, var
mov ax, [var]
Some programmers prefer to enclose the operand in brackets to show that the
contents, not the address, are used.

The minus operator behaves as you would expect. Both the following
instructions retrieve the value located at the word preceding array:
mov ax, array[-2]
mov ax, array-2
Structure Field
The structure operator (.) references a particular element of a structure or
“field,” to use C terminology:
mov bx, structvar.field1
The address of the structure operand is the sum of the offsets of structvar
and field1. For more information about structures, see “Structures and
Unions” in Chapter 5.
Segment Override
The segment override operator (:) specifies a segment portion of the address
that is different from the default segment. When used with instructions, this
operator can apply to segment registers or segment names:
mov ax, es:farvar ; Use segment override
The assembler will not generate a segment override if the default segment is
explicitly provided. Thus, the following two statements assemble in exactly the
same way:
mov [bx], ax
mov ds:[bx], ax
A segment name override or the segment override operator identifies the

operand as an address expression.
mov WORD PTR FARSEG:0, ax ; Segment name override
mov WORD PTR es:100h, ax ; Legal and equivalent
mov WORD PTR es:[100h], ax ; expressions
; mov WORD PTR [100h], ax ; Illegal, not an address
As the example shows, a constant expression cannot be an address expression

unless it has a segment override.
Indirect Memory Operands

Like direct memory operands, indirect memory operands specify the contents of
a given address. However, the processor calculates the address at run time by
referring to the contents of registers. Since values in the registers can change at
run time, indirect memory operands provide dynamic access to memory.

Indirect memory operands make possible run-time operations such as pointer

indirection and dynamic indexing of array elements, including indexing of
multidimensional arrays.
Strict rules govern which registers you can use for indirect memory operands
under 16-bit versions of the 8086-based processors. The rules change
significantly for 32-bit processors starting with the 80386. However, the new
rules apply only to code that does not need to be compatible with earlier
processors.
This section covers features of indirect operands in either mode. The specific
16-bit rules and 32-bit rules are then explained separately.
Indirect Operands with 16- and 32-Bit Registers

Some rules and options for indirect memory operands always apply, regardless
of the size of the register. For example, you must always specify the register and
operand size for indirect memory operands. But you can use various syntaxes to
indicate an indirect memory operand. This section describes the rules that apply
to both 16-bit and 32-bit register modes.
Specifying Indirect Memory Operands

The index operator specifies the register or registers for indirect operands. The
processor uses the data pointed to by the register. For example, the following
instruction moves into AX the word value at the address in DS:BX.
mov ax, WORD PTR [bx]
When you specify more than one register, the processor adds the contents of the
two addresses together to determine the effective address (the address of the
data to operate on):
mov ax, [bx+si]
Specifying Displacements
You can specify an address displacement, which is a constant value added to the
effective address. A direct memory specifier is the most common displacement:
mov ax, table[si]
In this relocatable expression, the displacement table is the base address of an

array; SI holds an index to an array element. The SI value is calculated at run
time, often in a loop. The element loaded into AX depends on the value of SI at
the time the instruction executes.

Each displacement can be an address or numeric constant. If there is more than

one displacement, the assembler totals them at assembly time and encodes the
total displacement. For example, in the statement
table WORD 100 DUP (0)
.
.
.
mov ax, table[bx][di]+6
both table and 6 are displacements. The assembler adds the value of 6 to
table to get the total displacement. However, the statement
mov ax, mem1[si] + mem2
is not legal, because it attempts to use a single command to join the contents of
two different addresses.
Specifying Operand Size

You must give the size of an indirect memory operand in one of three ways:
u By the variable’s declared size
u With the PTR operator
u Implied by the size of the other operand
The following lines illustrate all three methods. Assume the size of the table
array is WORD, as declared earlier.
mov table[bx], 0 ; 2 bytes - from size of table
mov BYTE PTR table, 0 ; 1 byte - specified by BYTE
mov ax, [bx] ; 2 bytes - implied by AX
Syntax Options
The assembler allows a variety of syntaxes for indirect memory operands.
However, all registers must be inside brackets. You can enclose each register in
its own pair of brackets, or you can place the registers in the same pair of
brackets separated by a plus operator (+). All the following variations are legal
and assemble the same way:
mov ax, table[bx][di]
mov ax, table[di][bx]
mov ax, table[bx+di]
mov ax, [table+bx+di]
mov ax, [bx][di]+table
All of these statements move the value in table indexed by BX+DI into AX.

Scaling Indexes
The value of index registers pointing into arrays must often be adjusted for zero-
based arrays and scaled according to the size of the array items. For a word
array, the item number must be multiplied by two (shifted left by one place).
When using 16-bit registers, you must scale with separate instructions, as shown
here:
mov bx, 5 ; Get sixth element (adjust for 0)
shl bx, 1 ; Scale by two (word size)
inc wtable[bx] ; Increment sixth element in table
When using 32-bit registers on the 80386/486 processor, you can include scaling
in the operand, as described in “Indirect Memory Operands with 32-Bit
Registers,” following.
Accessing Structure Elements

The structure member operator can be used in indirect memory operands to
access structure elements. In this example, the structure member operator loads
the year field of the fourth element of the students array into AL:
STUDENT STRUCT
grade WORD ?
name BYTE 20 DUP (?)
year BYTE ?
STUDENT ENDS
students STUDENT < >

.
. ; Assume array is initialized
mov bx, OFFSET students; Point to array of students
mov ax, 4 ; Get fourth element
mov di, SIZE STUDENT ; Get size of STUDENT
mul di ; Multiply size times
mov di, ax ; elements to point DI
; to current element
mov al, (STUDENT PTR[bx+di]).year
For more information on MASM structures, see “Structures and Unions” in

Chapter 5.
Indirect Memory Operands with 16-Bit Registers

For 8086-based computers and MS-DOS, you must follow the strict indexing
rules established for the 8086 processor. Only four registers are allowed — BP,
BX, SI, and DI — those only in certain combinations.

BP and BX are base registers. SI and DI are index registers. You can use either
a base or an index register by itself. But if you combine two registers, one must
be a base and one an index. Here are legal and illegal forms:
mov ax, [bx+di] ; Legal
mov ax, [bx+si] ; Legal
mov ax, [bp+di] ; Legal
mov ax, [bp+si] ; Legal
; mov ax, [bx+bp] ; Illegal - two base registers
; mov ax, [di+si] ; Illegal - two index registers
Table 3.1 shows the register modes in which you can specify indirect memory
operands.
Table 3.1 Indirect Addressing with 16-Bit Registers
Mode Syntax Effective Address
Register indirect [BX] Contents of register
[BP]
[DI]
[SI]
Base or index displacement[BX] Contents of register plus
displacement[BP] displacement
displacement[DI]
displacement[SI]
Base plus index [BX][DI] Contents of base register plus
[BP][DI] contents of index register
[BX][SI]
[BP][SI]
Base plus index with displacement[BX][DI] Sum of base register, index
displacement displacement[BP][DI] register, and displacement
displacement[BX][SI]
displacement[BP][SI]
Different combinations of registers and displacements have different timings, as

shown in Reference.
Indirect Memory Operands with 32-Bit Registers

You can write instructions for the 80386/486 processor using either 16-bit or
32-bit segments. Indirect memory operands are different in each case.
In 16-bit real mode, the 80386/486 operates the same way as earlier 8086-based
processors, with one difference: you can use 32-bit registers. If the 80386/486
processor is enabled (with the .386 or .486 directive), 32-bit general-purpose
registers are available with either 16-bit or 32-bit segments. Thirty-two–bit

registers eliminate many of the limitations of 16-bit indirect memory operands.

You can use 80386/486 features to make your MS-DOS programs run faster
and more efficiently if you are willing to sacrifice compatibility with earlier
processors.
In 32-bit mode, an offset address can be up to 4 gigabytes. (Segments are still
represented in 16 bits.) This effectively eliminates size restrictions on each
segment, since few programs need 4 gigabytes of memory. Windows NT uses
32-bit mode and flat model, which spans all segments. XENIX 386 uses 32-bit
mode with multiple segments.
80386/486 Enhancements
On the 80386/486, the processor allows you to use any general-purpose 32-bit
register as a base or index register, except ESP, which can be a base but not an
index. However, you cannot combine 16-bit and 32-bit registers. Several
examples are shown here:
add edx, [eax] ; Add double
mov dl, [esp+10] ; Copy byte from stack
dec WORD PTR [edx][eax] ; Decrement word
cmp ax, array[ebx][ecx] ; Compare word from array
jmp FWORD PTR table[ecx] ; Jump into pointer table
Scaling Factors
With 80386/486 registers, the index register can have a scaling factor of 1, 2, 4,
or 8. Any register except ESP can be the index register and can have a scaling
factor. To specify the scaling factor, use the multiplication operator (*) adjacent
to the register.
You can use scaling to index into arrays with different sizes of elements. For
example, the scaling factor is 1 for byte arrays (no scaling needed), 2 for word
arrays, 4 for doubleword arrays, and 8 for quadword arrays. There is no
performance penalty for using a scaling factor. Scaling is illustrated in the
following examples:
mov eax, darray[edx*4] ; Load double of double array
mov eax, [esi*8][edi] ; Load double of quad array
mov ax, wtbl[ecx+2][edx*2] ; Load word of word array
Scaling is also necessary on earlier processors, but it must be done with separate
instructions before the indirect memory operand is used, as described in
“Indirect Memory Operands with 16-Bit Registers,” previous.
The default segment register is SS if the base register is EBP or ESP. However,
if EBP is scaled, the processor treats it as an index register with a value relative
to DS, not SS.

All other base registers are relative to DS. If two registers are used, only one can
have a scaling factor. The register with the scaling factor is defined as the index
register. The other register is defined as the base. If scaling is not used, the first
register is the base. If only one register is used, it is considered the base for
deciding the default segment unless it is scaled. The following examples illustrate
how to determine the base register:
mov eax, [edx][ebp*4] ; EDX base (not scaled - seg DS)
mov eax, [edx*1][ebp] ; EBP base (not scaled - seg SS)
mov eax, [edx][ebp] ; EDX base (first - seg DS)
mov eax, [ebp][edx] ; EBP base (first - seg SS)
mov eax, [ebp] ; EBP base (only - seg SS)
mov eax, [ebp*2] ; EBP*2 index (seg DS)
Mixing 16-Bit and 32-Bit Registers

Assembly statements can mix 16-bit and 32-bit registers. For example, the
following statement is legal for 16-bit and 32-bit segments:
mov eax, [bx]
This statement moves the 32-bit value pointed to by BX into the EAX register.
Although BX is a 16-bit pointer, it can still point into a 32-bit segment.
However, the following statement is never legal, since you cannot use the CX
register as a 16-bit pointer:
; mov eax, [cx] ; illegal
Operands that mix 16-bit and 32-bit registers are also illegal:
; mov eax, [ebx+si] ; illegal
The following statement is legal in either 16-bit or 32-bit mode:

mov bx, [eax]
This statement moves the 16-bit value pointed to by EAX into the BX register.
This works in 32-bit mode. However, in 16-bit mode, moving a 32-bit pointer
into a 16-bit segment is illegal. If EAX contains a 16-bit value (the top half of
the 32-bit register is 0), the statement works. However, if the top half of the
EAX register is not 0, the operand points into a part of the segment that doesn’t
exist, generating an error. If you use 32-bit registers as indexes in 16-bit mode,
you must make sure that the index registers contain valid 16-bit addresses.
The Program Stack

The preceding discussion on memory operands lays the groundwork for
understanding the important data area known as the “stack.”

A stack is an area of memory for storing data temporarily. Unlike other

segments that store data starting from low memory, the stack stores data starting
from high memory. Data is always pushed onto, or “popped” from the top of
the stack.
The stack gets its name from its similarity to the spring-loaded plate holders in
cafeterias. You add and remove plates from only the top of the stack. To
retrieve the third plate, you must remove — that is, “pop” — the first two plates.
Stacks are often referred to as LIFO buffers, from their last-in-first-out
operation.
A stack is an essential part of any nontrivial program. A program continually
uses its stack to temporarily store return addresses, procedure arguments,
memory data, flags, or registers.
The SP register serves as an indirect memory operand to the top of the stack. At
first, the stack is an uninitialized segment of a finite size. As your program adds
data to the stack, the stack grows downward from high memory to low
memory. When you remove items from the stack, it shrinks upward from low to
high memory.
Saving Operands on the Stack

The PUSH instruction stores a 2-byte operand on the stack. The POP
instruction retrieves the most recent pushed value. When a value is pushed onto
the stack, the assembler decreases the SP (Stack Pointer) register by 2. On
8086-based processors, the SP register always points to the top of the stack.
The PUSH and POP instructions use the SP register to keep track of the
current position.
When a value is popped off the stack, the assembler increases the SP register by
2. Since the stack always contains word values, the SP register changes in
multiples of two. When a PUSH or POP instruction executes in a 32-bit code
segment (one with USE32 use type), the assembler transfers a 4-byte value, and
ESP changes in multiples of four.
Note The 8086 and 8088 processors differ from later Intel processors in how
they push and pop the SP register. If you give the statement push sp with the
8086 or 8088, the word pushed is the word in SP after the push operation.

Figure 3.1 illustrates how pushes and pops change the SP register.
Figure 3.1 Stack Status Before and After Pushes and Pops
On the 8086, PUSH and POP take only registers or memory expressions as
their operands. The other processors allow an immediate value to be an operand
for PUSH. For example, the following statement is legal on the 80186–80486
processors:
push 7 ; 3 clocks on 80286
That statement is faster than these equivalent statements, which are required on
the 8088 or 8086:
mov ax, 7 ; 2 clocks plus
push ax ; 3 clocks on 80286
Words are popped off the stack in reverse order: the last item pushed is the first
popped. To return the stack to its original status, you do the same number of

pops as pushes. You can subtract the correct number of words from the SP
register if you want to restore the stack without using the values on it.
To reference operands on the stack, remember that the values pointed to by the
BP (Base Pointer) and SP registers are relative to the SS (Stack Segment)
register. The BP register is often used to point to the base of a frame of
reference (a stack frame) within the stack. This example shows how you can
access values on the stack using indirect memory operands with BP as the base
register.
push bp ; Save current value of BP
mov bp, sp ; Set stack frame
push ax ; Push first; SP = BP - 2
push bx ; Push second; SP = BP - 4
push cx ; Push third; SP = BP - 6
.
.
.
mov ax, [bp-6] ; Put third word in AX
mov bx, [bp-4] ; Put second word in BX
mov cx, [bp-2] ; Put first word in CX
.
.
.
add sp, 6 ; Restore stack pointer
; (two bytes per push)
pop bp ; Restore BP
If you often use these stack values in your program, you may want to give them
labels. For example, you can use TEXTEQU to create a label such as count
TEXTEQU <[bp-6]>. Now you can replace the mov ax, [bp - 6]
statement in the previous example with mov ax, count. For more information
about the TEXTEQU directive, see “Text Macros” in Chapter 9.
Saving Flags on the Stack

Your program can push and pop flags onto the stack with the PUSHF and
POPF instructions. These instructions save and then restore the status of the
flags. You can also use them within a procedure to save and restore the flag
status of the caller. The 32-bit versions of these instructions are PUSHFD and
POPFD.
This example saves the flags register before calling the systask procedure:
pushf
call systask
popf

If you do not need to store the entire flags register, you can use the LAHF
instruction to manually load and store the status of the lower byte of the flag
register in the AH register. SAHF restores the value.
Saving Registers on the Stack (80186–80486 Only)

Starting with the 80186 processor, the PUSHA and POPA instructions push or
pop all the general-purpose registers with only one instruction. These
instructions save the status of all registers before a procedure call and restore
them after the return. Using PUSHA and POPA is significantly faster and takes
fewer bytes of code than pushing and popping each register individually.
The processor pushes the registers in the following order: AX, CX, DX, BX,
SP, BP, SI, and DI. The SP word pushed is the value before the first register is
pushed.
The processor pops the registers in the opposite order. The 32-bit versions of
these instructions are PUSHAD and POPAD.
Accessing Data with Pointers and Addresses

A pointer is simply a variable that contains an address of some other variable.
The address in the pointer “points” to the other object. Pointers are useful when
transferring a large data object (such as an array) to a procedure. The caller
places only the pointer on the stack, which the called procedure uses to locate
the array. This eliminates the impractical step of having to pass the entire array
back and forth through the stack.
There is a difference between a far address and a far pointer. A “far address” is
the address of a variable located in a far data segment. A “far pointer” is a
variable that contains the segment address and offset of some other data. Like
any other variable, a pointer can be located in either the default (near) data
segment or in a far segment.
Previous versions of MASM allow pointer variables but provide little support for
them. In previous versions, any address loaded into a variable can be considered
a pointer, as in the following statements:
Var BYTE 0 ; Variable
npVar WORD Var ; Near pointer to variable
fpVar DWORD Var ; Far pointer to variable
If a variable is initialized with the name of another variable, the initialized

variable is a pointer, as shown in this example. However, in previous versions of
MASM, the CodeView debugger recognizes npVar and fpVar as word and
doubleword variables. CodeView does not treat them as pointers, nor does it
recognize the type of data they point to (bytes, in the example).

The TYPEDEF directive and enhanced capabilities of ASSUME (introduced in

MASM 6.0) make it easier to manage pointers in registers and variables. The
rest of this chapter describes these directives and how they apply to basic
pointer operations.
Defining Pointer Types with TYPEDEF

The TYPEDEF directive can define types for pointer variables. A type so
defined is considered the same as the intrinsic types provided by the assembler
and can be used in the same contexts. When used to define pointers, the syntax
for TYPEDEF is:
typename TYPEDEF [[distance]] PTR qualifiedtype
The typename is the name assigned to the new type. The distance can be
NEAR, FAR, or any distance modifier. The qualifiedtype can be any
previously intrinsic or defined MASM type, or a type previously defined with
TYPEDEF. (For a full definition of qualifiedtype, see “Data Types” in Chapter
1.)
Here are some examples of user-defined types:
PBYTE TYPEDEF PTR BYTE ; Pointer to bytes
NPBYTE TYPEDEF NEAR PTR BYTE ; Near pointer to bytes
FPBYTE TYPEDEF FAR PTR BYTE ; Far pointer to bytes
PWORD TYPEDEF PTR WORD ; Pointer to words
NPWORD TYPEDEF NEAR PTR WORD ; Near pointer to words
FPWORD TYPEDEF FAR PTR WORD ; Far pointer to words
PPBYTE TYPEDEF PTR PBYTE ; Pointer to pointer to bytes

; (in C, an array of strings)
PVOID TYPEDEF PTR ; Pointer to any type of data
PERSON STRUCT ; Structure type

name BYTE 20 DUP (?)
num WORD ?
PERSON ENDS
PPERSON TYPEDEF PTR PERSON ; Pointer to structure type
The distance of a pointer can be set specifically or determined automatically by

the memory model (set by .MODEL) and the segment size (16 or 32 bits). If
you don’t use .MODEL, near pointers are the default.
In 16-bit mode, a near pointer is 2 bytes that contain the offset of the object
pointed to. A far pointer requires 4 bytes, and contains both the segment and
offset. In 32-bit mode, a near pointer is 4 bytes and a far pointer is 6 bytes,
since segments are

still word values in 32-bit mode. If you specify the distance with NEAR or
FAR, the processor uses the default distance of the current segment size. You
can use NEAR16, NEAR32, FAR16, and FAR32 to override the defaults set
by the current segment size. In flat model, NEAR is the default.
You can declare pointer variables with a pointer type created with TYPEDEF.
Here are some examples using these pointer types.
; Type declarations
Array WORD 25 DUP (0)
Msg BYTE "This is a string", 0
pMsg PBYTE Msg ; Pointer to string
pArray PWORD Array ; Pointer to word array
npMsg NPBYTE Msg ; Near pointer to string
npArray NPWORD Array ; Near pointer to word array
fpArray FPWORD Array ; Far pointer to word array
fpMsg FPBYTE Msg ; Far pointer to string
S1 BYTE "first", 0 ; Some strings

S2 BYTE "second", 0
S3 BYTE "third", 0
pS123 PBYTE S1, S2, S3, 0 ; Array of pointers to strings
ppS123 PPBYTE pS123 ; A pointer to pointers to strings
Andy PERSON <> ; Structure variable

pAndy PPERSON Andy ; Pointer to structure variable
; Procedure prototype
EXTERN ptrArray:PBYTE ; External variable

Sort PROTO pArray:PBYTE ; Parameter for prototype
; Parameter for procedure

Sort PROC pArray:PBYTE
LOCAL pTmp:PBYTE ; Local variable
.
.
.
ret
Sort ENDP
Once defined, pointer types can be used in any context where intrinsic types are
allowed.

Defining Register Types with ASSUME

You can use the ASSUME directive with general-purpose registers to specify
that a register is a pointer to a certain size of object. For example:
ASSUME bx:PTR WORD ; Assume BX is now a word pointer
inc [bx] ; Increment word pointed to by BX
add bx, 2 ; Point to next word
mov [bx], 0 ; Word pointed to by BX = 0
.
. ; Other pointer operations with BX
.
ASSUME bx:NOTHING ; Cancel assumption
In this example, BX is specified as a pointer to a word. After a sequence of

using BX as a pointer, the assumption is canceled by assuming NOTHING.
Without the assumption to PTR WORD, many instructions need a size
specifier. The INC and MOV statements from the previous examples would
have to be written like this to specify the sizes of the memory operands:
inc WORD PTR [bx]
mov WORD PTR [bx], 0
When you have used ASSUME, attempts to use the register for other purposes
generate assembly errors. In this example, while the PTR WORD assumption is
in effect, any use of BX inconsistent with its ASSUME declaration generates an
error. For example,
; mov al, [bx] ; Can't move word to byte register
You can also use the PTR operator to override defaults:

mov al, BYTE PTR [bx] ; Legal
Similarly, you can use ASSUME to prevent the use of a register as a pointer, or
even to disable a register:
ASSUME bx:WORD, dx:ERROR
; mov al, [bx] ; Error - BX is an integer, not a pointer
; mov ax, dx ; Error - DX disabled
For information on using ASSUME with segment registers, refer to “Setting the
ASSUME Directive for Segment Registers” in Chapter 2.

Basic Pointer and Address Operations

A program can perform the following basic operations with pointers and
addresses:
u Initialize a pointer variable by storing an address in it.
u Load an address into registers, directly or from a pointer.
The sections in the rest of this chapter describe variations of these tasks with
pointers and addresses. The examples are used with the assumption that you
have previously defined the following pointer types with the TYPEDEF
directive:
PBYTE TYPEDEF PTR BYTE ; Pointer to bytes
NPBYTE TYPEDEF NEAR PTR BYTE ; Near pointer to bytes
FPBYTE TYPEDEF FAR PTR BYTE ; Far pointer to bytes
Initializing Pointer Variables

If the value of a pointer is known at assembly time, the assembler can initialize it
automatically so that no processing time is wasted on the task at run time. The
following example shows how to do this, placing the address of msg in the
pointer pmsg.
Msg BYTE "String", 0
pMsg PBYTE Msg
If a pointer variable can be conditionally defined to one of several constant

addresses, initialization must be delayed until run time. The technique is
different for near pointers than for far pointers, as shown here:
Msg1 BYTE "String1"
Msg2 BYTE "String2"
npMsg NPBYTE ?
fpMsg FPBYTE ?
.
.
.
mov npMsg, OFFSET Msg1 ; Load near pointer
mov WORD PTR fpMsg[0], OFFSET Msg2 ; Load far offset

mov WORD PTR fpMsg[2], SEG Msg2 ; Load far segment
If you know that the segment for a far pointer is in a register, you can load it
directly:
mov WORD PTR fpMsg[2], ds ; Load segment of
; far pointer

Dynamic Addresses
Often a pointer must point to a dynamic address, meaning the address depends
on a run-time condition. Typical situations include memory allocated by MS-
DOS (see “Interrupt 21h Function 48h” in Help) and addresses found by the
SCAS or CMPS instructions (see “Processing Strings” in Chapter 5). The
following illustrates the technique for saving dynamic addresses:
; Dynamically allocated buffer
fpBuf FPBYTE 0 ; Initialize so offset will be zero
.
.
.
mov ah, 48h ; Allocate memory
mov bx, 10h ; Request 16 paragraphs
int 21h ; Call DOS
jc error ; Return segment in AX
mov WORD PTR fpBuf[2], ax ; Load segment
. ; (offset is already 0)
.
.
error: ; Handle error
Copying Pointers
Sometimes one pointer variable must be initialized by copying from another.
Here are two ways to copy a far pointer:
fpBuf1 FPBYTE ?
fpBuf2 FPBYTE ?
.
.
.
; Copy through registers is faster, but requires a spare register
mov ax, WORD PTR fpBuf1[0]
mov WORD PTR fpBuf2[0], ax
mov ax, WORD PTR fpBuf1[2]
mov WORD PTR fpBuf2[2], ax
; Copy through stack is slower, but does not use a register

push WORD PTR fpBuf1[0]
push WORD PTR fpBuf1[2]
pop WORD PTR fpBuf2[2]
pop WORD PTR fpBuf2[0]

Pointers as Arguments
Most high-level-language procedures and library functions accept arguments
passed on the stack. “Passing Arguments on the Stack” in Chapter 7 covers this
subject in detail. A pointer is passed in the same way as any other variable, as
this fragment shows:
; Push a far pointer (segment always pushed first)
push WORD PTR fpMsg[2] ; Push segment
push WORD PTR fpMsg[0] ; Push offset
Pushing an address has the same result as pushing a pointer to the address:
; Push a far address as a far pointer
mov ax, SEG fVar ; Load and push segment
push ax
mov ax, OFFSET fVar ; Load and push offset
push ax
On the 80186 and later processors, you can push a constant in one step:
push SEG fVar ; Push segment
push OFFSET fVar ; Push offset
Loading Addresses into Registers

Loading a near address into a register (or a far address into a pair of registers) is
a common task in assembly-language programming. To reference data pointed
to by a pointer, your program must first place the pointer into a register or pair
of registers.
Load far addresses as segment:offset pairs. The following pairs have specific
uses:
Segment:Offset Pair Standard Use
DS:SI Source for string operations
ES:DI Destination for string operations
DS:DX Input for certain DOS functions
ES:BX Output from certain DOS functions
Addresses from Data Segments

For near addresses, you need only load the offset; the segment is assumed as SS
for stack-based data and as DS for other data. You must load both segment and
offset for far pointers.

Here is an example of loading an address into DS:BX from a near data segment:
.DATA
Msg BYTE "String"
.
.
.
mov bx, OFFSET Msg ; Load address to BX
; (DS already loaded)
Far data can be loaded like this:

.FARDATA
Msg BYTE "String"
.
.
.
mov ax, SEG Msg ; Load address to ES:BX
mov es, ax
mov bx, OFFSET Msg
You can also read a far address from a pointer in one step, using the LES and
LDS instructions described next.
Far Pointers
The LES and LDS instructions load a far pointer into a segment pair. The
instructions copy the pointer’s low word into either ES or DS, and the high
word into a given register. The following example shows how to load a far
pointer into ES:DI:
OutBuf BYTE 20 DUP (0)
fpOut FPBYTE OutBuf

.
.
.
les di, fpOut ; Load far pointer into ES:DI

Stack Variables
The technique for loading the address of a stack variable is significantly different
from the technique for loading near addresses. You may need to put the correct
segment value into ES for string operations. The following example illustrates
how to load the address of a local (stack) variable to ES:DI:
Task PROC
LOCAL Arg[4]:BYTE
push ss ; Since it's stack-based, segment is SS

pop es ; Copy SS to ES
lea di, Arg ; Load offset to DI
The local variable in this case actually evaluates to SS:[BP-4]. This is an offset
from the stack frame (described in “Passing Arguments on the Stack,” Chapter
7). Since you cannot use the OFFSET operator to get the offset of an indirect
memory operand, you must use the LEA (Load Effective Address) instruction.
Direct Memory Operands

To get the address of a direct memory operand, use either the LEA instruction
or the MOV instruction with OFFSET. Though both methods have the same
effect, the MOV instruction produces smaller and faster code, as shown in this
example:
lea si, Msg ; Four byte instruction
mov si, OFFSET Msg ; Three byte equivalent
Copying Between Segment Pairs

Copying from one register pair to another is complicated by the fact that you
cannot copy one segment register directly to another. Two copying methods are
shown here. Timings are for the 8088 processor.
; Copy DS:SI to ES:DI, generating smaller code
push ds ; 1 byte, 14 clocks
pop es ; 1 byte, 12 clocks
mov di, si ; 2 bytes, 2 clocks
; Copy DS:SI to ES:DI, generating faster code

mov di, ds ; 2 bytes, 2 clocks
mov es, di ; 2 bytes, 2 clocks
mov di, si ; 2 bytes, 2 clocks

Model-Independent Techniques
Often you may want to write code that is memory-model independent. If you
are writing libraries that must be available for different memory models, you can
use conditional assembly to handle different sizes of pointers. You can use the
predefined symbols @DataSize and @Model to test the current assumptions.
You can use conditional assembly to write code that works with pointer
variables that have no specified distance. The predefined symbol @DataSize
tests the pointer size for the current memory model:
Msg1 BYTE "String1"
pMsg PBYTE ?
.
.
.
IF @DataSize ; @DataSize > 0 for far
mov WORD PTR pMsg[0], OFFSET Msg1 ; Load far offset
mov WORD PTR pMsg[2], SEG Msg1 ; Load far segment
ELSE ; @DataSize = 0 for near
mov pMsg, OFFSET Msg1 ; Load near pointer
ENDIF
In the following example, a procedure receives as an argument a pointer to a

word variable. The code inside the procedure uses @DataSize to determine
whether the current memory model supports far or near data. It loads and
processes the data accordingly:
; Procedure that receives an argument by reference
mul8 PROC arg:PTR WORD
IF @DataSize
les bx, arg ; Load far pointer to ES:BX
mov ax, es:[bx] ; Load the data pointed to
ELSE
mov bx, arg ; Load near pointer to BX (assume DS)
mov ax, [bx] ; Load the data pointed to
ENDIF
shl ax, 1 ; Multiply by 8
shl ax, 1
shl ax, 1
ret
mul8 ENDP

If you have many routines, writing the conditionals for each case can be tedious.
The following conditional statements automatically generate the proper
instructions and segment overrides.
; Equates for conditional handling of pointers
IF @DataSize
lesIF TEXTEQU <les>
ldsIF TEXTEQU <lds>
esIF TEXTEQU <es:>
ELSE
lesIF TEXTEQU <mov>
ldsIF TEXTEQU <mov>
esIF TEXTEQU <>
ENDIF
Once you define these conditionals, you can use them to simplify code that must
handle several types of pointers. This next example rewrites the above mul8
procedure to use conditional code.
mul8 PROC arg:PTR WORD
lesIF bx, arg ; Load pointer to BX or ES:BX

mov ax, esIF [bx] ; Load the data from [BX] or ES:[BX]
shl ax, 1 ; Multiply by 8
shl ax, 1
shl ax, 1
ret
mul8 ENDP
The conditional statements from these examples can be defined once in an

include file and used whenever you need to handle pointers.

85
C H A P T E R 4
Defining and Using Simple

Data Types
This chapter covers the concepts essential for working with simple data types in
assembly-language programs. The first section shows how to declare integer
variables. The second section describes basic operations including moving,
loading, and sign-extending numbers, as well as calculating. The last section
describes how to do various operations with numbers at the bit level, such as
using bitwise logical instructions and shifting and rotating bits.
The complex data types introduced in the next chapter — arrays, strings,
structures, unions, and records — use many of the operations illustrated in this
chapter. Floating-point operations require a different set of instructions and
techniques. These are covered in Chapter 6, “Using Floating-Point and Binary
Coded Decimal
Numbers.”
Declaring Integer Variables

An integer is a whole number, such as 4 or 4,444. Integers have no fractional
part, as do the real numbers discussed in Chapter 6. You can initialize integer
variables in several ways with the data allocation directives. This section
explains how to use the SIZEOF and TYPE operators to provide information to
the assembler about the types in your program. For information on symbolic
integer constants, see “Integer Constants and Constant Expressions” in Chapter
1.
Allocating Memory for Integer Variables

When you declare an integer variable by assigning a label to a data allocation
directive, the assembler allocates memory space for the integer. The variable’s
name becomes a label for the memory space. The syntax is:
[[name]] directive initializer

Template: MSGRIDA1.DOT Author: Ruth L Silverio Last Saved By: Ruth L Silverio
The following directives indicate the integer’s size and value range:
Directive Description of Initializers
BYTE, DB (byte) Allocates unsigned numbers from 0 to 255.
SBYTE (signed byte) Allocates signed numbers from –128 to +127.
WORD, DW (word = 2 bytes) Allocates unsigned numbers from
0 to 65,535 (64K).
SWORD (signed word) Allocates signed numbers from
–32,768 to +32,767.
DWORD, DD (doubleword = 4 Allocates unsigned numbers from
bytes), 0 to 4,294,967,295 (4 megabytes).
SDWORD (signed doubleword) Allocates signed numbers from
–2,147,483,648 to +2,147,483,647.
FWORD, DF (farword = 6 bytes) Allocates 6-byte (48-bit) integers. These values are
normally used only as pointer variables on the
80386/486 processors.
QWORD, DQ (quadword = 8 bytes) Allocates 8-byte integers used with 8087-family
coprocessor instructions.
TBYTE, DT (10 bytes), Allocates 10-byte (80-bit) integers if the initializer has
a radix specifying the base of the number.
See Chapter 6 for information on the REAL4, REAL8, and REAL10 directives
that allocate real numbers.
The SIZEOF and TYPE operators, when applied to a type, return the size of an
integer of that type. The size attribute associated with each data type is:
Data Type Bytes
BYTE, SBYTE 1
WORD, SWORD 2
DWORD, SDWORD 4
FWORD 6
QWORD 8
TBYTE 10
The data types SBYTE, SWORD, and SDWORD tell the assembler to treat the
initializers as signed data. It is important to use these signed types with high-level
constructs such as .IF, .WHILE, and .REPEAT, and with PROTO and
INVOKE directives. For descriptions of these directives, see the sections
“Loop-Generating Directives,” “Declaring Procedure Prototypes,” and “Calling
Procedures with INVOKE” in Chapter 7.

Chapter 4 Defining and Using Simple Data Types 87
The assembler stores integers with the least significant bytes lowest in memory.
Note that assembler listings and most debuggers show the bytes of a word in the
opposite order — high byte first.
Figure 4.1 illustrates the integer formats.
Figure 4.1 Integer Formats
Although the TYPEDEF directive’s primary purpose is to define pointer

variables (see “Defining Pointer Types with TYPEDEF” in Chapter 3), you can
also use TYPEDEF to create an alias for any integer type. For example, these
declarations
char TYPEDEF SBYTE
long TYPEDEF DWORD
float TYPEDEF REAL4
double TYPEDEF REAL8
allow you to use char, long, float, or double in your programs if you
prefer the C data labels.
Data Initialization
You can initialize variables when you declare them with constants or expressions
that evaluate to constants. The assembler generates an error if you specify an
initial value too large for the variable type.
A ? in place of an initializer indicates you do not require the assembler to
initialize the variable. The assembler allocates the space but does not write in it.
Use ? for buffer areas or variables your program will initialize at run time.

You can declare and initialize variables in one step with the data directives, as
these examples show.
integer BYTE 16 ; Initialize byte to 16
negint SBYTE -16 ; Initialize signed byte to -16
expression WORD 4*3 ; Initialize word to 12
signedexp SWORD 4*3 ; Initialize signed word to 12
empty QWORD ? ; Allocate uninitialized long int
BYTE 1,2,3,4,5,6 ; Initialize six unnamed bytes
long DWORD 4294967295 ; Initialize doubleword to
; 4,294,967,295
longnum SDWORD -2147433648 ; Initialize signed doubleword
; to -2,147,433,648
tb TBYTE 2345t ; Initialize 10-byte binary number
For information on arrays and on using the DUP operator to allocate initializer
lists, see “Arrays and Strings” in Chapter 5.
Working with Simple Variables

Once you have declared integer variables in your program, you can use them to
copy, move, and sign-extend integer variables in your MASM code. This section
shows how to do these operations as well as how to add, subtract, multiply, and
divide numbers and do bit-level manipulations with logical, shift, and rotate
instructions.
Since MASM instructions require operands to be the same size, you may need
to operate on data in a size other than that originally declared. You can do this
with the PTR operator. For example, you can use the PTR operator to access
the high-order word of a DWORD-size variable. The syntax for the PTR
operator is
type PTR expression
where the PTR operator forces expression to be treated as having the type
specified. An example of this use is
.DATA
num DWORD 0
.CODE
mov ax, WORD PTR num[0] ; Loads a word-size value from

mov dx, WORD PTR num[2] ; a doubleword variable

Copying Data
The primary instructions for moving data from operand to operand and loading
them into registers are MOV (Move), XCHG (Exchange), CWD (Convert
Word to Double), and CBW (Convert Byte to Word).
Moving Data
The most common method of moving data, the MOV instruction, is essentially
a copy instruction, since it always copies the source operand to the destination
operand without affecting the source. After a MOV instruction, the source and
destination operands contain the same value.
The following example illustrates the MOV instruction. As explained in
“General-Purpose Registers,” Chapter 1, you cannot move a value from one
location in memory to another in a single operation.
; Immediate value moves
mov ax, 7 ; Immediate to register
mov mem, 7 ; Immediate to memory direct
mov mem[bx], 7 ; Immediate to memory indirect
; Register moves
mov mem, ax ; Register to memory direct
mov mem[bx], ax ; Register to memory indirect
mov ax, bx ; Register to register
mov ds, ax ; General register to segment register
; Direct memory moves

mov ax, mem ; Memory direct to register
mov ds, mem ; Memory to segment register
; Indirect memory moves

mov ax, mem[bx] ; Memory indirect to register
mov ds, mem[bx] ; Memory indirect to segment register
; Segment register moves

mov mem, ds ; Segment register to memory
mov mem[bx], ds ; Segment register to memory indirect
mov ax, ds ; Segment register to general register

The following example shows several common types of moves that require two
instructions.
; Move immediate to segment register
mov ax, DGROUP ; Load AX with immediate value
mov ds, ax ; Copy AX to segment register
; Move memory to memory

mov ax, mem1 ; Load AX with memory value
mov mem2, ax ; Copy AX to other memory
; Move segment register to segment register

mov ax, ds ; Load AX with segment register
mov es, ax ; Copy AX to segment register
The MOVSX and MOVZX instructions for the 80386/486 processors extend
and copy values in one step. See “Extending Signed and Unsigned Integers,”
following.
Exchanging Integers
The XCHG (Exchange) instruction exchanges the data in the source and
destination operands. You can exchange data between registers or between
registers and memory, but not from memory to memory:
xchg ax, bx ; Put AX in BX and BX in AX
xchg memory, ax ; Put "memory" in AX and AX in "memory"
; xchg mem1, mem2 ; Illegal- can't exchange memory locations
Extending Signed and Unsigned Integers

Since moving data between registers of different sizes is illegal, you must “sign-
extend” integers to convert signed data to a larger size. Sign-extending means
copying the sign bit of the unextended operand to all bits of the operand’s next
larger size. This widens the operand while maintaining its sign and value.
8086-based processors provide four instructions specifically for sign-extending.
The four instructions act only on the accumulator register (AL, AX, or EAX), as
shown in the following list.
Instruction Sign-extend
CBW (convert byte to word) AL to AX
CWD (convert word to doubleword) AX to DX:AX
CWDE (convert word to doubleword extended)* AX to EAX
CDQ (convert doubleword to quadword)* EAX to EDX:EAX
*Requires an extended register and applies only to 80386/486 processors.

On the 80386/486 processors, the CWDE instruction converts a signed 16-bit

value in AX to a signed 32-bit value in EAX. The CDQ instruction converts a
signed 32-bit value in EAX to a signed 64-bit value in the EDX:EAX register
pair.
This example converts signed integers using CBW, CWD, CWDE, and CDQ.
.DATA
mem8 SBYTE -5
mem16 SWORD +5
mem32 SDWORD -5
.CODE
.
.
.
mov al, mem8 ; Load 8-bit -5 (FBh)
cbw ; Convert to 16-bit -5 (FFFBh) in AX
mov ax, mem16 ; Load 16-bit +5
cwd ; Convert to 32-bit +5 (0000:0005h) in DX:AX
mov ax, mem16 ; Load 16-bit +5
cwde ; Convert to 32-bit +5 (00000005h) in EAX
mov eax, mem32 ; Load 32-bit -5 (FFFFFFFBh)
cdq ; Convert to 64-bit -5
; (FFFFFFFF:FFFFFFFBh) in EDX:EAX
These four instructions efficiently convert unsigned values as well, provided the
sign bit is zero. This example, for instance, correctly widens mem16 whether
you treat the variable as signed or unsigned.
The processor does not differentiate between signed and unsigned values. For
instance, the value of mem8 in the previous example is literally 251 (0FBh) to
the processor. It ignores the human convention of treating the highest bit as an
indicator of sign. The processor can ignore the distinction between signed and
unsigned numbers because binary arithmetic works the same in either case.
If you add 7 to mem8, for example, the result is 258 (102h), a value too large to
fit into a single byte. The byte-sized mem8 can accommodate only the least-
significant digits of the result (02h), and so receives the value of 2. The result is
the same whether we treat mem8 as a signed value (-5) or unsigned value (251).
This overview illustrates how the programmer, not the processor, must keep
track of which values are signed or unsigned, and treat them accordingly. If
AL=127 (01111111y), the instruction CBW sets AX=127 because the sign bit is
zero. If AL=128 (10000000y), however, the sign bit is 1. CBW thus sets
AX=65,280

(FF00h), which may not be what you had in mind if you assumed AL originally
held an unsigned value.To widen unsigned values, explicitly set the higher
register to zero, as shown in the following example:
.DATA
mem8 BYTE 251
mem16 WORD 251
.CODE
.
.
.
mov al, mem8 ; Load 251 (FBh) from 8-bit memory
sub ah, ah ; Zero upper half (AH)
mov ax, mem16 ; Load 251 (FBh) from 16-bit memory

sub dx, dx ; Zero upper half (DX)
sub eax, eax ; Zero entire extended register (EAX)

mov ax, mem16 ; Load 251 (FBh) from 16-bit memory
The 80386/486 processors provide instructions that move and extend a value to
a larger data size in a single step. MOVSX moves a signed value into a register
and sign-extends it. MOVZX moves an unsigned value into a register and zero-
extends it.
; 80386/486 instructions
movzx dx, bl ; Load unsigned 8-bit value into
; 16-bit register and zero-extend
These special 80386/486 instructions usually execute much faster than the
equivalent 8086/286 instructions.
Adding and Subtracting Integers

You can use the ADD, ADC, INC, SUB, SBB, and DEC instructions for
adding, incrementing, subtracting, and decrementing values in single registers.
You can also combine them to handle larger values that require two registers for
storage.
Adding and Subtracting Integers Directly

The ADD, INC (Increment), SUB, and DEC (Decrement) instructions operate
on 8- and 16-bit values on the 8086–80286 processors, and on 8-, 16-, and 32-
bit values on the 80386/486 processors. They can be combined with the ADC
and SBB instructions to work on 32-bit values on the 8086 and 64-bit values on
the 80386/486 processors. (See “Adding and Subtracting in Multiple Registers,”
following.)

These instructions have two requirements:

1. If there are two operands, only one operand can be a memory operand.
2. If there are two operands, both must be the same size.
To meet the second requirement, you can use the PTR operator to force an
operand to the size required. (See “Working with Simple Variables,” previous.)
For example, if Buffer is an array of bytes and BX points to an element of the
array, you can add a word from Buffer with
add ax, WORD PTR Buffer[bx] ; Add word from byte array
The next example shows 8-bit signed and unsigned addition and subtraction.
.DATA
mem8 BYTE 39
.CODE
; Addition
; signed unsigned
mov al, 26 ; Start with register 26 26
inc al ; Increment 1 1
add al, 76 ; Add immediate 76 + 76
; ---- ----
; 103 103
add al, mem8 ; Add memory 39 + 39
; ---- ----
mov ah, al ; Copy to AH -114 142
+overflow
add al, ah ; Add register 142
; ----
; 28+carry
; Subtraction
; signed unsigned
mov al, 95 ; Load register 95 95
dec al ; Decrement -1 -1
sub al, 23 ; Subtract immediate -23 -23
; ---- ----
; 71 71
sub al, mem8 ; Subtract memory -122 -122
; ---- ----
; -51 205+sign
mov ah, 119 ; Load register 119

sub al, ah ; and subtract -51
; ----
; 86+overflow

The INC and DEC instructions treat integers as unsigned values and do not
update the carry flag for signed carries and borrows.
When the sum of 8-bit signed operands exceeds 127, the processor sets the
overflow flag. (The overflow flag is also set if both operands are negative and
the sum is less than or equal to -128.) Placing a JO (Jump on Overflow) or
INTO (Interrupt on Overflow) instruction in your program at this point can
transfer control to error-recovery statements. When the sum exceeds 255, the
processor sets the carry flag. A JC (Jump on Carry) instruction at this point can
transfer control to error-recovery statements.
In the previous subtraction example, the processor sets the sign flag if the result
goes below 0. At this point, you can use a JS (Jump on Sign) instruction to
transfer control to error-recovery statements. Jump instructions are described in
the “Jumps” section in Chapter 7.
Adding and Subtracting in Multiple Registers

You can add and subtract numbers larger than the register size on your
processor with the ADC (Add with Carry) and SBB (Subtract with Borrow)
instructions. If the operations prior to an ADC or SBB instruction do not set the
carry flag, these instructions are identical to ADD and SUB. When you operate
on large values in more than one register, use ADD and SUB for the least
significant part of the number and ADC or SBB for the most significant part.
The following example illustrates multiple-register addition and subtraction. You
can also use this technique with 64-bit operands on the 80386/486 processors.
.DATA
mem32 DWORD 316423
mem32a DWORD 316423
mem32b DWORD 156739
.CODE
.
.
.
; Addition
mov ax, 43981 ; Load immediate 43981
sub dx, dx ; into DX:AX
add ax, WORD PTR mem32[0] ; Add to both + 316423
adc dx, WORD PTR mem32[2] ; memory words ------
; Result in DX:AX 360404
; Subtraction
mov ax, WORD PTR mem32a[0] ; Load mem32 316423
mov dx, WORD PTR mem32a[2] ; into DX:AX
sub ax, WORD PTR mem32b[0] ; Subtract low - 156739
sbb dx, WORD PTR mem32b[2] ; then high ------
; Result in DX:AX 159684

For 32-bit registers on the 80386/486 processors, only two steps are necessary.
If your program needs to be assembled for more than one processor, you can
assemble the statements conditionally, as shown in this example:
.DATA
mem32 DWORD 316423
mem32a DWORD 316423
mem32b DWORD 156739
p386 TEXTEQU (@Cpu AND 08h)
.CODE
.
.
.
; Addition
IF p386
mov eax, 43981 ; Load immediate
add eax, mem32 ; Result in EAX
ELSE
.
. ; do steps in previous example
.
ENDIF
; Subtraction
IF p386
mov eax, mem32a ; Load memory
sub eax, mem32b ; Result in EAX
ELSE
.
. ; do steps in previous example
.
ENDIF
Since the status of the carry flag affects the results of calculations with ADC
and SBB, be sure to turn off the carry flag with the CLC (Clear Carry Flag)
instruction or use ADD or SUB for the first calculation, when appropriate.
Multiplying and Dividing Integers

The 8086 family of processors uses different multiplication and division
instructions for signed and unsigned integers. Multiplication and division
instructions also have special requirements depending on the size of the
operands and the processor the code runs on.
Using Multiplication Instructions

The MUL instruction multiplies unsigned numbers. IMUL multiplies signed
numbers. For both instructions, one factor must be in the accumulator register
(AL for 8-bit numbers, AX for 16-bit numbers, EAX for 32-bit numbers). The

other factor can be in any single register or memory operand. The result
overwrites the contents of the accumulator register.
Multiplying two 8-bit numbers produces a 16-bit result returned in AX.
Multiplying two 16-bit operands yields a 32-bit result in DX:AX. The 80386/486
processor handles 64-bit products in the same way in the EDX:EAX pair.
This example illustrates multiplication of signed 16- and 32-bit integers.
.DATA
mem16 SWORD -30000
.CODE
.
.
.
; 8-bit unsigned multiply
mov al, 23 ; Load AL 23
mov bl, 24 ; Load BL * 24
mul bl ; Multiply BL -----
; Product in AX 552
; overflow and carry set
; 16-bit signed multiply

mov ax, 50 ; Load AX 50
; -30000
imul mem16 ; Multiply memory -----
; Product in DX:AX -1500000
; overflow and carry set
A nonzero number in the upper half of the result (AH for byte, DX or EDX for
word) sets the overflow and carry flags.
On the 80186–80486 processors, the IMUL instruction supports three
additional operand combinations. The first syntax option allows for 16-bit
multipliers producing a 16-bit product or 32-bit multipliers for 32-bit products on
the 80386/486. The result overwrites the destination. The syntax for this
operation is:
IMUL register16, immediate
The second syntax option specifies three operands for IMUL. The first operand
must be a 16-bit register operand, the second a 16-bit memory (or register)
operand, and the third a 16-bit immediate operand. IMUL multiplies the
memory (or register) and immediate operands and stores the product in the
register operand with this syntax:
IMUL register16,{ memory16 | register16}, immediate

For the 80386/486 only, a third option for IMUL allows an additional operand
for multiplication of a register value by a register or memory value. The syntax
is:
IMUL register,{register | memory}
The destination can be any 16-bit or 32-bit register. The source must be the
same size as the destination.
In all of these options, products too large to fit in 16 or 32 bits set the overflow
and carry flags. The following examples show these three options for IMUL.
imul dx, 456 ; Multiply DX times 456 on 80186-80486
imul ax, [bx],6 ; Multiply the value pointed to by BX
; by 6 and put the result in AX
imul dx, ax ; Multiply DX times AX on 80386

imul ax, [bx] ; Multiply AX by the value pointed to
; by BX on 80386
The IMUL instruction with multiple operands can be used for either signed or
unsigned multiplication, since the 16-bit product is the same in either case. To
get a 32-bit result, you must use the single-operand version of MUL or IMUL.
Using Division Instructions

The DIV instruction divides unsigned numbers, and IDIV divides signed
numbers. Both return a quotient and a remainder.
Table 4.1 summarizes the division operations. The dividend is the number to be
divided, and the divisor is the number to divide by. The quotient is the result.
The divisor can be in any register or memory location except the registers where
the quotient and remainder are returned.
Table 4.1 Division Operations
Size of Dividend Size of
Operand Register Divisor Quotient Remainder
16 bits AX 8 bits AL AH
32 bits DX:AX 16 bits AX DX
64 bits (80386 EDX:EAX 32 bits EAX EDX
and 80486)

Unsigned division does not require careful attention to flags. The following
examples illustrate signed division, which can be more complex.
.DATA
mem16 SWORD -2000
mem32 SDWORD 500000
.CODE
.
.
.
; Divide 16-bit unsigned by 8-bit
mov ax, 700 ; Load dividend 700
mov bl, 36 ; Load divisor DIV 36
div bl ; Divide BL ------
; Quotient in AL 19
; Remainder in AH 16
; Divide 32-bit signed by 16-bit

mov ax, WORD PTR mem32[0] ; Load into DX:AX
mov dx, WORD PTR mem32[2] ; 500000
idiv mem16 ; DIV -2000
; Divide memory ------
; Quotient in AX -250
; Remainder in DX 0
; Divide 16-bit signed by 16-bit

mov ax, WORD PTR mem16 ; Load into AX -2000
cwd ; Extend to DX:AX
mov bx,-421 ; DIV -421
idiv bx ; Divide by BX -----
; Quotient in AX 4
; Remainder in DX -316
If the dividend and divisor are the same size, sign-extend or zero-extend the
dividend so that it is the length expected by the division instruction. See
“Extending Signed and Unsigned Integers,” earlier in this chapter.
Manipulating Numbers at the Bit Level

The instructions introduced so far in this chapter access numbers at the byte or
word level. The logical, shift, and rotate instructions described in this section
access individual bits in a number. You can use logical instructions to evaluate
characters and do other text and screen operations. The shift and rotate
instructions do similar tasks by shifting and rotating bits through registers. This
section reviews some applications of these bit-level operations.

Logical Instructions
The logical instructions AND, OR, and XOR compare bits in two operands.
Based on the results of the comparisons, the instructions alter bits in the first
(destination) operand. The logical instruction NOT also changes bits, but
operates on a single operand.
The following list summarizes these four logical instructions. The list makes
reference to the “destination bit,” meaning the bit in the destination operand.
The terms “both bits” and “either bit” refer to the corresponding bits in the
source and destination operands. These instructions include:
Instruction Sets Destination Bit If Clears Destination Bit If
AND Both bits set Either or both bits clear
OR Either or both bits set Both bits clear
XOR Either bit (but not both) set Both bits set or both clear
NOT Destination bit clear Destination bit set
Note Do not confuse logical instructions with the logical operators, which
perform these operations at assembly time, not run time. Although the names
are the same, the assembler recognizes the difference.
The following example shows the result of the AND, OR, XOR, and NOT
instructions operating on a value in the AX register and in a mask. A mask is
any number with a pattern of bits set for an intended operation.
mov ax, 035h ; Load value 00110101
and ax, 0FBh ; Clear bit 2 AND 11111011
; --------
; Value is now 31h 00110001
or ax, 016h ; Set bits 4,2,1 OR 00010110
; --------
; Value is now 37h 00110111
xor ax, 0ADh ; Toggle bits 7,5,3,2,0 XOR 10101101
; --------
; Value is now 9Ah 10011010
not ax ; Value is now 65h 01100101
The AND instruction clears unmasked bits — that is, bits not protected by 1 in
the mask. To mask off certain bits in an operand and clear the others, use an
appropriate masking value in the source operand. The bits of the mask should
be 0 for any bit positions you want to clear and 1 for any bit positions you want
to remain unchanged.

The OR instruction forces specific bits to 1 regardless of their current settings.

The bits of the mask should be 1 for any bit positions you want to set and 0 for
any bit positions you want to remain unchanged.
The XOR instruction toggles the value of specific bits on and off — that is,
reverses them from their current settings. This instruction sets a bit to 1 if the
corresponding bits are different or to 0 if they are the same. The bits of the
mask should be 1 for any bit positions you want to toggle and 0 for any bit
positions you want to remain unchanged.
The following examples show an application for each of these instructions. The
code illustrating the AND instruction converts a “y” or “n” read from the
keyboard to uppercase, since bit 5 is always clear in uppercase letters. In the
example for OR, the first statement is faster and uses fewer bytes than cmp
bx, 0. When the operands for XOR are identical, each bit cancels itself,
producing 0.
;AND example - converts characters to uppercase
mov ah, 7 ; Get character without echo
int 21h
and al, 11011111y ; Convert to uppercase by clearing bit 5
cmp al, 'Y' ; Is it Y?
je yes ; If so, do Yes actions
. ; Else do No actions
.
yes: .
;OR example - compares operand to 0

or bx, bx ; Compare to 0
jg positive ; BX is positive
jl negative ; BX is negative
; else BX is zero
;XOR example - sets a register to 0

xor cx, cx ; 2 bytes, 3 clocks on 8088
sub cx, cx ; 2 bytes, 3 clocks on 8088
mov cx, 0 ; 3 bytes, 4 clocks on 8088
On the 80386/486 processors, the BSF (Bit Scan Forward) and the BSR (Bit
Scan Reverse) instructions perform operations like those of the logical
instructions. They scan the contents of a register to find the first-set or last-set
bit. You can use BSF or BSR to find the position of a set bit in a mask or to
check if a register value is 0.
Shifting and Rotating Bits

The 8086-based processors provide a complete set of instructions for shifting
and rotating bits. Shift instructions move bits a specified number of places to the

right or left. The last bit in the direction of the shift goes into the carry flag, and
the first bit is filled with 0 or with the previous value of the first bit.
Rotate instructions also move bits a specified number of places to the right or
left. For each bit rotated, the last bit in the direction of the rotate operation
moves into the first bit position at the other end of the operand. With some
variations, the carry bit is used as an additional bit of the operand. Figure 4.2
illustrates the eight variations of shift and rotate instructions for 8-bit operands.
Notice that SHL and SAL are identical.

Figure 4.2 Shifts and Rotates
All shift instructions use the same format. Before the instruction executes, the
destination operand contains the value to be shifted; after the instruction
executes, it contains the shifted operand. The source operand contains the
number of bits to shift or rotate. It can be the immediate value 1 or the CL
register. The 8088 and 8086 processors do not accept any other values or
registers with these instructions.
Starting with the 80186 processor, you can use 8-bit immediate values larger
than 1 as the source operand for shift or rotate instructions, as shown here:
shr bx, 4 ; 9 clocks, 3 bytes on 80286
The following statements are equivalent if the program must run on the 8088 or
8086 processor:
mov cl, 4 ; 2 clocks, 3 bytes on 80286
shr bx, cl ; 9 clocks, 2 bytes on 80286
; 11 clocks, 5 bytes total
Masks for logical instructions can be shifted to new bit positions. For example,
an operand that masks off a bit or group of bits can be shifted to move the mask
to a different position, allowing you to mask off a different bit each time the
mask is used. This technique, illustrated in the following example, is useful only
if the mask value is unknown until run time.
.DATA
masker BYTE 00000010y ; Mask that may change at run time
.CODE
.
.
.
mov cl, 2 ; Rotate two at a time
mov bl, 57h ; Load value to be changed 01010111y
rol masker, cl ; Rotate two to left 00001000y
or bl, masker ; Turn on masked values ---------
; New value is 05Fh 01011111y
rol masker, cl ; Rotate two more 00100000y
or bl, masker ; Turn on masked values ---------
; New value is 07Fh 01111111y
Multiplying and Dividing with Shift Instructions

You can use the shift and rotate instructions (SHR, SHL, SAR, and SAL) for
multiplication and division. Shifting a value right by one bit has the effect of
dividing by two; shifting left by 1 bit has the effect of multiplying by two. You
can take advantage of shifts to do fast multiplication and division by powers of

two. For example, shifting left twice multiplies by four, shifting left three times
multiplies by eight, and so on.
Use SHR (Shift Right) to divide unsigned numbers. You can use SAR (Shift
Arithmetic Right) to divide signed numbers, but SAR rounds negative numbers
down — IDIV always rounds negative numbers up (toward 0). Division using
SAR must adjust for this difference. Multiplication by shifting is the same for
signed and unsigned numbers, so you can use either SAL or SHL.
Multiply and divide instructions are relatively slow, particularly on the 8088 and
8086 processors. When multiplying or dividing by a power of two, use shifts to
speed operations by a factor of 10 or more. For example, these statements take
only four clocks on an 8088 or 8086 processor:
sub ah, ah ; Clear AH
shl ax, 1 ; Multiply byte in AL by 2
The following statements produce the same results, but take between 74 and 81
clocks on the 8088 or 8086 processors. The same statements take 15 clocks on
the 80286 and between 11 and 16 clocks on the 80386. (For a discussion about
instruction timings, see “A Word on Instruction Timings” in the Introduction.)
mov bl, 2 ; Multiply byte in AL by 2
mul bl
As the following macro shows, it’s possible to multiply by any number — in this
case, 10 — without resorting to the MUL instruction. However, such a
procedure is no more than an interesting arithmetic exercise, since the additional
code almost certainly takes more time to execute than a single MUL. You
should consider using shifts in your program only when multiplying or dividing
by a power of two.
mul_10 MACRO factor ; Factor must be unsigned
mov ax, factor ; Load into AX
shl ax, 1 ; AX = factor * 2
mov bx, ax ; Save copy in BX
add ax, bx ; AX = (factor * 8) + (factor * 2)
ENDM ; AX = factor * 10
Here’s another macro that divides by 512. In contrast to the previous example,
this macro uses little code and operates faster than an equivalent DIV
instruction.

div_512 MACRO dividend ; Dividend must be unsigned

mov ax, dividend ; Load into AX
shr ax, 1 ; AX = dividend / 2 (unsigned)
xchg al, ah ; XCHG is like rotate right 8
; AL = (dividend / 2) / 256
cbw ; Clear upper byte
ENDM ; AX = (dividend / 512)
If you need to shift a value that is too large to fit in one register, you can shift
each part separately. The RCR (Register Carry Right) and RCL (Register Carry
Left) instructions carry values from the first register to the second by passing the
leftmost or rightmost bit through the carry flag.
This example shifts a multiword value.
.DATA
mem32 DWORD 500000
.CODE
; Divide 32-bit unsigned by 16

mov cx, 4 ; Shift right 4 500000
again: shr WORD PTR mem32[2], 1 ; Shift into carry DIV 16
rcr WORD PTR mem32[0], 1 ; Rotate carry in ------
loop again ; 31250
Since the carry flag is treated as part of the operand (it’s like using a 9-bit or 17-
bit operand), the flag value before the operation is crucial. The carry flag can be
adjusted by a previous instruction, but you can also set or clear the flag directly
with the CLC (Clear Carry Flag), CMC (Complement Carry Flag), and STC
(Set Carry Flag) instructions.
On the 80386 and 80486 processors, an alternate method for multiplying quickly
by constants takes advantage of the LEA (Load Effective Address) instruction
and the scaling of indirect memory operands. By using a 32-bit value as both the
index and the base register in an indirect memory operand, you can multiply by
the constants 2, 3, 4, 5, 8, and 9 more quickly than you can by using the MUL
instruction. LEA calculates the offset of the source operand and stores it into
the destination register, EBX, as this example shows:
lea ebx, [eax*2] ; EBX = 2 * EAX
lea ebx, [eax*2+eax] ; EBX = 3 * EAX
Scaling of 80386 indirect memory operands is reviewed in “Indirect Memory

Operands with 32-Bit Registers” in Chapter 3. LEA is introduced in “Loading
Addresses into Registers” in Chapter 3.

The next chapter deals with more complex data types — arrays, strings,
structures, unions, and records. Many of the operations presented in this chapter
can also be applied to the data structures covered in Chapter 5, “Defining and
Using Complex Data Types.”

105
C H A P T E R 5
Defining and Using Complex

Data Types
With the complex data types available in MASM 6.1 — arrays, strings, records,
structures, and unions — you can access data as a unit or as individual elements
that make up a unit. The individual elements of complex data types are often the
integer types discussed in Chapter 4, “Defining and Using Simple Data Types.”
“Arrays and Strings” reviews how to declare, reference, and initialize arrays and
strings. This section summarizes the general steps needed to process arrays and
strings and describes the MASM instructions for moving, comparing, searching,
loading, and storing.
“Structures and Unions” covers similar information for structures and unions:
how to declare structure and union types, how to define structure and union
variables, and how to reference structures and unions and their fields.
“Records” explains how to declare record types, define record variables, and use
record operators.
Arrays and Strings

An array is a sequential collection of variables, all of the same size and type,
called “elements.” A string is an array of characters. For example, in the string
“ABC,” each letter is an element. You can access the elements in an array or
string relative to the first element. This section explains how to handle arrays
and strings in your programs.
Declaring and Referencing Arrays

Array elements occupy memory contiguously, so a program references each
element relative to the start of the array. To declare an array, supply a label
name, the element type, and a series of initializing values or ? placeholders. The
following examples declare the arrays warray and xarray:
warray WORD 1, 2, 3, 4
xarray DWORD 0FFFFFFFFh, 789ABCDEh

Initializer lists of array declarations can span multiple lines. The first initializer
must appear on the same line as the data type, all entries must be initialized,
and, if you want the array to continue to the new line, the line must end with a
comma. These examples show legal multiple-line array declarations:
big BYTE 21, 22, 23, 24, 25,
26, 27, 28
somelist WORD 10,

20,
30
If you do not use the LENGTHOF and SIZEOF operators discussed later in
this section, an array may span more than one logical line, although a separate
type declaration is needed on each logical line:
var1 BYTE 10, 20, 30
BYTE 40, 50, 60
BYTE 70, 80, 90
The DUP Operator

You can also declare an array with the DUP operator. This operator works with
any of the data allocation directives described in “Allocating Memory for Integer
Variables” in Chapter 4. In the syntax
count DUP (initialvalue [[, initialvalue]]...)
the count value sets the number of times to repeat all values within the
parentheses. The initialvalue can be an integer, character constant, or another
DUP operator, and must always appear within parentheses. For example, the
statement
barray BYTE 5 DUP (1)
allocates the integer 1 five times for a total of 5 bytes.

The following examples show various ways to allocate data elements with the
DUP operator:
array DWORD 10 DUP (1) ; 10 doublewords
; initialized to 1
buffer BYTE 256 DUP (?) ; 256-byte buffer
masks BYTE 20 DUP (040h, 020h, 04h, 02h) ; 80-byte buffer

; with bit masks
three_d DWORD 5 DUP (5 DUP (5 DUP (0))) ; 125 doublewords
; initialized to 0

Chapter 5 Defining and Using Complex Data Types 107
Referencing Arrays
Each element in an array is referenced with an index number, beginning with
zero. The array index appears in brackets after the array name, as in
array[9]
Assembly-language indexes differ from indexes in high-level languages, where

the index number always corresponds to the element’s position. In C, for
example, array[9] references the array’s tenth element, regardless of whether
each element is 1 byte or 8 bytes in size.
In assembly language, an element’s index refers to the number of bytes between
the element and the start of the array. This distinction can be ignored for arrays
of byte-sized elements, since an element’s position number matches its index.
For example, defining the array
prime BYTE 1, 3, 5, 7, 11, 13, 17
gives a value of 1 to prime[0], a value of 3 to prime[1], and so forth.

However, in arrays with elements larger than 1 byte, index numbers (except
zero) do not correspond to an element’s position. You must multiply an
element’s position by its size to determine the element’s index. Thus, for the
array
wprime WORD 1, 3, 5, 7, 11, 13, 17
wprime[4] represents the third element (5), which is 4 bytes from the
beginning of the array. Similarly, the expression wprime[6] represents the
fourth element (7) and wprime[10] represents the sixth element (13).
The following example determines an index at run time. It multiplies the position
by two (the size of a word element) by shifting it left:
mov si, cx ; CX holds position number
shl si, 1 ; Scale for word referencing
mov ax, wprime[si] ; Move element into AX
The offset required to access an array element can be calculated with the
following formula:
nth element of array = array[(n-1) * size of element]
Referencing an array element by distance rather than position is not difficult to
master, and is actually very consistent with how assembly language works.
Recall that a variable name is a symbol that represents the contents of a
particular address in memory. Thus, if the array wprime begins at address
DS:2400h, the reference wprime[6] means to the processor “the word value
contained in the DS segment at offset 2400h-plus-6-bytes.”

As described in “Direct Memory Operands,” Chapter 3, you can substitute the

plus operator (+) for brackets, as in:
wprime[9]
wprime+9
Since brackets simply add a number to an address, you don’t need them when
referencing the first element. Thus, wprime and wprime[0] both refer to the
first element of the array wprime.
If your program runs only on an 80186 processor or higher, you can use the
BOUND instruction to verify that an index value is within the bounds of an
array. For a description of BOUND, see the Reference.
LENGTHOF, SIZEOF, and TYPE for Arrays

When applied to arrays, the LENGTHOF, SIZEOF, and TYPE operators
return information about the length and size of the array and about the type of
the
initializers.
The LENGTHOF operator returns the number of elements in the array. The
SIZEOF operator returns the number of bytes used by the initializers in the
array definition. TYPE returns the size of the elements of the array. The
following examples illustrate these operators:
array WORD 40 DUP (5)
larray EQU LENGTHOF array ; 40 elements

sarray EQU SIZEOF array ; 80 bytes
tarray EQU TYPE array ; 2 bytes per element
num DWORD 4, 5, 6, 7, 8, 9, 10, 11
lnum EQU LENGTHOF num ; 8 elements

snum EQU SIZEOF num ; 32 bytes
tnum EQU TYPE num ; 4 bytes per element
warray WORD 40 DUP (40 DUP (5))
len EQU LENGTHOF warray ; 1600 elements

siz EQU SIZEOF warray ; 3200 bytes
typ EQU TYPE warray ; 2 bytes per element
Declaring and Initializing Strings

A string is an array of characters. Initializing a string like "Hello, there"
allocates and initializes 1 byte for each character in the string. An initialized
string can be no longer than 255 characters.

For data directives other than BYTE, a string may initialize only the first
element. The initializer value must fit into the specified size and conform to the
expression word size in effect (see “Integer Constants and Constant
Expressions” in Chapter 1), as shown in these examples:
wstr WORD "OK"
dstr DWORD "DATA" ; Legal under EXPR32 only
As with arrays, string initializers can span multiple lines. The line must end with
a comma if you want the string to continue to the next line.
str1 BYTE "This is a long string that does not ",
"fit on one line."
You can also have an array of pointers to strings.

PBYTE TYPEDEF PTR BYTE
.DATA
msg1 BYTE "Operation completed successfully."
msg2 BYTE "Unknown command"
msg3 BYTE "File not found"
pmsg PBYTE msg1 ; pmsg is an array
PBBYTE msg2 ; of pointers to
PBYTE msg3 ; above messages
Strings must be enclosed in single (') or double (") quotation marks. To put a
single quotation mark inside a string enclosed by single quotation marks, use two
single quotation marks. Likewise, if you need quotation marks inside a string
enclosed by double quotation marks, use two sets. These examples show the
various uses of quotation marks:
char BYTE 'a'
message BYTE "That's the message." ; That's the message.
warn BYTE 'Can''t find file.' ; Can't find file.
string BYTE "This ""value"" not found." ; This "value" not found.
You can always use single quotation marks inside a string enclosed by double
quotation marks, as the initialization for message shows, and vice versa.
The ? Initializer
You do not have to initialize an array. The ? operator lets you allocate space for
the array without placing specific values in it. Object files contain records for
initialized data. Unspecified space left in the object file means that no records
contain initialized data for that address. The actual values stored in arrays
allocated with ? depend on certain conditions. The ? initializer is treated as a
zero in a DUP statement that contains initializers in addition to the ? initializer.
If the ? initializer does not appear in a DUP statement, or if the DUP statement
contains only ? initializers, the assembler leaves the allocated space unspecified.

LENGTHOF, SIZEOF, and TYPE for Strings

Because strings are simply arrays of byte elements, the LENGTHOF, SIZEOF,
and TYPE operators behave as you would expect, as illustrated in this example:
msg BYTE "This string extends ",
"over three ",
"lines."
lmsg EQU LENGTHOF msg ; 37 elements

smsg EQU SIZEOF msg ; 37 bytes
tmsg EQU TYPE msg ; 1 byte per element
Processing Strings
The 8086-family instruction set has seven string instructions for fast and
efficient processing of entire strings and arrays. The term “string” in “string
instructions” refers to a sequence of elements, not just character strings. These
instructions work directly only on arrays of bytes and words on the 8086–80486
processors, and on arrays of bytes, words, and doublewords on the 80386/486
processors. Processing larger elements must be done indirectly with loops.
The following list gives capsule descriptions of the five instructions discussed in
this section.
Instruction Description
MOVS Copies a string from one location to another
STOS Stores contents of the accumulator register to a string
CMPS Compares one string with another
LODS Loads values from a string to the accumulator register
SCAS Scans a string for a specified value
All of these instructions use registers in a similar way and have a similar syntax.
Most are used with the repeat instruction prefixes REP, REPE (or REPZ), and
REPNE (or REPNZ). REPZ is a synonym for REPE (Repeat While Equal) and
REPNZ is a synonym for REPNE (Repeat While Not Equal).
This section first explains the general procedures for using all string instructions.
It then illustrates each instruction with an example.
Overview of String Instructions

The string instructions have specific requirements for the location of strings and
the use of registers. To operate on any string, follow these three steps:
1. Set the direction flag to indicate the direction in which you want to process
the string. The STD instruction sets the flag, while CLD clears it.

If the direction flag is clear, the string is processed upward (from low
addresses to high addresses, which is from left to right through the string). If
the direction flag is set, the string is processed downward (from high
addresses to low addresses, or from right to left). Under MS-DOS, the
direction flag is normally clear if your program has not changed it.
2. Load the number of iterations for the string instruction into the CX register.
If you want to process 100 elements in a string, move 100 into CX. If you
wish the string instruction to terminate conditionally (for example, during a
search when a match is found), load the maximum number of iterations that
can be performed without an error.
3. Load the starting offset address of the source string into DS:SI and the
starting address of the destination string into ES:DI. Some string instructions
take only a destination or source, not both (see Table 5.1).
Normally, the segment address of the source string should be DS, but you
can use a segment override to specify a different segment for the source
operand. You cannot override the segment address for the destination string.
Therefore, you may need to change the value of ES. For information on
changing segment registers, see “Programming Segmented Addresses” in
Chapter 3.
Note Although you can use a segment override on the source operand, a
segment override combined with a repeat prefix can cause problems in certain
situations on all processors except the 80386/486. If an interrupt occurs during
the string operation, the segment override is lost and the rest of the string
operation processes incorrectly. Segment overrides can be used safely when
interrupts are turned off or with the 80386/486 processors.
You can adapt these steps to the requirements of any particular string operation.
The syntax for the string instructions is:
[[prefix]] CMPS [[segmentregister:]] source, [[ES:]] destination
LODS [[segmentregister:]] source
[[prefix]] MOVS [[ES:]] destination, [[segmentregister:]] source
[[prefix]] SCAS [[ES:]] destination
[[prefix]] STOS [[ES:]] destination
Some instructions have special forms for byte, word, or doubleword operands.
If you use the form of the instruction that ends in B (BYTE), W (WORD), or D
(DWORD) with LODS, SCAS, and STOS, the assembler knows whether the
element is in the AL, AX, or EAX register. Therefore, these instruction forms
do not require operands.

Table 5.1 lists each string instruction with the type of repeat prefix it uses and
indicates whether the instruction works on a source, a destination, or both.
Table 5.1 Requirements for String Instructions
Instruction Repeat Prefix Source/Destination Register Pair
MOVS REP Both DS:SI, ES:DI
SCAS REPE/REPNE Destination ES:DI
CMPS REPE/REPNE Both DS:SI, ES:DI
LODS None Source DS:SI
STOS REP Destination ES:DI
INS REP Destination ES:DI
OUTS REP Source DS:SI
The repeat prefix causes the instruction that follows it to repeat for the number
of times specified in the count register or until a condition becomes true. After
each iteration, the instruction increments or decrements SI and DI so that it
points to the next array element. The direction flag determines whether SI and
DI are incremented (flag clear) or decremented (flag set). The size of the
instruction determines whether SI and DI are altered by 1, 2, or 4 bytes each
time.
Each prefix governs the number of repetitions as follows:
Prefix Description
REP Repeats instruction CX times
REPE, REPZ Repeats instruction maximum CX times while values are equal
REPNE, REPNZ Repeats instruction maximum CX times while values are not
equal
The prefixes apply to only one string instruction at a time. To repeat a block of
instructions, use a loop construction. (See “Loops” in Chapter 7.)
At run time, if a string instruction is preceded by a repeat sequence, the
processor:
1. Checks the CX register and exits if CX is 0.
2. Performs the string operation once.
3. Increases SI and/or DI if the direction flag is clear. Decreases SI and/or DI if
the direction flag is set. The amount of increase or decrease is 1 for byte
operations, 2 for word operations, and 4 for doubleword operations.
4. Decrements CX without modifying the flags.

5. Checks the zero flag (for SCAS or CMPS) if the REPE or REPNE prefix is
used. If the repeat condition holds, loops back to step 1. Otherwise, the loop
ends and execution proceeds to the next instruction.
When the repeat loop ends, SI (or DI) points to the position following a match
(when using SCAS or CMPS), so you need to decrement or increment DI or SI
to point to the element where the last match occurred.
Although string instructions (except LODS) are used most often with repeat
prefixes, they can also be used by themselves. In these cases, the SI and/or DI
registers are adjusted as specified by the direction flag and the size of operands.
Using String Instructions

To use the 8086-family string instructions, follow the steps outlined in the
previous section. Examples in this section illustrate each instruction.
You can also use the techniques in this section with structures and unions, since
arrays and strings can be fields in structures and unions. (See the section
“Structures and Unions,” following.)
Moving Array Data

The MOVS instruction copies data from one area of memory to another. To
move data, first load the count, source and destination addresses into the
appropriate registers. Then use REP with the MOVS instruction.
.MODEL small
.DATA
source BYTE 10 DUP ('0123456789')
destin BYTE 100 DUP (?)
.CODE
mov ax, @data ; Load same segment
mov ds, ax ; to both DS
mov es, ax ; and ES
.
.
.
cld ; Work upward
mov cx, LENGTHOF source ; Set iteration count to 100
mov si, OFFSET source ; Load address of source
mov di, OFFSET destin ; Load address of destination
rep movsb ; Move 100 bytes
Filling Arrays
The STOS instruction stores a specified value in each position of a string. The
string is the destination, so it must be pointed to by ES:DI. The value to store
must be in the accumulator.

The next example stores the character 'a' in each byte of a 100-byte string,
filling the entire string with “aaaa....” Notice how the code stores 50 words
rather than

100 bytes. This makes the fill operation faster by reducing the number of
iterations. To fill an odd number of bytes, you need to adjust for the last byte.
.MODEL small, C
.DATA
destin BYTE 100 DUP (?)
ldestin EQU (LENGTHOF destin) / 2
.CODE
. ; Assume ES = DS
.
.
cld ; Work upward
mov ax, 'aa' ; Load character to fill
mov cx, ldestin ; Load length of string
mov di, OFFSET destin ; Load address of destination
rep stosw ; Store 'aa' into array
Comparing Arrays
The CMPS instruction compares two strings and points to the address after
which a match or nonmatch occurs. If the values are the same, the zero flag is
set. Either string can be considered the destination or the source unless a
segment override is used. This example using CMPSB assumes that the strings
are in different segments. Both segments must be initialized to the appropriate
segment register.

.MODEL large, C
.DATA
string1 BYTE "The quick brown fox jumps over the lazy dog"
.FARDATA
string2 BYTE "The quick brown dog jumps over the lazy fox"
lstring EQU LENGTHOF string2
.CODE
mov ax, @data ; Load data segment
mov ds, ax ; into DS
mov ax, @fardata ; Load far data segment
mov es, ax ; into ES
.
.
.
cld ; Work upward
mov cx, lstring ; Load length of string
mov si, OFFSET string1 ; Load offset of string1
mov di, OFFSET string2 ; Load offset of string2
repe cmpsb ; Compare
je allmatch ; Jump if all match
.
.
.
allmatch: ; Special case for all match
Loading Data from Arrays

The LODS instruction loads a value from a string into the accumulator register.
This instruction is not used with a repeat instruction prefix, since continually
reloading the accumulator serves no purpose.
The code in this example loads, processes, and displays each byte in a string.

.DATA
info BYTE 0, 1, 2, 3, 4, 5, 6, 7, 8, 9
linfo WORD LENGTHOF info
.CODE
.
.
.
cld ; Work upward
mov cx, linfo ; Load length
mov si, OFFSET info ; Load offset of source
mov ah, 2 ; Display character function
get:
lodsb ; Get a character
add al, '0' ; Convert to ASCII
mov dl, al ; Move to DL
int 21h ; Call DOS to display character
loop get ; Repeat
Searching Arrays
The SCAS instruction compares the value pointed to by ES:DI with the value in
the accumulator. If both values are the same, it sets the zero flag.
A repeat prefix lets SCAS work on an entire string, scanning (from which SCAS
gets its name) for a particular value called the target. REPNE SCAS sets the
zero flag if it finds the target value in the array. REPE SCAS sets the zero flag
if the scanned array contains nothing but the target value.

This example assumes that ES is not the same as DS and that the address of the
string is stored in a pointer variable. The LES instruction loads the far address
of the string into ES:DI.
.DATA
string BYTE "The quick brown fox jumps over the lazy dog"
pstring PBYTE string ; Far pointer to string
lstring EQU LENGTHOF string ; Length of string
.CODE
.
.
.
cld ; Work upward
mov cx, lstring ; Load length of string
les di, pstring ; Load address of string
mov al, 'z' ; Load character to find
repne scasb ; Search
jne notfound ; Jump if not found
. ; ES:DI points to character
. ; after first 'z'
.
notfound: ; Special case for not found
Translating Data in Byte Arrays

The XLAT (Translate) instruction copies a byte from an array of bytes into the
AL register. The instruction takes its name from its ability to translate an
element’s number into the element itself. For example, given the number 7,
XLAT returns byte #7 from the array. The array may hold byte-sized integers
or, very often, a table or list of characters. The syntax for XLAT is:
XLAT[[B]] [[[[segment:]]memory]]
The optional B suffix (for “byte”) reflects the size of data the instruction
handles. Both XLAT and XLATB assemble to exactly the same machine code.
To use XLAT, place the offset of the start of the array in the BX register and
the desired index value in AL. Array indexes always begin with 0 in assembly
language. To retrieve the first byte of the array, set AL to 0; to retrieve the
second byte, set AL to 1, and so forth. XLAT returns the byte element in AL,
overwriting the index number.
By default, the DS register contains the segment of the table, but you can use a
segment override to specify a different segment. You need not give an operand
except when specifying a segment override. (For information about the segment
override operator, see “Direct Memory Operands” in Chapter 3.)

This example illustrates XLAT by looking up hexadecimal characters in a list.

The code converts an eight-bit binary number to a string representing a
hexadecimal number.
; Table of hexadecimal digits
hex BYTE "0123456789ABCDEF"
convert BYTE "You pressed the key with ASCII code "
key BYTE ?,?,"h",13,10,"$"
.CODE
.
.
.
mov ah, 8 ; Get a key in AL
int 21h ; Call DOS
mov bx, OFFSET hex ; Load table address
mov ah, al ; Save a copy in high byte
and al, 00001111y ; Mask out top character
xlat ; Translate
mov key[1], al ; Store the character
mov cl, 12 ; Load shift count
shr ax, cl ; Shift high char into position
xlat ; Translate
mov key, al ; Store the character
mov dx, OFFSET convert ; Load message
mov ah, 9 ; Display character
int 21h ; Call DOS
Although AL cannot contain an index value greater than 255, you can use
XLAT with arrays containing more than 256 elements. Simply treat each 256-
byte block of the array as a smaller sub-array. For example, to retrieve the
260th element of an array, add 256 to BX and set AL=3 (260-256-1).
Structures and Unions

A structure is a group of possibly dissimilar data types and variables that can be
accessed as a unit or by any of its components. The fields within the structure
can have different sizes and data types.
Unions are identical to structures, except that the fields of a union overlap in
memory, which allows you to define different data formats for the same
memory space. Unions can store different types of data depending on the
situation. They also can store data as one data type and retrieve it as another
data type.
Whereas each field in a structure has an offset relative to the first byte of the
structure, all the fields in a union start at the same offset. The size of a structure
is the sum of its components; the size of a union is the length of the longest
field.

A MASM structure is similar to a struct in the C language, a STRUCTURE in

FORTRAN, and a RECORD in Pascal. Unions in MASM are similar to unions
in C and FORTRAN, and to variant records in Pascal.
Follow these steps when using structures and unions:
1. Declare a structure (or union) type.
2. Define one or more variables having that type.
3. Reference the fields directly or indirectly with the field (dot) operator.
You can use the entire structure or union variable or just the individual fields as
operands in assembler statements. This section explains the allocating,
initializing, and nesting of structures and unions.
MASM 6.1 extends the functionality of structures and also makes some changes
to MASM 5.1 behavior. If you prefer, you can retain MASM 5.1 behavior by
specifying OPTION OLDSTRUCTS in your program.
Declaring Structure and Union Types

When you declare a structure or union type, you create a template for data. The
template states the sizes and, optionally, the initial values in the structure or
union, but allocates no memory.
The STRUCT keyword marks the beginning of a type declaration for a
structure. (STRUCT and STRUC are synonyms.) The format for STRUCT
and UNION type declarations is:
name {STRUCT | UNION} [[alignment]] [[,NONUNIQUE ]]
fielddeclarations
name ENDS
The fielddeclarations is a series of one or more variable declarations. You can
declare default initial values individually or with the DUP operator. (See
“Defining Structure and Union Variables,” following.) “Referencing Structures,
Unions, and Fields,” later in this chapter, explains the NONUNIQUE keyword.
You can nest structures and unions, as explained in “Nested Structures and
Unions,” also later in this chapter.
Initializing Fields
If you provide initializers for the fields of a structure or union when you declare
the type, these initializers become the default value for the fields when you
define a variable of that type. “Defining Structure and Union Variables,”
following, explains default initializers.

When you initialize the fields of a union type, the type and value of the first field
become the default value and type for the union. In this example of an initialized
union declaration, the default type for the union is DWORD:
DWB UNION
d DWORD 00FFh
w WORD ?
b BYTE ?
DWB ENDS
If the size of the first member is less than the size of the union, the assembler
initializes the rest of the union to zeros. When initializing strings in a type, make
sure the initial values are long enough to accommodate the largest possible
string.
Field Names
Structure and union field names must be unique within a nesting level because
they represent the offset from the beginning of the structure to the
corresponding field.
A label elsewhere in the code may have the same name as a structure field, but
a text macro cannot. Also, field names between structures need not be unique.
Field names must be unique if you place OPTION M510 or OPTION
OLDSTRUCTS in your code or use the /Zm option from the command line,
since versions of MASM prior to 6.0 require unique field names. (See Appendix
A.)
Alignment Value and Offsets for Structures

Data access to structures is faster on aligned fields than on unaligned fields.
Therefore, alignment gains speed at the cost of space. Alignment improves
access on 16-bit and 32-bit processors but makes no difference in programs
executing on an 8-bit 8088 processor.
The way the assembler aligns structure fields determines the amount of space
required to store a variable of that type. Each field in a structure has an offset
relative to 0. If you specify an alignment in the structure declaration (or with the
/Zpn command-line option), the offset for each field may be modified by the
alignment (or n).
The only values accepted for alignment are 1, 2, and 4. The default is 1. If the
type declaration includes an alignment, each field is aligned to either the field’s
size or the alignment value, whichever is less. If the field size in bytes is greater
than the alignment value, the field is padded so that its offset is evenly divisible
by the alignment value. Otherwise, the field is padded so that its offset is evenly
divisible by the field size.

Any padding required to reach the correct offset for the field is added prior to
allocating the field. The padding consists of zeros and always precedes the
aligned field. The size of the structure must also be evenly divisible by the
structure alignment value, so zeros may be added at the end of the structure.
If neither the alignment nor the /Zp command-line option is used, the offset is
incremented by the size of each data directive. This is the same as a default
alignment equal to 1. The alignment specified in the type declaration overrides
the /Zp command-line option.
These examples show how the assembler determines offsets:
STUDENT2 STRUCT 2 ; Alignment value is 2
score WORD 1 ; Offset = 0
id BYTE 2 ; Offset = 2 (1 byte padding added)
year DWORD 3 ; Offset = 4
sname BYTE 4 ; Offset = 8 (1 byte padding added)
STUDENT2 ENDS
One byte of padding is added at the end of the first byte-sized field. Otherwise,
the offset of the year field would be 3, which is not divisible by the alignment
value of 2. The size of this structure is now 9 bytes. Since 9 is not evenly
divisible by 2, 1 byte of padding is added at the end of student2.
STUDENT4 STRUCT 4 ; Alignment value is 4
sname BYTE 1 ; Offset = 0 (1 byte padding added)
score WORD 10 DUP (100) ; Offset = 2
year BYTE 2 ; Offset = 22 (1 byte padding
; added so offset of next field
; is divisible by 4)
id DWORD 3 ; Offset = 24
STUDENT4 ENDS
The alignment value affects the alignment of structure variables, so adding an

alignment value affects memory usage. This feature provides compatibility with
structures in Microsoft C. MASM 6.1 provides an improved H2INC utility,
which C programmers can use to translate C structures to assembly. (See
Environment and Tools, Chapter 20.)
The ALIGN, EVEN, and ORG directives can modify how field offsets are
placed during structure definition. The EVEN and ALIGN directives insert
padding bytes to round the field offset up to the specified alignment boundary.
The ORG directive changes the offset of the next field to a given value, either
positive or negative. If you use ORG when declaring a structure, you cannot
define a structure of that type. ORG is useful when accessing existing data
structures, such as a stack frame created by a high-level language.

Defining Structure and Union Variables

Once you have declared a structure or union type, you can define variables of
that type. For each variable defined, memory is allocated in the current segment
in the format declared by the type. The syntax for defining a structure or union
variable is:
[[name]] typename < [[initializer [[,initializer]]...]] >
[[name]] typename { [[initializer [[,initializer]]...]] }
[[name]] typename constant DUP ({ [[initializer [[,initializer]]...]] })
The name is the label assigned to the variable. If you do not provide a name, the
assembler allocates space for the variable but does not give it a symbolic name.
The typename is the name of a previously declared structure or union type.
You can give an initializer for each field. Each initializer must correspond in
type with the field defined in the type declaration. For unions, the type of the
initializer must be the same as the type for the first field. An initialization list can
also use the DUP operator.
The list of initializers can be broken only after a comma unless you end the line
with a continuation character (\). The last curly brace or angle bracket must
appear on the same line as the last initializer. You can also use the line
continuation character to extend a line as shown in the Item4 declaration that
follows. Angle brackets and curly braces can be intermixed in an initialization as
long as they match. This example illustrates the options for initializing lists in
structures of type ITEMS:

ITEMS STRUCT
Iname BYTE 'Item Name'
Inum WORD ?
UNION ITYPE ; UNION keyword appears first
oldtype BYTE 0 ; when nested in structure.
newtype WORD ? ; (See "Nested Structures
ENDS ; and Unions," following ).
ITEMS ENDS
.
.
.
.DATA
Item1 ITEMS < > ; Accepts default initializers
Item2 ITEMS { } ; Accepts default initializers
Item3 ITEMS <'Bolts', 126> ; Overrides default value of first
; 2 fields; use default of
; the third field
Item4 ITEMS { \
'Bolts', ; Item name
126 \ ; Part number
}
The example defines — that is, allocates space for — four structures of the
ITEMS type. The structures are named Item1 through Item4. Each definition
requires the angle brackets or curly braces even when not initialized. If you
initialize more than one field, separate the values with commas, as shown in
Item3 and Item4.
You need not initialize all fields in a structure. If a field is blank, the assembler
uses the structure’s initial value given for that field in the declaration. If there is
no default value, the field value is left unspecified.
For nested structures or unions, however, these are equivalent:
Item5 ITEMS {'Bolts', , }
Item6 ITEMS {'Bolts', , { } }
A variable and an array of union type WB look like this:

WB UNION
w WORD ?
b BYTE ?
WB ENDS
num WB {0Fh} ; Store 0Fh

array WB (40 / SIZEOF WB) DUP ({2}) ; Allocates and
; initializes 20 unions
Arrays as Field Initializers

The size of the initializer determines the length of the array that can override the
contents of a field in a variable definition. The override cannot contain more
elements than the default. Specifying fewer override array elements changes the
first n values of the default where n is the number of values in the override. The
rest of the array elements take their default values from the initializer.
Strings as Field Initializers

If the override is shorter, the assembler pads the override with spaces to equal
the length of the initializer. If the initializer is a string and the override value is
not a string, the override value must be enclosed in angle brackets or curly
braces.
A string can override any member of type BYTE (or SBYTE). You need not
enclose the string in angle brackets or curly braces unless mixed with other
override methods.
If a structure has an initialized string field or an array of bytes, any new string
assigned to a variable of the field that is smaller than the default is padded with
spaces. The assembler adds four spaces at the end of 'Bolts' in the variables
of type ITEMS previously shown. The Iname field in the ITEMS structure
cannot contain a field initializer longer than 'Item Name'.
Structures as Field Initializers

Initializers for structure variables must be enclosed in curly braces or angle
brackets, but you can specify overrides with fewer elements than the defaults.
This example illustrates the use of default values with structures as field
initializers:

DISKDRIVES STRUCT
a1 BYTE ?
b1 BYTE ?
c1 BYTE ?
DISKDRIVES ENDS
INFO STRUCT
buffer BYTE 100 DUP (?)
crlf BYTE 13, 10
query BYTE 'Filename: ' ; String <= can override
endmark BYTE 36
drives DISKDRIVES <0, 1, 1>
INFO ENDS
info1 INFO { , , 'Dir' }
; Next line illegal since name in query field is too long:

; info2 INFO {"TESTFILE", , "DirectoryName"}
lotsof INFO { , , 'file1', , {0,0,0} },

{ , , 'file2', , {0,0,1} },
{ , , 'file3', , {0,0,2} }
The following diagram shows how the assembler stores info1.
The initialization for drives gives default values for all three fields of the
structure. The fields left blank in info1 use the default values for those fields.
The info2 declaration is illegal because “DirectoryName” is longer than the
initial string for that field.
Arrays of Structures and Unions

You can define an array of structures using the DUP operator (see “Declaring
and Referencing Arrays,” earlier in this chapter) or by creating a list of
structures. For example, you can define an array of structure variables like this:

Item7 ITEMS 30 DUP ({,,{10}})
The Item7 array defined here has 30 elements of type ITEMS, with the third
field of each element (the union) initialized to 10.
You can also list array elements as shown in the following example.
Item8 ITEMS {'Bolts', 126, 10},
{'Pliers',139, 10},
{'Saws', 414, 10}
Redeclaring a Structure
The assembler generates an error when you declare a structure more than once
unless the following are the same:
u Field names
u Offsets of named fields
u Initialization lists
u Field alignment value
LENGTHOF, SIZEOF, and TYPE for Structures

The size of a structure determined by SIZEOF is the offset of the last field, plus
the size of the last field, plus any padding required for proper alignment. (For
information about alignment, see “Declaring Structure and Union Types,” earlier
in this chapter.)

This example, using the preceding data declarations, shows how to use the
LENGTHOF, SIZEOF, and TYPE operators with structures.
INFO STRUCT
buffer BYTE 100 DUP (?)
crlf BYTE 13, 10
query BYTE 'Filename: '
endmark BYTE 36
drives DISKDRIVES <0, 1, 1>
INFO ENDS
info1 INFO { , , 'Dir' }

lotsof INFO { , , 'file1', , {0,0,0} },
{ , , 'file2', , {0,0,1} },
{ , , 'file3', , {0,0,2} }
sinfo1 EQU SIZEOF info1 ; 116 = number of bytes in

; initializers
linfo1 EQU LENGTHOF info1 ; 1 = number of items
tinfo1 EQU TYPE info1 ; 116 = same as size
slotsof EQU SIZEOF lotsof ; 116 * 3 = number of bytes in

; initializers
llotsof EQU LENGTHOF lotsof ; 3 = number of items
tlotsof EQU TYPE lotsof ; 116 = same as size for structure
; of type INFO
LENGTHOF, SIZEOF, and TYPE for Unions

The size of a union determined by SIZEOF is the size of the longest field plus
any padding required. The length of a union variable determined by
LENGTHOF equals the number of initializers defined inside angle brackets or
curly braces. TYPE returns a value indicating the type of the longest field.
DWB UNION
d DWORD ?
w WORD ?
b BYTE ?
DWB ENDS
num DWB {0FFFFh}

array DWB (100 / SIZEOF DWB) DUP ({0})
snum EQU SIZEOF num ; = 4

lnum EQU LENGTHOF num ; = 1
tnum EQU TYPE num ; = 4
sarray EQU SIZEOF array ; = 100 (4*25)
larray EQU LENGTHOF array ; = 25
tarray EQU TYPE array ; = 4

Referencing Structures, Unions, and Fields

Like other variables, structure variables can be accessed by name. You can
access fields within structure variables with this syntax:
variable. field
References to fields must always be fully qualified, with the structure or union
names and the dot operator preceding the field name. The assembler requires
that you use the dot operator only with structure fields, not as an alternative to
the plus operator; nor can you use the plus operator as an alternative to the dot
operator.
The following example shows several ways to reference the fields of a structure
of type DATE.
DATE STRUCT ; Defines structure type
month BYTE ?
day BYTE ?
year WORD ?
DATE ENDS
yesterday DATE {1, 20, 1993} ; Declare structure

; variable
.
.
.
mov al, yesterday.day ; Use structure variables
mov bx, OFFSET yesterday ; Load structure address
mov al, (DATE PTR [bx]).month ; Use as indirect operand
mov al, [bx].date.month ; This is necessary only if
; month is already a
; field in a different
; structure
Under OPTION M510 or OPTION OLDSTRUCTS, unique structure names

do not need to be qualified. However, if the NONUNIQUE keyword appears in
a structure definition, all fields of the structure must be fully qualified when
referenced, even if the OPTION OLDSTRUCTS directive appears in the code.
Also, you must qualify all references to a field. (For information on the
OPTION directive, see Chapter 1.)
Even if the initialized union is the size of a WORD or DWORD, members of
structures or unions are accessible only through the field’s names.

In the following example, the two MOV statements show how you can access
the elements of an array of unions.
WB UNION
w WORD ?
b BYTE ?
WB ENDS
array WB (100 / SIZEOF WB) DUP ({0})
mov array[12].w, 40h

mov array[32].b, 2
As the preceding code illustrates, you can use unions to access the same data in
more than one form. One application of structures and unions is to simplify the
task of reinitializing a far pointer. For a far pointer declared as
FPWORD TYPEDEF FAR PTR WORD
.DATA
WordPtr FPWORD ?
you must follow these steps to point WordPtr to a word value named
ThisWord in the current data segment.
mov WORD PTR WordPtr[2], ds
mov WORD PTR WordPtr, OFFSET ThisWord
The preceding method requires that you remember whether the segment or the
offset is stored first. However, if your program declares a union like this:
uptr UNION
dwptr FPWORD 0
STRUCT
offs WORD 0
segm WORD 0
ENDS
uptr ENDS

You can initialize a far pointer with these steps:

.DATA
WrdPtr2 uptr <>
.
.
.
mov WrdPtr2.segm, ds
mov WrdPtr2.offs, OFFSET ThisWord
This code moves the segment and the offset into the pointer and then moves the
pointer into a register with the other field of the union. Although this technique
does not reduce the code size, it avoids confusion about the order for loading
the segment and offset.
Nested Structures and Unions

You can nest structures and unions in several ways. This section explains how
to refer to the fields in a nested structure or union. The following example
illustrates the four techniques for nesting, and how to reference the fields. Note
the syntax for nested structures. The techniques are reviewed following the
example.
ITEMS STRUCT
Inum WORD ?
Iname BYTE 'Item Name'
ITEMS ENDS
INVENTORY STRUCT
UpDate WORD ?
oldItem ITEMS { \
100,
'AF8' \ ; Named variable of
} ; existing structure
ITEMS { ?, '94C' } ; Unnamed variable of
; existing type
STRUCT ups ; Named nested structure
source WORD ?
shipmode BYTE ?
ENDS
STRUCT ; Unnamed nested structure
f1 WORD ?
f2 WORD ?
ENDS
INVENTORY ENDS
.DATA
yearly INVENTORY { }

; Referencing each type of data in the yearly structure:
mov ax, yearly.oldItem.Inum

mov yearly.ups.shipmode, 'A'
mov yearly.Inum, 'C'
mov ax, yearly.f1
To nest structures and unions, you can use any of these techniques:
u The field of a structure or union can be a named variable of an existing
structure or union type, as in the oldItem field. Because INVENTORY
contains two structures of type ITEMS , the field names in oldItem are not
unique. Therefore, you must use the full field names when referencing those
fields, as in the statement
mov ax, yearly.oldItem.Inum
u To declare a named structure or union inside another structure or union, give

the STRUCT or UNION keyword first and then define a label for it. Fields
of the nested structure or union must always be qualified:
mov yearly.ups.shipmode, 'A'
u As shown in the Items field of Inventory, you also can use unnamed
variables of existing structures or unions inside another structure or union. In
these cases, you can reference fields directly:
mov yearly.Inum, 'C'
mov ax, yearly.f1
Records
Records are similar to structures, except that fields in records are bit strings.
Each bit field in a record variable can be used separately in constant operands or
expressions. The processor cannot access bits individually at run time, but it can
access bit fields with instructions that manipulate bits.
Records are bytes, words, or doublewords in which the individual bits or groups
of bits are considered fields. In general, the three steps for using record variables
are the same as those for using other complex data types:
1. Declare a record type.
2. Define one or more variables having the record type.
3. Reference record variables using shifts and masks.
Once it is defined, you can use the record variable as an operand in assembler
statements.

This section explains the record declaration syntax and the use of the MASK
and WIDTH operators. It also shows some applications of record variables and
constants.
Declaring Record Types

A record type creates a template for data with the sizes and, optionally, the
initial values for bit fields in the record. It does not allocate memory space for
the
record.
The RECORD directive declares a record type for an 8-bit, 16-bit, or 32-bit
record that contains one or more bit fields. The maximum size is based on the
expression word size. See OPTION EXPR16 and OPTION EXPR32 in
Chapter 1. The syntax is:
recordname RECORD field [[, field]]...
The field declares the name, width, and initial value for the field. The syntax for
each field is:
fieldname:width[[=expression]]
Global labels, macro names, and record field names must all be unique, but
record field names can have the same names as structure field names. Width is
the number of bits in the field, and expression is a constant giving the initial (or
default) value for the field. Record definitions can span more than one line if the
continued lines end with commas.
If expression is given, it declares the initial value for the field. The assembler
generates an error message if an initial value is too large for the width of its field.
The first field in the declaration always goes into the most significant bits of the
record. Subsequent fields are placed to the right in the succeeding bits. If the
fields do not total exactly 8, 16, or 32 bits as appropriate, the entire record is
shifted right, so the last bit of the last field is the lowest bit of the record.
Unused bits in the high end of the record are initialized to 0.
The following example creates a byte record type COLOR having four fields:
blink, back, intense, and fore. The contents of the record type are shown
after the example. Since no initial values are given, all bits are set to 0. Note that
this is only a template maintained by the assembler. It allocates no space in the
data segment.

COLOR RECORD blink:1, back:3, intense:1, fore:3
The next example creates a record type CW that has six fields. Each record
declared with this type occupies 16 bits of memory. Initial (default) values are
given for each field. You can use them when declaring data for the record. The
bit diagram after the example shows the contents of the record type.
CW RECORD r1:3=0, ic:1=0, rc:2=0, pc:2=3, r2:2=1, masks:6=63
Defining Record Variables

Once you have declared a record type, you can define record variables of that
type. For each variable, the assembler allocates memory in the format declared
by the type. The syntax is:
[[name]] recordname <[[initializer [[,initializer]]...]] >
[[name]] recordname { [[initializer [[,initializer]]...]] }
[[name]] recordname constant DUP ( [[initializer [[,initializer]]...]] )
The recordname is the name of a record type previously declared with the
RECORD directive.
A fieldlist for each field in the record can be a list of integers, character
constants, or expressions that correspond to a value compatible with the size of
the field. You must include curly braces or angle brackets even when you do not
specify an initial value.
If you use the DUP operator (see “Declaring and Referencing Arrays,” earlier in
this chapter) to initialize multiple record variables, only the angle brackets and

any initial values need to be enclosed in parentheses. For example, you can
define an array of record variables with
xmas COLOR 50 DUP ( <1, 2, 0, 4> )
You do not have to initialize all fields in a record. If an initial value is blank, the
assembler automatically stores the default initial value of the field. If there is no
default value, the assembler clears each bit in the field.
The definition in the following example creates a variable named warning
whose type is given by the record type COLOR. The initial values of the fields in
the variable are set to the values given in the record definition. The initial values
override any default record values given in the declaration.
COLOR RECORD blink:1,back:3,intense:1,fore:3 ; Record
; declaration
warning COLOR <1, 0, 1, 4> ; Record
; definition
LENGTHOF, SIZEOF, and TYPE with Records

The SIZEOF and TYPE operators applied to a record name return the number
of bytes used by the record. SIZEOF returns the number of bytes a record
variable occupies. You cannot use LENGTHOF with a record declaration, but
you can use it with defined record variables. LENGTHOF returns the number
of records in an array of records, or 1 for a single record variable. The following
example illustrates these points.
; Record definition
; 9 bits stored in 2 bytes
RGBCOLOR RECORD red:3, green:3, blue:3
mov ax, RGBCOLOR ; Equivalent to "mov ax, 01FFh"

; mov ax, LENGTHOF RGBCOLOR ; Illegal since LENGTHOF can
; apply only to data label
mov ax, SIZEOF RGBCOLOR ; Equivalent to "mov ax, 2"
mov ax, TYPE RGBCOLOR ; Equivalent to "mov ax, 2"

; Record instance
; 8 bits stored in 1 byte
RGBCOLOR2 RECORD red:3, green:3, blue:2
rgb RGBCOLOR2 <1, 1, 1> ; Initialize to 00100101y
mov ax, RGBCOLOR2 ; Equivalent to

; "mov ax, 00FFh"
mov ax, LENGTHOF rgb ; Equivalent to "mov ax, 1"
mov ax, SIZEOF rgb ; Equivalent to "mov ax, 1"
mov ax, TYPE rgb ; Equivalent to "mov ax, 1"
Record Operators
The WIDTH operator (used only with records) returns the width in bits of a
record or record field. The MASK operator returns a bit mask for the bit
positions occupied by the given record field. A bit in the mask contains a 1 if
that bit corresponds to a bit field. The following example shows how to use
MASK and WIDTH.
.DATA
COLOR RECORD blink:1, back:3, intense:1, fore:3
message COLOR <1, 5, 1, 1>
wblink EQU WIDTH blink ; "wblink" = 1
wback EQU WIDTH back ; "wback" = 3
wintens EQU WIDTH intense ; "wintens" = 1
wfore EQU WIDTH fore ; "wfore" = 3
wcolor EQU WIDTH COLOR ; "wcolor" = 8
.CODE
.
.
.
mov ah, message ; Load initial 1101 1001
and ah, NOT MASK back ; Turn off AND 1000 1111
; "back" ---------
; 1000 1001
or ah, MASK blink ; Turn on OR 1000 0000
; "blink" ---------
; 1000 1001
xor ah, MASK intense ; Toggle XOR 0000 1000
; "intense" ---------
; 1000 0001
IF (WIDTH COLOR) GT 8 ; If color is 16 bit, load

mov ax, message ; into 16-bit register
ELSE ; else
mov al, message ; load into low 8-bit register
xor ah, ah ; and clear high 8-bits
ENDIF

The example continues by illustrating several ways in which record fields can
serve as operands and expressions:
; Rotate "back" of "message" without changing other values
mov al, message ; Load value from memory

mov ah, al ; Save a copy for work 1101 1001=ah/al
and al, NOT MASK back; Mask out old bits AND 1000 1111=mask
; to save old message ---------
; 1000 1001=al
mov cl, back ; Load bit position
shr ah, cl ; Shift to right 0000 1101=ah
inc ah ; Increment 0000 1110=ah
shl ah, cl ; Shift left again 1110 0000=ah

and ah, MASK back ; Mask off extra bits AND 0111 0000=mask
; to get new message ---------
; 0110 0000 ah
or ah, al ; Combine old and new OR 1000 1001 al
; ---------
mov message, ah ; Write back to memory 1110 1001 ah
Record variables are often used with the logical operators to perform logical
operations on the bit fields of the record, as in the previous example using the
MASK operator.

135
C H A P T E R 6
Using Floating-Point and

Binary Coded Decimal Numbers
MASM requires different techniques for handling floating-point (real) numbers

and binary coded decimal (BCD) numbers than for handling integers. You have
two choices for working with real numbers — a math coprocessor or emulation
routines.
Math coprocessors — the 8087, 80287, and 80387 chips — work with the main
processor to handle real-number calculations. The 80486 processor performs
floating-point operations directly. All information in this chapter pertaining to the
80387 coprocessor applies to the 80486DX processor as well. It does not apply
to the 80486SX, which does not provide an on-chip coprocessor.
This chapter begins with a summary of the directives and formats of floating-
point data that you need to allocate memory storage and initialize variables
before you can work with floating-point numbers.
The chapter then explains how to use a math coprocessor for floating-point
operations. It covers:
u The architecture of the registers.
u The operands for the coprocessor instruction formats.
u The coordination of coprocessor and main processor memory access.
u The basic groups of coprocessor instructions — for loading and storing data,
doing arithmetic calculations, and controlling program flow.
The next main section describes emulation libraries. The emulation routines
provided with all Microsoft high-level languages enable you to use coprocessor
instructions as though your computer had a math coprocessor. However, some
coprocessor instructions are not handled by emulation, as this section explains.
Finally, because math coprocessor and emulation routines can also operate on
BCD numbers, this chapter includes the instruction set for these numbers.

Using Floating-Point Numbers

Before using floating-point data in your program, you need to allocate the
memory storage for the data. You can then initialize variables either as real
numbers in decimal form or as encoded hexadecimals. The assembler stores
allocated data in 10-byte IEEE format. This section covers floating-point
declarations and floating-point data formats.
Declaring Floating-Point Variables and Constants

You can allocate real constants using the REAL4, REAL8, and REAL10
directives. These directives allocate the following floating-point numbers:
Directive Size
REAL4 Short (32-bit) real numbers
REAL8 Long (64-bit) real numbers
REAL10 10-byte (80-bit) real numbers and BCD numbers
Table 6.1 lists the possible ranges for floating-point variables. The number of
significant digits can vary in an arithmetic operation as the least-significant digit
may be lost through rounding errors. This occurs regularly for short and long
real numbers, so you should assume the lesser value of significant digits shown
in Table 6.1. Ten-byte real numbers are much less susceptible to rounding
errors for reasons described in the next section. However, under certain
circumstances, 10-byte real operations can have a precision of only 18 digits.
Table 6.1 Ranges of Floating-Point Variables
Significant
Data Type Bits Digits Approximate Range
Short real 32 6–7 1.18 x 10- 38 to 3.40 x 1038
Long real 64 15–16 2.23 x 10- 308 to 1.79 x 10308
10-byte real 80 19 3.37 x 10- 4932 to 1.18 x 104932
With versions of MASM prior to 6.0, the DD, DQ, and DT directives could
allocate real constants. MASM 6.1 still supports these directives, but the
variables are integers rather than floating-point values. Although this makes no
difference in the assembly code, CodeView displays the values incorrectly.
You can specify floating-point constants either as decimal constants or as
encoded hexadecimal constants. You can express decimal real-number constants
in the form:
[[+ | –]] integer[[fraction]][[E[[+ | –]]exponent]]

Chapter 6 Using Floating-Point and Binary Coded Decimal Numbers 137
For example, the numbers 2.523E1 and -3.6E-2 are written in the correct
decimal format. You can use these numbers as initializers for real-number
variables.
The assembler always evaluates digits of real numbers as base 10. It converts
real-number constants given in decimal format to a binary format. The sign,
exponent, and decimal part of the real number are encoded as bit fields within
the number.
You can also specify the encoded format directly with hexadecimal digits (0–9
plus A–F). The number must begin with a decimal digit (0–9) and end with the
real-number designator (R). It cannot be signed. For example, the hexadecimal
number 3F800000r can serve as an initializer for a doubleword-sized variable.
The maximum range of exponent values and the number of digits required in the
hexadecimal number depend on the directive. The number of digits for encoded
numbers used with REAL4, REAL8, and REAL10 must be 8, 16, and 20
digits, respectively. If the number has a leading zero, the number must be 9, 17,
or 21 digits.
Examples of decimal constant and hexadecimal specifications are shown here:
; Real numbers
short REAL4 25.23 ; IEEE format
double REAL8 2.523E1 ; IEEE format
tenbyte REAL10 2523.0E-2 ; 10-byte real format
; Encoded as hexadecimals
ieeeshort REAL4 3F800000r ; 1.0 as IEEE short
ieeedouble REAL8 3FF0000000000000r ; 1.0 as IEEE long
temporary REAL10 3FFF8000000000000000r ; 1.0 as 10-byte
; real
The section “Storing Numbers in Floating-Point Format,” following, explains the

IEEE formats — the way the assembler actually stores the data.
Pascal or C programmers may prefer to create language-specific TYPEDEF
declarations, as illustrated in this example:

; C-language specific
float TYPEDEF REAL4
double TYPEDEF REAL8
long_double TYPEDEF REAL10
; Pascal-language specific
SINGLE TYPEDEF REAL4
DOUBLE TYPEDEF REAL8
EXTENDED TYPEDEF REAL10
For applications of TYPEDEF, see “Defining Pointer Types with TYPEDEF,”

page 75.
Storing Numbers in Floating-Point Format

The assembler stores floating-point variables in the IEEE format. MASM 6.1
does not support .MSFLOAT and Microsoft binary format, which are available
in version 5.1 and earlier. Figure 6.1 illustrates the IEEE format for encoding
short (4-byte), long (8-byte), and 10-byte real numbers. Although this figure
places the most significant bit first for illustration, low bytes actually appear first
in memory.

Figure 6.1 Encoding for Real Numbers in IEEE Format
The following list explains how the parts of a real number are stored in the IEEE
format. Each item in the list refers to an item in Figure 6.1.
u Sign bit (0 for positive or 1 for negative) in the upper bit of the first byte.
u Exponent in the next bits in sequence (8 bits for a short real number, 11 bits
for a long real number, and 15 bits for a 10-byte real number).

u The integer part of the significand in bit 63 for the 10-byte real format. By
absorbing carry values, this bit allows 10-byte real operations to preserve
precision to 19 digits. The integer part is always 1 in short and long real
numbers; consequently, these formats do not provide a bit for the integer,
since there is no point in storing it.
u Decimal part of the significand in the remaining bits. The length is 23 bits for
short real numbers, 52 bits for long real numbers, and 63 bits for 10-byte real
numbers.
The exponent field represents a multiplier 2n. To accommodate negative

exponents (such as 2-6), the value in the exponent field is biased; that is, the
actual exponent is determined by subtracting the appropriate bias value from the
value in the exponent field. For example, the bias for short real numbers is 127.
If the value in the exponent field is 130, the exponent represents a value of 2130-
127, or 23. The bias for long real numbers is 1,023. The bias for 10-byte real
numbers is 16,383.
Once you have declared floating-point data for your program, you can use
coprocessor or emulator instructions to access the data. The next section
focuses on the coprocessor architecture, instructions, and operands required for
floating-point operations.
Using a Math Coprocessor

When used with real numbers, packed BCD numbers, or long integers,
coprocessors (the 8087, 80287, 80387, and 80486) calculate many times faster
than the 8086-based processors. The coprocessor handles data with its own
registers. The organization of these registers can be one of the four formats for
using operands explained in “Instruction and Operand Formats,” later in this
section.
This section describes how the coprocessor transfers data to and from the
coprocessor, coordinates processor and coprocessor operations, and controls
program flow.

Coprocessor Architecture
The coprocessor accesses memory as the CPU does, but it has its own data and
control registers — eight data registers organized as a stack and seven control
registers similar to the 8086 flag registers. The coprocessor’s instruction set
provides direct access to these registers.
The eight 80-bit data registers of the 8087-based coprocessors are organized as
a stack, although they need not be used as a stack. As data items are pushed
into the top register, previous data items move into higher-numbered registers,
which are lower on the stack. Register 0 is the top of the stack; register 7 is the
bottom. The syntax for specifying registers is:
ST [[(number)]]
The number must be a digit between 0 and 7 or a constant expression that
evaluates to a number from 0 to 7. ST is another way to refer to ST(0).
All coprocessor data is stored in registers in the 10-byte real format. The
registers and the register format are shown in Figure 6.2.
Figure 6.2 Coprocessor Data Registers
Internally, all calculations are done on numbers of the same type. Since 10-byte
real numbers have the greatest precision, lower-precision numbers are
guaranteed not to lose precision as a result of calculations. The instructions that
transfer values between the main memory and the coprocessor automatically
convert numbers to and from the 10-byte real format.

Instruction and Operand Formats

Because of the stack organization of registers, you can consider registers either
as elements on a stack or as registers much like 8086-family registers. Table 6.2
lists the four main groups of coprocessor instructions and the general syntax for
each. The names given to the instruction format reflect the way the instruction
uses the coprocessor registers. The instruction operands are placed in the
coprocessor data registers before the instruction executes.
Table 6.2 Coprocessor Operand Formats
Instruction Implied
Format Syntax Operands Example
Classical stack Finstruction ST, ST(1) fadd
Memory Finstruction memory ST fadd memloc
Register Finstruction ST(num), ST — fadd st(5), st
fadd st, st(3)
Finstruction ST, ST(num)
Register pop FinstructionP ST(num), ST — faddp st(4), st
You can easily recognize coprocessor instructions because, unlike all 8086-
family instruction mnemonics, they start with the letter F. Coprocessor
instructions can never have immediate operands and, with the exception of the
FSTSW instruction, they cannot have processor registers as operands.
Classical-Stack Format
Instructions in the classical-stack format treat the coprocessor registers like items
on a stack — thus its name. Items are pushed onto or popped off the top
elements of the stack. Since only the top item can be accessed on a traditional
stack, there is no need to specify operands. The first (top) register (and the
second, if the instruction needs two operands) is always assumed.
ST (the top of the stack) is the source operand in coprocessor arithmetic
operations. ST(1), the second register, is the destination. The result of the
operation replaces the destination operand, and the source is popped off the
stack. This leaves the result at the top of the stack.

The following example illustrates the classical-stack format; Figure 6.3 shows
the status of the register stack after each instruction.
fld1 ; Push 1 into first position
fldpi ; Push pi into first position
fadd ; Add pi and 1 and pop
Figure 6.3 Status of the Register Stack
Memory Format
Instructions that use the memory format, such as data transfer instructions, also
treat coprocessor registers like items on a stack. However, with this format,
items are pushed from memory onto the top element of the stack, or popped
from the top element to memory. You must specify the memory operand.
Some instructions that use the memory format specify how a memory operand
is to be interpreted — as an integer (I) or as a binary coded decimal (B). The
letter I or B follows the initial F in the syntax. For example, FILD interprets its
operand as an integer and FBLD interprets its operand as a BCD number. If the
instruction name does not include a type letter, the instruction works on real
numbers.
You can also use memory operands in calculation instructions that operate on
two values (see “Using Coprocessor Instructions,” later in this section). The
memory operand is always the source. The stack top (ST) is always the implied
destination.

The result of the operation replaces the destination without changing its stack
position, as shown in this example and in Figure 6.4:
.DATA
m1 REAL4 1.0
m2 REAL4 2.0
.CODE
.
.
.
fld m1 ; Push m1 into first position
fld m2 ; Push m2 into first position
fadd m1 ; Add m2 to first position
fstp m1 ; Pop first position into m1
fst m2 ; Copy first position to m2
Figure 6.4 Status of the Register Stack and Memory Locations
Register Format
Instructions that use the register format treat coprocessor registers as registers
rather than as stack elements. Instructions that use this format require two
register operands; one of them must be the stack top (ST).
In the register format, specify all operands by name. The first operand is the
destination; its value is replaced with the result of the operation. The second
operand is the source; it is not affected by the operation. The stack positions of
the operands do not change.

The only instructions that use the register operand format are the FXCH
instruction and arithmetic instructions for calculations on two values. With the
FXCH instruction, the stack top is implied and need not be specified, as shown
in this example and in Figure 6.5:
fadd st(1), st ; Add second position to first -
; result goes in second position
fadd st, st(2) ; Add first position to third -
; result goes in first position
fxch st(1) ; Exchange first and second positions
Figure 6.5 Status of the Previously Initialized Register Stack
Register-Pop Format
The register-pop format treats coprocessor registers as a modified stack. The
source register must always be the stack top. Specify the destination with the
register’s name.
Instructions with this format place the result of the operation into the destination
operand, and the top pops off the stack. The register-pop format is used only
for instructions for calculations on two values, as in this example and in Figure
6.6:
faddp st(2), st ; Add first and third positions and pop -
; first position destroyed;
; third moves to second and holds result

Figure 6.6 Status of the Already Initialized Register Stack
Coordinating Memory Access

The math coprocessor and main processor work simultaneously. However,
since the coprocessor cannot handle device input or output, data originates in the
main processor.
The main processor and the coprocessor have their own registers, which are
separate and inaccessible to each other. They exchange data through memory,
since memory is available to both.
When using the coprocessor, follow these three steps:
1. Load data from memory to coprocessor registers.
2. Process the data.
3. Store the data from coprocessor registers back to memory.
Step 2, processing the data, can occur while the main processor is handling other
tasks. Steps 1 and 3 must be coordinated with the main processor so that the
processor and coprocessor do not try to access the same memory at the same
time; otherwise, problems of coordinating memory access can occur. Since the
processor and coprocessor work independently, they may not finish working on
memory in the order in which you give instructions. The two potential timing
conflicts that can occur are handled in different ways.
One timing conflict results from a coprocessor instruction following a processor
instruction. The processor may have to wait until the coprocessor finishes if the
next processor instruction requires the result of the coprocessor’s calculation.
You do not have to write your code to avoid this conflict, however. The
assembler coordinates this timing automatically for the 8088 and 8086
processors, and the processor coordinates it automatically on the 80186–80486
processors. This is the case shown in the first example that follows.
Another conflict results from a processor instruction that accesses memory
following a coprocessor instruction that accesses the same memory. The
processor can try to load a variable that is still being used by the coprocessor.
You need careful synchronization to control the timing, and this synchronization
is not automatic on the 8087 coprocessor. For code to run correctly on the
8087, you must include WAIT or FWAIT (mnemonics for the same instruction)
to ensure that the coprocessor finishes before the processor begins, as shown in
the second example.

In this situation, the processor does not generate the FWAIT instruction
automatically.
; Processor instruction first - No wait needed
mov WORD PTR mem32[0], ax ; Load memory
mov WORD PTR mem32[2], dx
fild mem32 ; Load to register
; Coprocessor instruction first - Wait needed (for 8087)

fist mem32 ; Store to memory
fwait ; Wait until coprocessor
; is done
mov ax, WORD PTR mem32[0] ; Move to register
mov dx, WORD PTR mem32[2]
When generating code for the 8087 coprocessor, the assembler automatically
inserts a WAIT instruction before the coprocessor instruction. However, if you
use the .286 or .386 directive, the compiler assumes that the coprocessor
instructions are for the 80287 or 80387 and does not insert the WAIT
instruction. If your code does not need to run on an 8086 or 8088 processor,
you can make your programs smaller and more efficient by using the .286 or
.386 directive.
Using Coprocessor Instructions

The 8087 family of coprocessors has separate instructions for each of the
following operations:
u Loading and storing data
u Doing arithmetic calculations
u Controlling program flow
The following sections explain the available instructions and show how to use
them for each of these operations. For general syntax information, see
“Instruction and Operand Formats,” earlier in this section.
Loading and Storing Data

Data-transfer instructions copy data between main memory and the coprocessor
registers or between different coprocessor registers. Two principles govern data
transfers:
u The choice of instruction determines whether a value in memory is
considered an integer, a BCD number, or a real number. The value is always
considered a 10-byte real number once transferred to the coprocessor.

u The size of the operand determines the size of a value in memory. Values in
the coprocessor always take up 10 bytes.
You can transfer data to stack registers using load commands. These commands
push data onto the stack from memory or from coprocessor registers. Store
commands remove data. Some store commands pop data off the register stack
into memory or coprocessor registers; others simply copy the data without
changing it on the stack.
If you use constants as operands, you cannot load them directly into
coprocessor registers. You must allocate memory and initialize a variable to a
constant value. That variable can then be loaded by using one of the load
instructions in the following list.
The math coprocessor offers a few special instructions for loading certain
constants. You can load 0, 1, pi, and several common logarithmic values
directly. Using these instructions is faster and often more precise than loading
the values from initialized variables.
All instructions that load constants have the stack top as the implied destination
operand. The constant to be loaded is the implied source operand.
The coprocessor data area, or parts of it, can also be moved to memory and
later loaded back. You may want to do this to save the current state of the
coprocessor before executing a procedure. After the procedure ends, restore the
previous status. Saving coprocessor data is also useful when you want to modify
coprocessor behavior by writing certain data to main memory, operating on the
data with 8086-family instructions, and then loading it back to the coprocessor
data area.
Use the following instructions for transferring numbers to and from
registers:
Instruction(s) Description
FLD, FST, FSTP Loads and stores real numbers
FILD, FIST, FISTP Loads and stores binary integers
FBLD Loads BCD
FBSTP Stores BCD
FXCH Exchanges register values
FLDZ Pushes 0 into ST
FLD1 Pushes 1 into ST
FLDPI Pushes the value of pi into ST
FLDCW mem2byte Loads the control word into the coprocessor
F[[N]]STCW mem2byte Stores the control word in memory

FLDENV mem14byte Loads environment from memory

F[[N]]STENV mem14byte Stores environment in memory
Instruction(s) Description
FRSTOR mem94byte Restores state from memory
F[[N]]SAVE mem94byte Saves state in memory
FLDL2E Pushes the value of log2e into ST
FLDL2T Pushes log210 into ST
FLDLG2 Pushes log102 into ST
FLDLN2 Pushes loge2 into ST
The following example and Figure 6.7 illustrate some of these instructions:
.DATA
m1 REAL4 1.0
m2 REAL4 2.0
.CODE
fld m1 ; Push m1 into first item
fld st(2) ; Push third item into first
fst m2 ; Copy first item to m2
fxch st(2) ; Exchange first and third items
fstp m1 ; Pop first item into m1
Figure 6.7 Status of the Register Stack: Main Memory and Coprocessor

Doing Arithmetic Calculations

Most of the coprocessor instructions for arithmetic operations have several
forms, depending on the operand used. You do not need to specify the operand
type in the

instruction if both operands are stack registers, since register values are always
10-byte real numbers. In most of the arithmetic instructions listed here, the
result replaces the destination register. The instructions include:
FADD Adds the source and destination
FSUB Subtracts the source from the destination
FSUBR Subtracts the destination from the source
FMUL Multiplies the source and the destination
FDIV Divides the destination by the source
FDIVR Divides the source by the destination
FABS Sets the sign of ST to positive
FCHS Reverses the sign of ST
FRNDINT Rounds ST to an integer
FSQRT Replaces the contents of ST with its square root
FSCALE Multiplies the stack-top value by 2 to the power contained in ST(1)
FPREM Calculates the remainder of ST divided by ST(1)
80387 Only
FSIN Calculates the sine of the value in ST
FCOS Calculates the cosine of the value in ST
FSINCOS Calculates the sine and cosine of the value in ST
FPREM1 Calculates the partial remainder by performing modulo division on the top
two stack registers
FXTRACT Breaks a number down into its exponent and mantissa and pushes the
mantissa onto the register stack
F2XM1 Calculates 2x–1
FYL2X Calculates Y * log2 X
FYL2XP1 Calculates Y * log2 (X+1)
FPTAN Calculates the tangent of the value in ST
FPATAN Calculates the arctangent of the ratio Y/X
F[[N]]INIT Resets the coprocessor and restores all the default conditions in the control
and status words
F[[N]]CLEX Clears all exception flags and the busy flag of the status word
FINCSTP Adds 1 to the stack pointer in the status word
FDECSTP Subtracts 1 from the stack pointer in the status word
FFREE Marks the specified register as empty

The following example illustrates several arithmetic instructions. The code

solves quadratic equations, but does no error checking and fails for some values
because it attempts to find the square root of a negative number. Both Help and
the MATH.ASM sample file show a complete version of this procedure. The
complete form uses the FTST (Test for Zero) instruction to check for a negative
number or 0 before calculating the square root.
.DATA
a REAL4 3.0
b REAL4 7.0
cc REAL4 2.0
posx REAL4 0.0
negx REAL4 0.0
.CODE
.
.
.
; Solve quadratic equation - no error checking
; The formula is: -b +/- squareroot(b2 - 4ac) / (2a)
fld1 ; Get constants 2 and 4
fadd st,st ; 2 at bottom
fld st ; Copy it
fmul a ; = 2a
fmul st(1),st ; = 4a
fxch ; Exchange
fmul cc ; = 4ac
fld b ; Load b
fmul st,st ; = b2
fsubr ; = b2 - 4ac
; Negative value here produces error
fsqrt ; = square root(b2 - 4ac)
fld b ; Load b
fchs ; Make it negative
fxch ; Exchange
fld st ; Copy square root

fadd st,st(2) ; Plus version = -b + root(b2 - 4ac)
fxch ; Exchange
fsubp st(2),st ; Minus version = -b - root(b2 - 4ac)
fdiv st,st(2) ; Divide plus version

fstp posx ; Store it
fdivr ; Divide minus version
fstp negx ; Store it

Controlling Program Flow

The math coprocessor has several instructions that set control flags in the status
word. The 8087-family control flags can be used with conditional jumps to
direct program flow in the same way that 8086-family flags are used. Since the
coprocessor does not have jump instructions, you must transfer the status word
to memory so that the flags can be used by 8086-family instructions.
An easy way to use the status word with conditional jumps is to move its upper
byte into the lower byte of the processor flags, as shown in this example:
fstsw mem16 ; Store status word in memory
fwait ; Make sure coprocessor is done
mov ax, mem16 ; Move to AX
sahf ; Store upper word in flags
The SAHF (Store AH into Flags) instruction in this example transfers AH into
the low bits of the flags register.
You can save several steps by loading the status word directly to AX on the
80287 with the FSTSW and FNSTSW instructions. This is the only case in
which data can be transferred directly between processor and coprocessor
registers, as shown in this example:
fstsw ax
The coprocessor control flags and their relationship to the status word are
described in “Control Registers,” following.
The 8087-family coprocessors provide several instructions for comparing
operands and testing control flags. All these instructions compare the stack top
(ST) to a source operand, which may either be specified or implied as ST(1).
The compare instructions affect the C3, C2, and C0 control flags, but not the
C1 flag. Table 6.3 shows the flags’ settings for each possible result of a
comparison or test.
Table 6.3 Control-Flag Settings After Comparison or Test
After FCOM After FTEST C3 C2 C0
ST > source ST is positive 0 0 0
ST < source ST is negative 0 0 1
ST = source ST is 0 1 0 0
Not comparable ST is NAN or projective infinity 1 1 1
Variations on the compare instructions allow you to pop the stack once or twice
and to compare integers and zero. For each instruction, the stack top is always

the implied destination operand. If you do not give an operand, ST(1) is the
implied source. With some compare instructions, you can specify the source as
a memory or register operand.
All instructions summarized in the following list have implied operands: either
ST as a single-destination operand or ST as the destination and ST(1) as the
source. Each instruction in the list has implied operands. Some instructions have
a wait version and a no-wait version. The no-wait versions have N as the
second letter. The instructions for comparing and testing flags include:
FCOM Compares the stack top to the source. The
source and destination are unaffected by the comparison.
FTST Compares ST to 0.
FCOMP Compares the stack top to the source and then pops the
stack.
FUCOM , FUCOMP, Compares the source to ST and sets the condition codes of
FUCOMPP the status word according to the result (80386/486 only).
F[[N]]STSW mem2byte Stores the status word in memory.
FXAM Sets the value of the control flags based on the type of the
number in ST.
FPREM Finds a correct remainder for large operands. It uses the
C2 flag to indicate whether the remainder returned is partial
(C2 is set) or complete (C2 is clear). If the bit is set, the
operation should be repeated. It also returns the least-
significant three bits of the quotient in C0, C3, and C1.
FNOP Copies the stack top onto itself, thus padding the
executable file and taking up processing time without having
any effect on registers or memory.
FDISI , FNDISI , FENI, FNENI Enables or disables interrupts (8087 only).
FSETPM Sets protected mode. Requires a .286P or .386P directive
(80287, 80387, and 80486 only).
The following example illustrates some of these instructions. Notice how

conditional blocks are used to enhance 80287 code.
.DATA
down REAL4 10.35 ; Sides of a rectangle
across REAL4 13.07
diamtr REAL4 12.93 ; Diameter of a circle
status WORD ?

P287 EQU (@Cpu AND 00111y)

.CODE
.
.
.
; Get area of rectangle
fld across ; Load one side
fmul down ; Multiply by the other
; Get area of circle: Area = PI * (D/2)2

fld1 ; Load one and
fadd st, st ; double it to get constant 2
fdivr diamtr ; Divide diameter to get radius
fmul st, st ; Square radius
fldpi ; Load pi
fmul ; Multiply it
; Compare area of circle and rectangle

fcompp ; Compare and throw both away
IF p287
fstsw ax ; (For 287+, skip memory)
ELSE
fnstsw status ; Load from coprocessor to memory
mov ax, status ; Transfer memory to register
ENDIF
sahf ; Transfer AH to flags register
jp nocomp ; If parity set, can't compare
jz same ; If zero set, they're the same
jc rectangle ; If carry set, rectangle is bigger
jmp circle ; else circle is bigger
nocomp: ; Error handler

.
.
.
same: ; Both equal
.
.
.
rectangle: ; Rectangle bigger
.
.
.
circle: ; Circle bigger

Additional instructions for the 80387/486 are FLDENVD and FLDENVW for
loading the environment; FNSTENVD, FNSTENVW, FSTENVD, and
FSTENVW for storing the environment state; FNSAVED, FNSAVEW,
FSAVED, and FSAVEW for saving the coprocessor state; and FRSTORD and
FRSTORW for restoring the coprocessor state.
The size of the code segment, not the operand size, determines the number of
bytes loaded or stored with these instructions. The instructions ending with W
store the 16-bit form of the control register data, and the instructions ending
with D store the 32-bit form. For example, in 16-bit mode FSAVEW saves the
16-bit control register data. If you need to store the 32-bit form of the control
register data, use FSAVED.
Control Registers
Some of the flags of the seven 16-bit control registers control coprocessor
operations, while others maintain the current status of the coprocessor. In this
sense, they are much like the 8086-family flags registers (see Figure 6.8).
Figure 6.8 Coprocessor Control Registers
The status word register is the only commonly used control register. (The others
are used mostly by systems programmers.) The format of the status word
register is shown in Figure 6.9, which shows how the coprocessor control flags
align with the processor flags. C3 overwrites the zero flag, C2 overwrites the
parity flag, and C0 overwrites the carry flag. C1 overwrites an undefined bit, so
it cannot be used directly with conditional jumps, although you can use the
TEST instruction to

check C1 in memory or in a register. The status word register also overwrites

the sign and auxiliary-carry flags, so you cannot count on their being unchanged
after the operation.
Figure 6.9 Coprocessor and Processor Control Flags
Using An Emulator Library

If you do not have a math coprocessor or an 80486 processor, you can do most
floating-point operations by writing assembly-language procedures and accessing
an emulator from a high-level language. All Microsoft high-level languages come
with emulator libraries for all memory models.
To use emulator functions, first write your assembly-language procedure using
coprocessor instructions. Then assemble the module with the /FPi option and
link it with your high-level – language modules. You can enter options in the
Programmer’s WorkBench (PWB) environment, or you can use the OPTION
EMULATOR in your source code.
In emulation mode, the assembler generates instructions for the linker that the
Microsoft emulator can use. The form of the OPTION directive in the following
example tells the assembler to use emulation mode. This option (introduced in
Chapter 1) can be defined only once in a module.
OPTION EMULATOR

You can use emulator functions in a stand-alone assembler program by

assembling with the /Cx command-line option and linking with the appropriate
emulator library. The following fragment outlines a small-model program that
contains floating-point instructions served by an emulator:
.MODEL small, c
OPTION EMULATOR
.
.
.
PUBLIC main
.CODE
main: ; Program entry point must
.STARTUP ; have name 'main'
.
fadd st, st ; Floating-point instructions
fldpi ; emulated
Emulator libraries do not allow for all of the coprocessor instructions. The
following floating-point instructions are not emulated:
FBLD FLDENV FSAVE FSTENV
FBSTP FNOP FSAVEW FUCOM
FCOS FPREM1 FSAVED FUCOMP
FDECSTP FRSTOR FSETPM FUCOMPP
FINCSTP FRSTORW FSIN FXTRACT
FINIT FRSTORD FSINCOS
For information about writing assembly-language procedures for high-level
languages, see Chapter 12, “Mixed-Language Programming.”
Using Binary Coded Decimal Numbers

Binary coded decimal (BCD) numbers allow calculations on large numbers
without rounding errors. This characteristic makes BCD numbers a common
choice for monetary calculations. Although BCDs can represent integers of any
precision, the 8087-based coprocessors accommodate BCD numbers only in the
range ±999,999,999,999,999,999.
This section explains how to define BCD numbers, how to access them with a
math coprocessor or emulator, and how to perform simple BCD calculations on
the main processor.
Defining BCD Constants and Variables

Unpacked BCD numbers are made up of bytes containing a single decimal digit
in the lower 4 bits of each byte. Packed BCD numbers are made up of bytes

containing two decimal digits: one in the upper 4 bits and one in the lower 4 bits.
The leftmost digit holds the sign (0 for positive, 1 for negative).
Packed BCD numbers are encoded in the 8087 coprocessor’s packed BCD
format. They can be up to 18 digits long, packed two digits per byte. The
assembler zero-pads BCDs initialized with fewer than 18 digits. Digit 20 is the
sign bit, and digit 19 is reserved.
When you define an integer constant with the TBYTE directive and the current
radix is decimal (t), the assembler interprets the number as a packed BCD
number.
The syntax for specifying packed BCDs is the same as for other integers.
pos1 TBYTE 1234567890 ; Encoded as 00000000001234567890h
neg1 TBYTE -1234567890 ; Encoded as 80000000001234567890h
Unpacked BCD numbers are stored one digit to a byte, with the value in the
lower
4 bits. They can be defined using the BYTE directive. For example, an
unpacked BCD number could be defined and initialized as follows:
unpackedr BYTE 1,5,8,2,5,2,9 ; Initialized to 9,252,851
unpackedf BYTE 9,2,5,2,8,5,1 ; Initialized to 9,252,851
As these two lines show, you can arrange digits backward or forward, depending
on how you write the calculation routines that handle the numbers.
BCD Calculations on a Coprocessor

As the previous section explains, BCDs differ from other numbers only in the
way a program stores them in memory. Internally, a math coprocessor does not
distinguish BCD integers from any other type. The coprocessor can load,
calculate, and store packed BCD integers up to 18 digits long.
The coprocessor instruction
fbld bcd1
pushes the packed BCD number at bcd1 onto the coprocessor stack. When
your code completes calculations on the number, place the result back into
memory in BCD format with the instruction
fbstp bcd1
which discards the variable from the stack top.

BCD Calculations on the Main Processor

The 8086-family of processors can perform simple arithmetic operations on
BCD integers, but only one digit at a time. The main processor, like the
coprocessor, operates internally on the number’s binary value. It requires
additional code to translate the binary result back into BCD format.
The main processor provides instructions specifically designed to translate to
and from BCD format. These instructions are called “ASCII-adjust” and
“decimal-adjust” instructions. They get their names from Intel mnemonics that
use the term “ASCII” to refer to unpacked BCD numbers and “decimal” to
refer to packed BCD numbers.
Unpacked BCD Numbers

When a calculation using two one-digit values produces a two-digit result, the
instructions AAA, AAS, AAM, and AAD place the first digit in AL and the
second in AH. If the digit in AL needs to carry to or borrow from the digit in
AH, the instructions set the carry and auxiliary carry flags. The four ASCII-
adjust instructions for unpacked BCDs are:
AAA Adjusts after an addition operation.
AAS Adjusts after a subtraction operation.
AAM Adjusts after a multiplication operation. Always use with MUL, not with
IMUL.
AAD Adjusts before a division operation. Unlike other BCD instructions, AAD
converts a BCD value to a binary value before the operation. After the
operation, use AAM to adjust the quotient. The remainder is lost. If you
need the remainder, save it in another register before adjusting the quotient.
Then move it back to AL and adjust if necessary.
For processor arithmetic on unpacked BCD numbers, you must do the 8-bit
arithmetic calculations on each digit separately, and assign the result to the AL
register. After each operation, use the corresponding BCD instruction to adjust
the result. The ASCII-adjust instructions do not take an operand and always
work on the value in the AL register.

The following examples show how to use each of these instructions in BCD
addition, subtraction, multiplication, and division.
; To add 9 and 3 as BCDs:
mov ax, 9 ; Load 9
mov bx, 3 ; and 3 as unpacked BCDs
add al, bl ; Add 09h and 03h to get 0Ch
aaa ; Adjust 0Ch in AL to 02h,
; increment AH to 01h, set carry
; Result 12 (unpacked BCD in AX)
; To subtract 4 from 13:

mov ax, 103h ; Load 13
mov bx, 4 ; and 4 as unpacked BCDs
sub al, bl ; Subtract 4 from 3 to get FFh (-1)
aas ; Adjust 0FFh in AL to 9,
; decrement AH to 0, set carry
; Result 9 (unpacked BCD in AX)
; To multiply 9 times 3:
mov ax, 903h ; Load 9 and 3 as unpacked BCDs
mul ah ; Multiply 9 and 3 to get 1Bh
aam ; Adjust 1Bh in AL
; to get 27 (unpacked BCD in AX)
; To divide 25 by 2:
mov ax, 205h ; Load 25
mov bl, 2 ; and 2 as unpacked BCDs
aad ; Adjust 0205h in AX
; to get 19h in AX
div bl ; Divide by 2 to get
; quotient 0Ch in AL
; remainder 1 in AH
aam ; Adjust 0Ch in AL
; to 12 (unpacked BCD in AX)
; (remainder destroyed)
If you process multidigit BCD numbers in loops, each digit is processed and
adjusted in turn.
Packed BCD Numbers

Packed BCD numbers are made up of bytes containing two decimal digits: one
in the upper 4 bits and one in the lower 4 bits. The 8086-family processors
provide instructions for adjusting packed BCD numbers after addition and
subtraction. You must write your own routines to adjust for multiplication and
division.

For processor calculations on packed BCD numbers, you must do the 8-bit
arithmetic calculations on each byte separately, placing the result in the AL
register. After each operation, use the corresponding decimal-adjust instruction
to adjust the result. The decimal-adjust instructions do not take an operand and
always work on the value in the AL register.
The 8086-family processors provide the instructions DAA (Decimal Adjust after
Addition) and DAS (Decimal Adjust after Subtraction) for adjusting packed
BCD numbers after addition and subtraction.
These examples use DAA and DAS to add and subtract BCDs.
;To add 88 and 33:
mov ax, 8833h ; Load 88 and 33 as packed BCDs
add al, ah ; Add 88 and 33 to get 0BBh
daa ; Adjust 0BBh to 121 (packed BCD:)
; 1 in carry and 21 in AL
;To subtract 38 from 83:

mov ax, 3883h ; Load 83 and 38 as packed BCDs
sub al, ah ; Subtract 38 from 83 to get 04Bh
das ; Adjust 04Bh to 45 (packed BCD:)
; 0 in carry and 45 in AL
Unlike the ASCII-adjust instructions, the decimal-adjust instructions never affect

AH. The assembler sets the auxiliary carry flag if the digit in the lower 4 bits
carries to or borrows from the digit in the upper 4 bits, and it sets the carry flag
if the digit in the upper 4 bits needs to carry to or borrow from another byte.
Multidigit BCD numbers are usually processed in loops. Each byte is processed
and adjusted in turn.

161
C H A P T E R 7
Controlling Program Flow
Very few programs execute all lines sequentially from .STARTUP to .EXIT.
Rather, complex program logic and efficiency dictate that you control the flow
of your program — jumping from one point to another, repeating an action until
a condition is reached, and passing control to and from procedures. This chapter
describes various ways for controlling program flow and several features that
simplify coding program-control constructs.
The first section covers jumps from one point in the program to another. It
explains how MASM 6.1 optimizes both unconditional and conditional jumps
under certain circumstances, so that you do not have to specify every attribute.
The section also describes instructions you can use to test conditional jumps.
The next section describes loop structures that repeat actions or evaluate
conditions. It discusses MASM directives, such as .WHILE and .REPEAT, that
generate appropriate compare, loop, and jump instructions for you, and the .IF,
.ELSE, and .ELSEIF directives that generate jump instructions.
The “Procedures” section in this chapter explains how to write an assembly-
language procedure. It covers the extended functionality for PROC, a PROTO
directive that lets you write procedure prototypes similar to those used in C, an
INVOKE directive that automates parameter passing, and options for the stack-
frame setup inside procedures.
The last section explains how to pass program control to an interrupt routine.
Jumps
Jumps are the most direct way to change program control from one location to
another. At the processor level, jumps work by changing the value of the IP
(Instruction Pointer) register to a target offset and, for far jumps, by changing
the CS register to a new segment address. Jump instructions fall into only two
categories: conditional and unconditional.

Unconditional Jumps
The JMP instruction transfers control unconditionally to another instruction.
JMP’s single operand contains the address of the target instruction.
Unconditional jumps skip over code that should not be executed, as shown here:
; Handle one case
label1: .
.
.
jmp continue
; Handle second case

label2: .
.
.
jmp continue
.
.
.
continue:
The distance of the target from the jump instruction and the size of the operand
determine the assembler’s encoding of the instruction. The longer the distance,
the more bytes the assembler uses to code the instruction. In versions of MASM
prior to 6.0, unconditional NEAR jumps sometimes generated inefficient code,
but MASM can now optimize unconditional jumps.
Jump Optimizing
The assembler determines the smallest encoding possible for the direct
unconditional jump. MASM does not require a distance operator, so you do not
have to determine the correct distance of the jump. If you specify a distance, it
overrides any assembler optimization. If the specified distance falls short of the
target address, the assembler generates an error. If the specified distance is
longer than the jump requires, the assembler encodes the given distance and
does not optimize it.
The assembler optimizes jumps when the following conditions are met:
u You do not specify SHORT, NEAR, FAR, NEAR16, NEAR32, FAR16,
FAR32, or PROC as the distance of the target.
u The target of the jump is not external and is in the same segment as the jump
instruction. If the target is in a different segment (but in the same group), it is
treated as though it were external.

Chapter 7 Controlling Program Flow 163
If these two conditions are met, MASM uses the instruction, distance, and size
of the operand to determine how to optimize the encoding for the jump. No
syntax changes are necessary.
Note This information about jump optimizing also applies to conditional jumps
on the 80386/486.
Indirect Operands
An indirect operand provides a pointer to the target address, rather than the
address itself. A pointer is a variable that contains an address. The processor
distinguishes indirect (pointer) operands from direct (address) operands by the
instruction’s context.
You can specify the pointer’s size with the WORD, DWORD, or FWORD
attributes. Default sizes are based on .MODEL and the default segment size.
jmp [bx] ; Uses .MODEL and segment size defaults
jmp WORD PTR [bx] ; A NEAR16 indirect call
If the indirect operand is a register, the jump is always a NEAR16 jump for a
16-bit register, and NEAR32 for a 32-bit register:
jmp bx ; NEAR16 jump
jmp ebx ; NEAR32 jump
A DWORD indirect operand, however, is ambiguous to the assembler.

jmp DWORD PTR [var] ; A NEAR32 jump in a 32-bit segment;
; a FAR16 jump in a 16-bit segment
In this case, your code must clear the ambiguity with the NEAR32 or FAR16
keywords. The following example shows how to use TYPEDEF to define
NEAR32 and FAR16 pointer types.
NFP TYPEDEF PTR NEAR32
FFP TYPEDEF PTR FAR16
jmp NFP PTR [var] ; NEAR32 indirect jump
jmp FFP PTR [var] ; FAR16 indirect jump
You can use an unconditional jump as a form of conditional jump by specifying

the address in a register or indirect memory operand. Also, you can use indirect
memory operands to construct jump tables that work like C switch statements,
Pascal CASE statements, or Basic ON GOTO, ON GOSUB, or SELECT
CASE statements, as shown in the following example.

NPVOID TYPEDEF NEAR PTR

.DATA
ctl_tbl NPVOID extended, ; Null key (extended code)
ctrla, ; Address of CONTROL-A key routine
ctrlb ; Address of CONTROL-B key routine
.CODE
.
.
.
mov ah, 8h ; Get a key
int 21h
cbw ; Stretch AL into AX
mov bx, ax ; Copy
shl bx, 1 ; Convert to address
jmp ctl_tbl[bx] ; Jump to key routine
extended:
mov ah, 8h ; Get second key of extended key
int 21h
. ; Use another jump table
. ; for extended keys
.
jmp next
ctrla: . ; CONTROL-A code here
.
.
jmp next
ctrlb: . ; CONTROL-B code here
.
.
jmp next
.
.
next: . ; Continue
In this instance, the indirect memory operands point to addresses of routines for
handling different keystrokes.
Conditional Jumps
The most common way to transfer control in assembly language is to use a
conditional jump. This is a two-step process:
1. First test the condition.
2. Then jump if the condition is true or continue if it is false.

All conditional jumps except two (JCXZ and JECXZ) use the processor flags
for their criteria. Thus, any statement that sets or clears a flag can serve as a test
basis for a conditional jump. The jump statement can be any one of 30
conditional-jump instructions. A conditional-jump instruction takes a single
operand containing the target address. You cannot use a pointer value as a target
as you can with unconditional jumps.
Jumping Based on the CX Register

JCXZ and JECXZ are special conditional jumps that do not consult the
processor flags. Instead, as their names imply, these instructions cause a jump
only if the CX or ECX register is zero. The use of JCXZ and JECXZ with
program loops is covered in the next section, “Loops.”
Jumping Based on the Processor Flags

The remaining conditional jumps in the processor’s repertoire all depend on the
status of the flags register. As the following list shows, several conditional jumps
have two or three names — JE (Jump if Equal) and JZ (Jump if Zero), for
example. Shared names assemble to exactly the same machine instruction, so
you may choose whichever mnemonic seems more appropriate. Jumps that
depend on the status of the flags register include:
Instruction Jumps if
JC/JB/JNAE Carry flag is set
JNC/JNB/JAE Carry flag is clear
JBE/JNA Either carry or zero flag is set
JA/JNBE Carry and zero flag are clear
JE/JZ Zero flag is set
JNE/JNZ Zero flag is clear
JL/JNGE Sign flag ≠ overflow flag
JGE/JNL Sign flag = overflow flag
JLE/JNG Zero flag is set or sign ≠ overflow
JG/JNLE Zero flag is clear and sign = overflow
JS Sign flag is set
JNS Sign flag is clear
JO Overflow flag is set
JNO Overflow flag is clear
JP/JPE Parity flag is set (even parity)
JNP/JPO Parity flag is clear (odd parity)
The last two jumps in the list, JPE (Jump if Parity Even) and JPO (Jump if
Parity Odd), are useful only for communications programs. The processor sets

the parity flag if an operation produces a result with an even number of set bits.
A communications program can compare the flag against the parity bit received
through the serial port to test for transmission errors.
The conditional jumps in the preceding list can follow any instruction that
changes the processor flags, as these examples show:
; Uses JO to handle overflow condition
add ax, bx ; Add two values
jo overflow ; If value too large, adjust
; Uses JNZ to check for zero as the result of subtraction

sub ax, bx ; Subtract
mov cx, Count ; First, initialize CX
jnz skip ; If the result is not zero, continue
call zhandler ; Else do special case
As the second example shows, the jump does not have to immediately follow
the instruction that alters the flags. Since MOV does not change the flags, it can
appear between the SUB instruction and the dependent jump.
There are three categories of conditional jumps:
u Comparison of two values
u Individual bit settings in a value
u Whether a value is zero or nonzero
Jumps Based on Comparison of Two Values

The CMP instruction is the most common way to test for conditional jumps. It
compares two values without changing either, then sets or clears the processor
flags according to the results of the comparison.
Internally, the CMP instruction is the same as the SUB instruction, except that
CMP does not change the destination operand. Both set flags according to the
result of the subtraction.

You can compare signed or unsigned values, but you must choose the
subsequent conditional jump to reflect the correct value type. For example, JL
(Jump if Less Than) and JB (Jump if Below) may seem conceptually similar,
but a failure to understand the difference between them can result in program
bugs. Table 7.1 shows the correct conditional jumps for comparisons of signed
and unsigned values. The table shows the zero, carry, sign, and overflow flags
as ZF, CF, SF, and OF, respectively.
Table 7.1 Conditional Jumps Based on Comparisons of Two Values
Signed Comparisons Unsigned Comparisons
Instruction Jump if True Instruction Jump if True
JE ZF = 1 JE ZF = 1
JNE ZF = 0 JNE ZF = 0
JG/JNLE ZF = 0 and SF = OF JA/JNBE CF = 0 and ZF = 0
JLE/JNG ZF = 1 or SF ≠ OF JBE/JNA CF = 1 or ZF = 1
JL/JNGE SF ≠ OF JB/JNAE CF = 1
JGE/JNL SF = OF JAE/JNB CF = 0
The mnemonic names of jumps always refer to the comparison of CMP’s first
operand (destination) with the second operand (source). For instance, in this
example, JG tests whether the first operand is greater than the second.
cmp ax, bx ; Compare AX and BX
jg next1 ; Equivalent to: If ( AX > BX ) goto next1
jl next2 ; Equivalent to: If ( AX < BX ) goto next2
Jumps Based on Bit Settings

The individual bit settings in a single value can also serve as the criteria for a
conditional jump. The TEST instruction tests whether specific bits in an operand
are on or off (set or clear), and sets the zero flag accordingly.

The TEST instruction is the same as the AND instruction, except that TEST
changes neither operand. The following example shows an application of TEST.
.DATA
bits BYTE ?
.CODE
.
.
.
; If bit 2 or bit 4 is set, then call task_a
; Assume "bits" is 0D3h 11010011
test bits, 10100y ; If 2 or 4 is set AND 00010100
jz skip1 ; --------
call task_a ; Then call task_a 00010000
skip1: ; Jump taken
.
.
.
; If bits 2 and 4 are clear, then call task_b
; Assume "bits" is 0E9h 11101001
test bits, 10100y ; If 2 and 4 are clear AND 00010100
jnz skip2 ; --------
call task_b ; Then call task_b 00000000
skip2: ; Jump taken
The source operand for TEST is often a mask in which the test bits are the only
bits set. The destination operand contains the value to be tested. If all the bits
set in the mask are clear in the destination operand, TEST sets the zero flag. If
any of the flags set in the mask are also set in the destination operand, TEST
clears the zero flag.
The 80386/486 processors provide additional bit-testing instructions. The BT
(Bit Test) series of instructions copy a specified bit from the destination operand
to the carry flag. A JC or JNC can then route program flow depending on the
result. For variations on the BT instruction, see the Reference.
Jumps Based on a Value of Zero

A program often needs to jump based on whether a particular register contains a
value of zero. We’ve seen how the JCXZ instruction jumps depending on the
value in the CX register. You can test for zero in other data registers nearly as
efficiently with the OR instruction. A program can OR a register with itself
without changing the register’s contents, then act on the resulting flags status.
For example, the following example tests whether BX is zero:
or bx, bx ; Is BX = 0?
jz is_zero ; Jump if so

This code is functionally equivalent to:

cmp bx, 0 ; Is BX = 0?
je is_zero ; Jump if so
but produces smaller and faster code, since it does not use an immediate number
as an operand. The same technique also lets you test a register’s sign bit:
or dx, dx ; Is DX sign bit set?
js sign_set ; Jump if so
Jump Extending
Unlike an unconditional jump, a conditional jump cannot reference a label more
than 128 bytes away. For example, the following statement is valid as long as
target is within a distance of 128 bytes:
; Jump to target less than 128 bytes away
jz target ; If previous operation resulted
; in zero, jump to target
However, if target is too distant, the following sequence is necessary to

enable a longer jump. Note this sequence is logically equivalent to the preceding
example:
; Jumps to distant targets previously required two steps
jnz skip ; If previous operation result is
; NOT zero, jump to "skip"
jmp target ; Otherwise, jump to target
skip:
MASM can automate jump-extending for you. If you target a conditional jump
to a label farther than 128 bytes away, MASM rewrites the instruction with an
unconditional jump, which ensures that the jump can reach its target. If target
lies within a 128-byte range, the assembler encodes the instruction jz target
as is. Otherwise, MASM generates two substitute instructions:
jne $ + 2 + (length in bytes of the next instruction)
jmp NEAR PTR target
The assembler generates this same code sequence if you specify the distance
with NEAR PTR, FAR PTR, or SHORT. Therefore,
jz NEAR PTR target
becomes

jne $ + 5
jmp NEAR PTR target
even if target is less than 128 bytes away.

MASM enables automatic jump expansion by default, but you can turn it off
with the NOLJMP form of the OPTION directive. For information about the
OPTION directive, see page 24.
If the assembler generates code to extend a conditional jump, it issues a level 3
warning saying that the conditional jump has been lengthened. You can set the
warning level to 1 for development and to level 3 for a final optimizing pass to
see if you can shorten jumps by reorganizing.
If you specify the distance for the jump and the target is out of range for that
distance, a “Jump out of Range” error results.
Since the JCXZ and JECXZ instructions do not have logical negations,
expansion of the jump instruction to handle targets with unspecified distances
cannot be performed for those instructions. Therefore, the distance must always
be short.
The size and distance of the target operand determines the encoding for
conditional or unconditional jumps to externals or targets in different segments.
The jump-extending and optimization features do not apply in this case.
Note Conditional jumps on the 80386 and 80486 processors can be to targets
up to 32K away, so jump extension occurs only for targets greater than that
distance.
Anonymous Labels
When you code jumps in assembly language, you must invent many label
names. One alternative to continually thinking up new label names is to use
anonymous labels, which you can use anywhere in your program. But because
anonymous labels do not provide meaningful names, they are best used for
jumping over only a few lines of code. You should mark major divisions of a
program with actual named labels.
Use two at signs (@@) followed by a colon (:) as an anonymous label. To jump
to the nearest preceding anonymous label, use @B (back) in the jump
instruction’s operand field; to jump to the nearest following anonymous label,
use @F (forward) in the operand field.
The jump in the following example targets an anonymous label:

jge @F
.
.
.
@@:
The items @B and @F always refer to the nearest occurrences of @@:, so there
is never any conflict between different anonymous labels.
Decision Directives
The high-level structures you can use for decision-making are the .IF, .ELSEIF,
and .ELSE statements. These directives generate conditional jumps. The
expression following the .IF directive is evaluated, and if true, the following
instructions are executed until the next .ENDIF, .ELSE, or .ELSEIF directive is
reached. The .ELSE statements execute if the expression is false. Using the
.ELSEIF directive puts a new expression inside the alternative part of the
original .IF statement to be evaluated. The syntax is:
.IF condition1
statements
[[.ELSEIF condition2
statements]]
[[.ELSE
statements]]
.ENDIF
The decision structure
.IF cx == 20
mov dx, 20
.ELSE
mov dx, 30
.ENDIF
generates this code:

.IF cx == 20
0017 83 F9 14 * cmp cx, 014h
001A 75 05 * jne @C0001
001C BA 0014 mov dx, 20
.ELSE
001F EB 03 * jmp @C0003
0021 *@C0001:
0021 BA 001E mov dx, 30
.ENDIF
0024 *@C0003:

Loops
Loops repeat an action until a termination condition is reached. This condition
can be a counter or the result of an expression’s evaluation. MASM 6.1 offers
many ways to set up loops in your programs. The following list compares
MASM loop structures:
Instructions Action
LOOP Automatically decrements CX. When CX = 0, the loop ends. The top
of the loop cannot be greater than 128 bytes from the LOOP
instruction. (This is true for all LOOP instructions.)
LOOPE/LOOPZ, Loops while equal or not equal. Checks both CX and the state of the
LOOPNE/LOOPNZ zero flag. LOOPZ ends when either CX=0 or the zero flag is clear,
whichever occurs first. LOOPNZ ends when either CX=0 or the zero
flag is set, whichever occurs first. LOOPE and LOOPZ assemble to
the same machine instruction, as do LOOPNE and LOOPNZ. Use
whichever mnemonic best fits the context of your loop. Set CX to a
number out of range if you don’t want a count to control the loop.
JCXZ, JECXZ Branches to a label only if CX = 0 or ECX = 0. Unlike other
conditional-jump instructions, which can jump to either a near or a
short label under the 80386 or 80486, JCXZ and JECXZ always
jump to a short label.
Conditional jumps Acts only if certain conditions met. Necessary if several conditions
must be tested. See “Conditional Jumps,” page 164.
The following examples illustrate these loop constructions.

; The LOOP instruction: For 200 to 0 do task
mov cx, 200 ; Set counter
next: . ; Do the task here
.
.
loop next ; Do again
; Continue after loop
; The LOOPNE instruction: While AX is not 'Y', do task

mov cx, 256 ; Set count too high to interfere
wend: . ; But don't do more than 256 times
. ; Some statements that change AX
.
cmp al, 'Y' ; Is it Y or too many times?
loopne wend ; No? Repeat
; Yes? Continue
The JCXZ and JECXZ instructions provide an efficient way to avoid executing
loops when the loop counter CX is empty. For example, consider the following
loops:

mov cx, LoopCount ; Load loop counter

next: . ; Iterate loop CX times
.
.
If LoopCount is zero, CX decrements to -1 on the first pass. It then must

decrement 65,535 more times before reaching 0. Use a JCXZ to avoid this
problem:
mov cx, LoopCount ; Load loop counter
jcxz done ; Skip loop if count is 0
next: . ; Else iterate loop CX times
.
.
done: ; Continue after loop
Loop-Generating Directives
The high-level control structures generate loop structures for you. These
directives are similar to the while and repeat loops of C or Pascal, and can
make your assembly programs easier to code and to read. The assembler
generates the appropriate assembly code. These directives are summarized as
follows:
Directives Action
.WHILE ... .ENDW The statements between .WHILE condition and .ENDW
execute while the condition is true.
.REPEAT ... .UNTIL The loop executes at least once and continues until the condition
given after .UNTIL is true. Generates conditional jumps.
.REPEAT ... .UNTILCXZ Compares label to an expression and generates appropriate
loop instructions.
.BREAK End a .REPEAT or a .WHILE loop unconditionally.
.CONTINUE Jump unconditionally past any remaining code to bottom of
loop.

These constructs work much as they do in a high-level language such as C or

Pascal. Keep in mind the following points:
u These directives generate appropriate processor instructions. They are not
new instructions.
u They require proper use of signed and unsigned data declarations.
These directives cause a set of instructions to execute based on the evaluation of

some condition. This condition can be an expression that evaluates to a signed
or unsigned value, an expression using the binary operators in C (&&, ||, or !),
or the state of a flag. For more information about expression operators, see page
178.
The evaluation of the condition requires the assembler to know if the operands
in the condition are signed or unsigned. To state explicitly that a named memory
location contains a signed integer, use the signed data allocation directives
SBYTE, SWORD, and SDWORD.
.WHILE Loops
As with while loops in C or Pascal, the test condition for .WHILE is checked
before the statements inside the loop execute. If the test condition is false, the
loop does not execute. While the condition is true, the statements inside the loop
repeat.
Use the .ENDW directive to mark the end of the .WHILE loop. When the
condition becomes false, program execution begins at the first statement
following the .ENDW directive. The .WHILE directive generates appropriate
compare and jump statements. The syntax is:
.WHILE condition
statements
.ENDW
For example, this loop copies the contents of one buffer to another until a ‘$’
character (marking the end of the string) is found:
.DATA
buf1 BYTE "This is a string",'$'
buf2 BYTE 100 DUP (?)
.CODE
sub bx, bx ; Zero out bx
.WHILE (buf1[bx] != '$')
mov al, buf1[bx] ; Get a character
mov buf2[bx], al ; Move it to buffer 2
inc bx ; Count forward
.ENDW

.REPEAT Loops
MASM’s .REPEAT directive allows for loop constructions like the do loop of C
and the REPEAT loop of Pascal. The loop executes until the condition
following the .UNTIL (or .UNTILCXZ) directive becomes true. Since the
condition is checked at the end of the loop, the loop always executes at least
once. The .REPEAT directive generates conditional jumps. The syntax is:
.REPEAT
statements
.UNTIL condition
.REPEAT
statements
.UNTILCXZ [[condition]]
where condition can also be expr1 == expr2 or expr1 != expr2. When two
conditions are used, expr2 can be an immediate expression, a register, or (if
expr1 is a register) a memory location.
For example, the following code fills a buffer with characters typed at the
keyboard. The loop ends when the ENTER key (character 13) is pressed:
.DATA
buffer BYTE 100 DUP (0)
.CODE
sub bx, bx ; Zero out bx
.REPEAT
mov ah, 01h
int 21h ; Get a key
mov buffer[bx], al ; Put it in the buffer
inc bx ; Increment the count
.UNTIL (al == 13) ; Continue until al is 13

The .UNTIL directive generates conditional jumps, but the .UNTILCXZ

directive generates a LOOP instruction, as shown by the listing file code for
these examples. In a listing file, assembler-generated code is preceded by an
asterisk.
ASSUME bx:PTR SomeStruct
.REPEAT
*@C0001:
inc ax
.UNTIL ax==6
* cmp ax, 006h
* jne @C0001
.REPEAT
*@C0003:
mov ax, 1
.UNTILCXZ
* loop @C0003
.REPEAT
*@C0004:
.UNTILCXZ [bx].field != 6
* cmp [bx].field, 006h
* loope @C0004
.BREAK and .CONTINUE Directives

The .BREAK and .CONTINUE directives terminate a .REPEAT or .WHILE
loop prematurely. These directives allow an optional .IF clause for conditional
breaks. The syntax is:
.BREAK [[.IF condition]]
.CONTINUE [[.IF condition]]
Note that .ENDIF is not used with the .IF forms of .BREAK and .CONTINUE
in this context. The .BREAK and .CONTINUE directives work the same way
as the break and continue instructions in C. Execution continues at the
instruction following the .UNTIL, .UNTILCXZ, or .ENDW of the nearest
enclosing loop.
Instead of ending the loop execution as .BREAK does, .CONTINUE causes
loop execution to jump directly to the code that evaluates the loop condition of
the nearest enclosing loop.
The following loop accepts only the keys in the range ‘0’ to ‘9’ and terminates
when you press ENTER.

.WHILE 1 ; Loop forever

mov ah, 08h ; Get key without echo
int 21h
.BREAK .IF al == 13 ; If ENTER, break out of the loop
.CONTINUE .IF (al < '0') || (al > '9')
; If not a digit, continue looping
mov dl, al ; Save the character for processing
mov ah, 02h ; Output the character
int 21h
.ENDW
If you assemble the preceding source code with the /Fl and /Sg command-line
options and then view the results in the listing file, you will see this code:
.WHILE 1
0017 *@C0001:
0017 B4 08 mov ah, 08h
0019 CD 21 int 21h
.BREAK .IF al == 13
001B 3C 0D * cmp al, 00Dh
001D 74 10 * je @C0002
.CONTINUE .IF (al '0') || (al '9')
001F 3C 30 * cmp al, '0'
0021 72 F4 * jb @C0001
0023 3C 39 * cmp al, '9'
0025 77 F0 * ja @C0001
0027 8A D0 mov dl, al
0029 B4 02 mov ah, 02h
002B CD 21 int 21h
.ENDW
002D EB E8 * jmp @C0001
002F *@C0002:
The high-level control structures can be nested. That is, .REPEAT or .WHILE
loops can contain .REPEAT or .WHILE loops as well as .IF statements.
If the code generated by a .WHILE loop, .REPEAT loop, or .IF statement
generates a conditional or unconditional jump, MASM encodes the jump using
the jump extension and jump optimization techniques described in
“Unconditional Jumps,” page 162, and “Conditional Jumps,” page 164.

Writing Loop Conditions

You can express the conditions of the .IF, .REPEAT, and .WHILE directives
using relational operators, and you can express the attributes of the operand with
the PTR operator. To write loop conditions, you also need to know how the
assembler evaluates the operators and operands in the condition. This section
explains the operators, attributes, precedence level, and expression evaluation
order for the conditions used with loop-generating directives.
Expression Operators
The binary relational operators in MASM 6.1 are the same binary operators
used in C. These operators generate MASM compare, test, and conditional
jump instructions. High-level control instructions include:
Operator Meaning
== Equal
!= Not equal
> Greater than
>= Greater than or equal to
< Less than
<= Less than or equal to
& Bit test
! Logical NOT
&& Logical AND
|| Logical OR
A condition without operators (other than !) tests for nonzero as it does in C.

For example, .WHILE (x) is the same as .WHILE (x != 0), and .WHILE
(!x) is the same as .WHILE (x == 0).
You can also use the flag names (ZERO?, CARRY?, OVERFLOW?, SIGN?,
and PARITY?) as operands in conditions with the high-level control structures.
For example, in .WHILE (CARRY?), the value of the carry flag determines the
outcome of the condition.
Signed and Unsigned Operands

Expression operators generate unsigned jumps by default. However, if either
side of the operation is signed, the assembler considers the entire operation
signed.

You can use the PTR operator to tell the assembler that a particular operand in
a register or constant is a signed number, as in these examples:
.WHILE SWORD PTR [bx] <= 0
.IF SWORD PTR mem1 > 0
Without the PTR operator, the assembler would treat the contents of BX as an
unsigned value.
You can also specify the size attributes of operands in memory locations with
SBYTE, SWORD, and SDWORD, for use with .IF, .WHILE, and .REPEAT.
.DATA
mem1 SBYTE ?
mem2 WORD ?
.IF mem1 > 0
.WHILE mem2 < bx
.WHILE SWORD PTR ax < count
Precedence Level
As with C, you can concatenate conditions with the && operator for AND, the
|| operator for OR, and the ! operator for negate. The precedence level is !, &&,
and ||, with ! having the highest priority. Like expressions in high-level languages,
precedence is evaluated left to right.
Expression Evaluation
The assembler evaluates conditions created with high-level control structures
according to short-circuit evaluation. If the evaluation of a particular condition
automatically determines the final result (such as a condition that evaluates to
false in a compound statement concatenated with AND), the evaluation does not
continue.
For example, in this .WHILE statement,
.WHILE (ax > 0) && (WORD PTR [bx] == 0)
the assembler evaluates the first condition. If this condition is false (that is, if
AX is less than or equal to 0), the evaluation is finished. The second condition is
not checked and the loop does not execute, because a compound condition
containing && requires both expressions to be true for the entire condition to be
true.

Procedures
Organizing your code into procedures that execute specific tasks divides large
programs into manageable units, allows for separate testing, and makes code
more efficient for repetitive tasks.
Assembly-language procedures are similar to functions, subroutines, and
procedures in high-level languages such as C, FORTRAN, and Pascal. Two
instructions control the use of assembly-language procedures. CALL pushes the
return address onto the stack and transfers control to a procedure, and RET
pops the return address off the stack and returns control to that location.
The PROC and ENDP directives mark the beginning and end of a procedure.
Additionally, PROC can automatically:
u Preserve register values that should not change but that the procedure might
otherwise alter.
u Set up a local stack pointer, so that you can access parameters and local
variables placed on the stack.
u Adjust the stack when the procedure ends.
Defining Procedures
Procedures require a label at the start of the procedure and a RET instruction at
the end. Procedures are normally defined by using the PROC directive at the
start of the procedure and the ENDP directive at the end. The RET instruction
normally is placed immediately before the ENDP directive. The assembler
makes sure the distance of the RET instruction matches the distance defined by
the PROC directive. The basic syntax for PROC is:
label PROC [[NEAR | FAR]]
.
.
.
RET [[constant]]
label ENDP
The CALL instruction pushes the address of the next instruction in your code
onto the stack and passes control to a specified address. The syntax is:
CALL {label | register | memory}
The operand contains a value calculated at run time. Since that operand can be a
register, direct memory operand, or indirect memory operand, you can write call
tables similar to the example code on page 164.

Calls can be near or far. Near calls push only the offset portion of the calling
address and therefore must target a procedure within the same segment or
group. You can specify the type for the target operand. If you do not, MASM
uses the declared distance (NEAR or FAR) for operands that are labels and for
the size of register or memory operands. The assembler then encodes the call
appropriately, as it does with unconditional jumps. (See previous “Unconditional
Jumps” and “Conditional Jumps.”)
MASM optimizes a call to a far non-external label when the label is in the
current segment by generating the code for a near call, saving one byte.
You can define procedures without PROC and ENDP, but if you do, you must
make sure that the size of the CALL matches the size of the RET. You can
specify the RET instruction as RETN (Return Near) or RETF (Return Far) to
override the default size:
call NEAR PTR task ; Call is declared near
. ; Return comes to here
.
.
task: ; Procedure begins with near label
.
. ; Instructions go here
.
retn ; Return declared near
The syntax for RETN and RETF is:

label: | label LABEL NEAR
statements
RETN [[constant]]
label LABEL FAR
statements
RETF [[constant]]
The RET instruction (and its RETF and RETN variations) allows an optional
constant operand that specifies a number of bytes to be added to the value of
the SP register after the return. This operand adjusts for arguments passed to
the procedure before the call, as shown in the example in “Using Local
Variables,” following.
When you define procedures without PROC and ENDP, you must make sure
that calls have the same size as corresponding returns. For example, RETF pops
two words off the stack. If a NEAR call is made to a procedure with a far
return, the popped value is meaningless, and the stack status may cause the
execution to return to a random memory location, resulting in program failure.

An extended PROC syntax automates many of the details of accessing

arguments and saving registers. See “Declaring Parameters with the PROC
Directive,” later in this chapter.
Passing Arguments on the Stack

Each time you call a procedure, you may want it to operate on different data.
This data, called “arguments,” can be passed to the procedure in various ways.
Although you can pass arguments to a procedure in registers or in variables, the
most common method is the stack. Microsoft languages have specific
conventions for passing arguments. These conventions for assembly-language
modules shared with modules from high-level languages are explained in Chapter
12, “Mixed-Language Programming.”
This section describes how a procedure accesses the arguments passed to it on
the stack. Each argument is accessed as an offset from BP. However, if you use
the PROC directive to declare parameters, the assembler calculates these offsets
for you and lets you refer to parameters by name. The next section, “Declaring
Parameters with the PROC Directive,” explains how to use PROC this way.
This example shows how to pass arguments to a procedure. The procedure
expects to find those arguments on the stack. As this example shows, arguments
must be accessed as offsets of BP.
; C-style procedure call and definition
mov ax, 10 ; Load and

push ax ; push constant as third argument
push arg2 ; Push memory as second argument
push cx ; Push register as first argument
call addup ; Call the procedure
add sp, 6 ; Destroy the pushed arguments
. ; (equivalent to three pops)
.
.
addup PROC NEAR ; Return address for near call
; takes two bytes
push bp ; Save base pointer - takes two bytes
; so arguments start at fourth byte
mov bp, sp ; Load stack into base pointer
mov ax, [bp+4] ; Get first argument from
; fourth byte above pointer
add ax, [bp+6] ; Add second argument from
; sixth byte above pointer
add ax, [bp+8] ; Add third argument from
; eighth byte above pointer
pop bp ; Restore BP
ret ; Return result in AX
addup ENDP

Figure 7.1 shows the stack condition at key points in the process.
Figure 7.1 Procedure Arguments on the Stack
Starting with the 80186 processor, the ENTER and LEAVE instructions
simplify the stack setup and restore instructions at the beginning and end of
procedures. However, ENTER uses a lot of time. It is necessary only with
nested, statically-scoped procedures. Thus, a Pascal compiler may sometimes
generate ENTER. The LEAVE instruction, on the other hand, is an efficient
way to do the stack cleanup. LEAVE reverses the effect of the last ENTER
instruction by restoring BP and SP to their values before the procedure call.

Declaring Parameters with the PROC Directive

With the PROC directive, you can specify registers to be saved, define param-
eters to the procedure, and assign symbol names to parameters (rather than as
offsets from BP). This section describes how to use the PROC directive to
automate the parameter-accessing techniques described in the last section.
For example, the following diagram shows a valid PROC statement for a
procedure called from C. It takes two parameters, var1 and arg1, and uses
(and must save) the DI and SI registers:
The syntax for PROC is:

label PROC [[attributes]] [[USES reglist]] [[, ]] [[parameter[[:tag]]... ]]
The parts of the PROC directive include:
label The name of the procedure.
attributes Any of several attributes of the procedure, including the distance, langtype, and
visibility of the procedure. The syntax for attributes is given on the following
page.
reglist A list of registers following the USES keyword that the procedure uses, and that
should be saved on entry. Registers in the list must be separated by blanks or
tabs, not by commas. The assembler generates prologue code to push these
registers onto the stack. When you exit, the assembler generates epilogue code
to pop the saved register values off the stack.
parameter The list of parameters passed to the procedure on the stack. The list can have a
variable number of parameters. See the discussion following for the syntax of
parameter. This list can be longer than one line if the continued line ends with a
comma.
This diagram shows a valid PROC definition that uses several attributes:

Attributes
The syntax for the attributes field is:
[[distance]] [[langtype]] [[visibility]] [[<prologuearg>]]
The explanations for these options include:
distance Controls the form of the RET instruction generated. Can be NEAR or FAR. If
distance is not specified, it is determined from the model declared with the
.MODEL directive. NEAR distance is assumed for TINY, SMALL,
COMPACT, and FLAT. The assembler assumes FAR distance for MEDIUM ,
LARGE , and HUGE . For 80386/486 programming with 16- and 32-bit
segments, you can specify NEAR16, NEAR32, FAR16, or FAR32.
langtype Determines the calling convention used to access parameters and restore the stack.
The BASIC, FORTRAN, and PASCAL langtypes convert procedure names to
uppercase, place the last parameter in the parameter list lowest on the stack, and
generate a RET num instruction to end the procedure. The RET adjusts the stack
upward by num, which represents the number of bytes in the argument list. This
step, called “cleaning the stack,” returns the stack pointer SP to the value it had
before the caller pushed any arguments.
The C and STDCALL langtype prefixes an underscore to the procedure name
when the procedure’s scope is PUBLIC or EXPORT and places the first
parameter lowest on the stack. SYSCALL is equivalent to the C calling convention
with no underscore prefixed to the procedure’s name. STDCALL uses caller
stack cleanup when :VARARG is specified; otherwise the called routine must
clean up the stack (see Chapter 12).
visibility Indicates whether the procedure is available to other modules. The visibility can
be PRIVATE, PUBLIC, or EXPORT. A procedure name is PUBLIC unless it
is explicitly declared as PRIVATE. If the visibility is EXPORT, the linker places
the procedure’s name in the export table for segmented executables. EXPORT
also enables PUBLIC visibility.
You can explicitly set the default visibility with the OPTION directive. OPTION
PROC:PUBLIC sets the default to public. For more information, see Chapter 1,
“Using the Option Directive.”
prologuearg Specifies the arguments that affect the generation of prologue and epilogue code
(the code MASM generates when it encounters a PROC directive or the end of a
procedure). For an explanation of prologue and epilogue code, see “Generating
Prologue and Epilogue Code,” later in this chapter.

Parameters
The comma that separates parameters from reglist is optional, if both fields
appear on the same line. If parameters appears on a separate line, you must end
the reglist field with a comma. In the syntax:
parmname [[:tag]
parmname is the name of the parameter. The tag can be the qualifiedtype or
the keyword VARARG. However, only the last parameter in a list of param-
eters can use the VARARG keyword. The qualifiedtype is discussed in “Data
Types,” Chapter 1. An example showing how to reference VARARG param-
eters appears later in this section. You can nest procedures if they do not have
parameters or USES register lists. This diagram shows a procedure definition
with one parameter definition.
The procedure presented in “Passing Arguments on the Stack,” page 182, is

here rewritten using the extended PROC functionality. Prior to the procedure
call, you must push the arguments onto the stack unless you use INVOKE.
(See “Calling Procedures with INVOKE,” later in this chapter.)
addup PROC NEAR C,
arg1:WORD, arg2:WORD, count:WORD
mov ax, arg1
add ax, count
add ax, arg2
ret
addup ENDP
If the arguments for a procedure are pointers, the assembler does not generate
any code to get the value or values that the pointers reference; your program
must still explicitly treat the argument as a pointer. (For more information about
using pointers, see Chapter 3, “Using Addresses and Pointers.”)

In the following example, even though the procedure declares the parameters as
near pointers, you must code two MOV instructions to get the values of the
param-
eters. The first MOV gets the address of the parameters, and the second MOV
gets the parameter.
; Call from C as a FUNCTION returning an integer
.MODEL medium, c
.CODE
myadd PROC arg1:NEAR PTR WORD, arg2:NEAR PTR WORD
mov bx, arg1 ; Load first argument

mov ax, [bx]
mov bx, arg2 ; Add second argument
add ax, [bx]
ret
myadd ENDP
You can use conditional-assembly directives to make sure your pointer

parameters are loaded correctly for the memory model. For example, the
following version of myadd treats the parameters as FAR parameters, if
necessary.
.MODEL medium, c ; Could be any model
.CODE
myadd PROC arg1:PTR WORD, arg2:PTR WORD
IF @DataSize
les bx, arg1 ; Far parameters
mov ax, es:[bx]
les bx, arg2
add ax, es:[bx]
ELSE
mov bx, arg1 ; Near parameters
mov ax, [bx]
mov bx, arg2
add ax, [bx]
ENDIF
ret
myadd ENDP

Using VARARG
In the PROC statement, you can append the :VARARG keyword to the last
parameter to indicate that the procedure accepts a variable number of
arguments. However, :VARARG applies only to the C, SYSCALL, or
STDCALL calling conventions (see Chapter 12). A symbol must precede
:VARARG so the procedure can access arguments as offsets from the given
variable name, as this example illustrates:
addup3 PROTO NEAR C, argcount:WORD, arg1:VARARG
invoke addup3, 3, 5, 2, 4
addup3 PROC NEAR C, argcount:WORD, arg1:VARARG

sub ax, ax ; Clear work register
sub si, si
.WHILE argcount > 0 ; Argcount has number of arguments

add ax, arg1[si] ; Arg1 has the first argument
dec argcount ; Point to next argument
inc si
inc si
.ENDW
ret ; Total is in AX
addup3 ENDP
You can pass non-default-sized pointers in the VARARG portion of the

parameter list by separately passing the segment portion and the offset portion
of the address.
Note When you use the extended PROC features and the assembler encounters
a RET instruction, it automatically generates instructions to pop saved registers,
remove local variables from the stack, and, if necessary, remove parameters. It
generates this code for each RET instruction it encounters. You can reduce code
size by having only one return and jumping to it from various locations.
Using Local Variables

In high-level languages, local variables are visible only within a procedure. In
Microsoft languages, these variables are usually stored on the stack. In
assembly-language programs, you can also have local variables. These variables
should not be confused with labels or variable names that are local to a module,
as described in Chapter 8, “Sharing Data and Procedures Among Modules and
Libraries.”

This section outlines the standard methods for creating local variables. The next
section shows how to use the LOCAL directive to make the assembler

automatically generate local variables. When you use this directive, the
assembler generates the same instructions as those demonstrated in this section
but handles some of the details for you.
If your procedure has relatively few variables, you can usually write the most
efficient code by placing these values in registers. Use local (stack) data when
you have a large amount of temporary data for the procedure.
To use a local variable, you must save stack space for it at the start of the
procedure. A procedure can then reference the variable by its position in the
stack. At the end of the procedure, you must clean the stack by restoring the
stack pointer. This effectively throws away all local variables and regains the
stack space they occupied.
This example subtracts 2 bytes from the SP register to make room for a local
word variable, then accesses the variable as [bp-2].
push ax ; Push one argument
call task ; Call
.
.
.
task PROC NEAR

push bp ; Save base pointer
mov bp, sp ; Load stack into base pointer
sub sp, 2 ; Save two bytes for local variable
.
.
.
mov WORD PTR [bp-2], 3 ; Initialize local variable
add ax, [bp-2] ; Add local variable to AX
sub [bp+4], ax ; Subtract local from argument
. ; Use [bp-2] and [bp+4] in
. ; other operations
.
mov sp, bp ; Clear local variables
pop bp ; Restore base
ret 2 ; Return result in AX and pop
task ENDP ; two bytes to clear parameter
Notice the instruction mov sp,bp at the end of the procedure restores the
original value of SP. The statement is required only if the value of SP changes
inside the procedure (usually by allocating local variables). The argument passed
to the procedure is removed with the RET instruction. Contrast this to the
example in “Passing Arguments on the Stack,” page 182, in which the calling
code adjusts the stack for the argument.

Figure 7.2 shows the stack at key points in the process.
Figure 7.2 Local Variables on the Stack
Creating Local Variables Automatically

MASM’s LOCAL directive automates the process for creating local variables
on the stack. LOCAL frees you from having to count stack words, and it
makes your code easier to write and maintain. This section illustrates the
advantages of creating temporary data with the LOCAL directive.
To use the LOCAL directive, list the variables you want to create, giving a type
for each one. The assembler calculates how much space is required on the
stack. It also generates instructions to properly decrement SP (as described in
the previous section) and to reset SP when you return from the procedure.
When you create local variables this way, your source code can refer to each
local variable by name rather than as an offset of the stack pointer. Moreover,

the assembler generates debugging information for each local variable. If you
have programmed before in a high-level language that allows scoping, local
variables will seem familiar. For example, a C compiler sets up variables with
automatic storage class in the same way as the LOCAL directive.
We can simplify the procedure in the previous section with the following code:
task PROC NEAR arg:WORD
LOCAL loc:WORD
.
.
.
mov loc, 3 ; Initialize local variable
add ax, loc ; Add local variable to AX
sub arg, ax ; Subtract local from argument
. ; Use "loc" and "arg" in other operations
.
.
ret
task ENDP
The LOCAL directive must be on the line immediately following the PROC
statement with the following syntax:
LOCAL vardef [[, vardef]]...
Each vardef defines a local variable. A local variable definition has this form:
label[[[count]]][[:qualifiedtype]]
These are the parameters in local variable definitions:
label The name given to the local variable. You can use this name to access the
variable.
count The number of elements of this name and type to allocate on the stack. You
can allocate a simple array on the stack with count. The brackets around
count are required. If this field is omitted, one data object is assumed.
qualifiedtype A simple MASM type or a type defined with other types and attributes. For
more information, see “Data Types” in Chapter 1.
If the number of local variables exceeds one line, you can place a comma at the
end of the first line and continue the list on the next line. Alternatively, you can
use several consecutive LOCAL directives.

The assembler does not initialize local variables. Your program must include
code to perform any necessary initializations. For example, the following code
fragment sets up a local array and initializes it to zero:
arraysz EQU 20
aproc PROC USES di

LOCAL var1[arraysz]:WORD, var2:WORD
.
.
.
; Initialize local array to zero
push ss
pop es ; Set ES=SS
lea di, var1 ; ES:DI now points to array
mov cx, arraysz ; Load count
sub ax, ax
rep stosw ; Store zeros
; Use the array...
.
.
.
ret
aproc ENDP
Even though you can reference stack variables by name, the assembler treats
them as offsets of BP, and they are not visible outside the procedure. In the
following procedure, array is a local variable.
index EQU 10
test PROC NEAR
LOCAL array[index]:WORD
.
.
.
mov bx, index
; mov array[bx], 5 ; Not legal!
The second MOV statement may appear to be legal, but since array is an
offset of BP, this statement is the same as
; mov [bp + bx + arrayoffset], 5 ; Not legal!
BP and BX can be added only to SI and DI. This example would be legal,
however, if the index value were moved to SI or DI. This type of error in your
program can be difficult to find unless you keep in mind that local variables in
procedures are offsets of BP.

Declaring Procedure Prototypes

MASM provides the INVOKE directive to handle many of the details important
to procedure calls, such as pushing parameters according to the correct calling
conventions. To use INVOKE, the procedure called must have been declared
previously with a PROC statement, an EXTERNDEF (or EXTERN) statement,
or a TYPEDEF. You can also place a prototype defined with PROTO before
the INVOKE if the procedure type does not appear before the INVOKE.
Procedure prototypes defined with PROTO inform the assembler of types and
numbers of arguments so the assembler can check for errors and provide
automatic conversions when INVOKE calls the procedure.
Declaring procedure prototypes is good programming practice, but is optional.
Prototypes in MASM perform the same function as prototypes in C and other
high-level languages. A procedure prototype includes the procedure name, the
types, and (optionally) the names of all parameters the procedure expects.
Prototypes usually are placed at the beginning of an assembly program or in a
separate include file so the assembler encounters the prototype before the actual
procedure.
Prototypes enable the assembler to check for unmatched parameters and are
especially useful for procedures called from other modules and other languages.
If you write routines for a library, you may want to put prototypes into an
include file for all the procedures used in that library. For more information
about using include files, see Chapter 8, “Sharing Data and Procedures among
Modules and Libraries.”
The PROTO directive provides one way to define a procedure prototype. The
syntax for a prototype definition is the same as for a procedure declaration (see
“Declaring Parameters with the PROC Directive,” earlier in this chapter), except
that you do not include the list of registers, prologuearg list, or the scope of the
procedure.
Also, the PROTO keyword precedes the langtype and distance attributes. The
attributes (like C and FAR) are optional. However, if they are not specified, the
defaults are based on any .MODEL or OPTION LANGUAGE statement. The
names of the parameters are also optional, but you must list parameter types. A
label preceding :VARARG is also optional in the prototype but not in the
PROC statement.
If a PROTO and a PROC for the same function appear in the same module,
they must match in attribute, number of parameters, and parameter types. The
easiest way to create prototypes with PROTO is to write your procedure and
then copy the first line (the line that contains the PROC keyword) to a location
in your program that follows the data declarations. Change PROC to PROTO
and remove the USES reglist, the prologuearg field, and the visibility field. It

is important that the prototype follow the declarations for any types used in it to
avoid any forward references used by the parameters in the prototype.
The following example illustrates how to define and then declare two typical
procedures. In both prototype and declaration, the comma before the argument
list is optional only when the list does not appear on a separate line:
; Procedure prototypes.
addup PROTO NEAR C argcount:WORD, arg2:WORD, arg3:WORD

myproc PROTO FAR C, argcount:WORD, arg2:VARARG
; Procedure declarations
addup PROC NEAR C, argcount:WORD, arg2:WORD, arg3:WORD

.
.
.
myproc PROC FAR C PUBLIC <callcount> USES di si,
argcount:WORD,
arg2:VARARG
When you call a procedure with INVOKE, the assembler checks the arguments
given by INVOKE against the parameters expected by the procedure. If the
data types of the arguments do not match, MASM reports an error or converts
the type to the expected type. These conversions are explained in the next
section.
Calling Procedures with INVOKE

INVOKE generates a sequence of instructions that push arguments and call a
procedure. This helps maintain code if arguments or langtype for a procedure
are changed. INVOKE generates procedure calls and automatically:
u Converts arguments to the expected types.
u Pushes arguments on the stack in the correct order.
u Cleans the stack when the procedure returns.
If arguments do not match in number or if the type is not one the assembler can
convert, an error results.
If the procedure uses VARARG, INVOKE can pass a number of arguments
different from the number in the parameter list without generating an error or
warning. Any additional arguments must be at the end of the INVOKE
argument list. All other arguments must match those in the prototype parameter
list.

The syntax for INVOKE is:

INVOKE expression [[, arguments]]
where expression can be the procedure’s label or an indirect reference to a
procedure, and arguments can be an expression, a register pair, or an expression
preceded with ADDR. (The ADDR operator is discussed later in this chapter.)
Procedures with these prototypes
addup PROTO NEAR C argcount:WORD, arg2:WORD, arg3:WORD
myproc PROTO FAR C, argcount:WORD, arg2:VARARG
and these procedure declarations

addup PROC NEAR C, argcount:WORD, arg2:WORD, arg3:WORD
.
.
.
myproc PROC FAR C PUBLIC <callcount> USES di si,
argcount:WORD,
arg2:VARARG
can be called with INVOKE statements like this:

INVOKE addup, ax, x, y
INVOKE myproc, bx, cx, 100, 10
The assembler can convert some arguments and parameter type combinations
so that the correct type can be passed. The signed or unsigned qualities of the
arguments in the INVOKE statements determine how the assembler converts
them to the types expected by the procedure.
The addup procedure, for example, expects parameters of type WORD, but
the arguments passed by INVOKE to the addup procedure can be any of these
types:
u BYTE, SBYTE, WORD, or SWORD
u An expression whose type is specified with the PTR operator to be one of
those types
u An 8-bit or 16-bit register
u An immediate expression in the range –32K to +64K
u A NEAR PTR
If the type is smaller than that expected by the procedure, MASM widens the
argument to match.

Widening Arguments
For INVOKE to correctly handle type conversions, you must use the signed
data types for any signed assignments. MASM widens an argument to match the
type expected by a procedure’s parameters in these cases:
Type Passed Type Expected
BYTE, SBYTE WORD, SWORD, DWORD, SDWORD
WORD, SWORD DWORD, SDWORD
The assembler can extend a segment if far data is expected, and it can convert
the type given in the list to the types expected. If the assembler cannot convert
the type, however, it generates an error.
Detecting Errors
If the assembler needs to widen an argument, it first copies the value to AL or
AX. It widens an unsigned value by placing a zero in the higher register area,
and widens a signed value with a CBW, CWD, or CWDE instruction as
required. Similarly, the assembler copies a constant argument value into AL or
AX when the .8086 directive is in effect. You can see these generated
instructions in the listing file when you include the /Sg command-line option.
Using the accumulator register to widen or copy an argument may lead to an
error if you attempt to pass AX as another argument. For example, consider the
following INVOKE statement for a procedure with the C calling convention
INVOKE myprocA, ax, cx, 100, arg
where arg is a BYTE variable and myproc expects four arguments of type
WORD. The assembler widens and then pushes arg like this:
mov al, DGROUP:arg
xor ah, ah
push ax
The generated code thus overwrites the last argument (AX) passed to the
procedure. The assembler generates an error in this case, requiring you to
rewrite the INVOKE statement.
To summarize, the INVOKE directive overwrites AX and perhaps DX when
widening arguments. It also uses AX to push constants on the 8088 and 8086. If
you use these registers (or EAX and EDX on an 80386/486) to pass arguments,
they may be overwritten. The assembler’s error detection prevents this from
ever becoming a run-time bug, but AX and DX should remain your last choice
for holding arguments.

Invoking Far Addresses

You can pass a FAR pointer in a segment::offset pair, as shown in the following.
Note the use of double colons to separate the register pair. The registers could
be any other register pair, including a pair that an MS-DOS call uses to return
values.
FPWORD TYPEDEF FAR PTR WORD
SomeProc PROTO var1:DWORD, var2:WORD, var3:WORD
pfaritem FPWORD faritem

.
.
.
les bx, pfaritem
INVOKE SomeProc, ES::BX, arg1, arg2
However, INVOKE cannot combine into a single address one argument for the
segment and one for the offset.
Passing an Address
You can use the ADDR operator to pass the address of an expression to a
procedure that expects a NEAR or FAR pointer. This example generates code
to pass a far pointer (to arg1) to the procedure proc1.
PBYTE TYPEDEF FAR PTR BYTE
arg1 BYTE "This is a string"
proc1 PROTO NEAR C fparg:PBYTE
.
.
.
INVOKE proc1, ADDR arg1
For information on defining pointers with TYPEDEF, see “Defining Pointer

Types with TYPEDEF” in Chapter 3.

Invoking Procedures Indirectly

You can make an indirect procedure call such as call [bx + si] by using a
pointer to a function prototype with TYPEDEF, as shown in this example:
FUNCPROTO TYPEDEF PROTO NEAR ARG1:WORD
FUNCPTR TYPEDEF PTR FUNCPROTO
.DATA
pfunc FUNCPTR OFFSET proc1, OFFSET proc2
.CODE
.
.
.
mov bx, OFFSET pfunc ; BX points to table
mov si, Num ; Num contains 0 or 2
INVOKE FUNCPTR PTR [bx+si], arg1 ; Call proc1 if Num=0
; or proc2 if Num=2
You can also use ASSUME to accomplish the same task. The following
ASSUME statement associates the type FUNCPTR with the BX register.
ASSUME BX:FUNCPTR
mov bx, OFFSET pfunc
mov si, Num
INVOKE [bx+si], arg1
Checking the Code Generated

Code generated by the INVOKE directive may vary depending on the processor
mode and calling conventions in effect. You can check your listing files to see
the code generated by the INVOKE directive if you use the /Sg command-line
option.
Generating Prologue and Epilogue Code

When you use the PROC directive with its extended syntax and argument list,
the assembler automatically generates the prologue and epilogue code in your
procedure. “Prologue code” is generated at the start of the procedure. It sets up
a stack pointer so you can access parameters from within the procedure. It also
saves space on the stack for local variables, initializes registers such as DS, and
pushes registers that the procedure uses. Similarly, “epilogue code” is the code
at the end of the procedure that pops registers and returns from the procedure.

The assembler automatically generates the prologue code when it encounters the
first instruction or label after the PROC directive. This means you cannot label
the prologue for the purpose of jumping to it. The assembler generates the
epilogue code when it encounters a RET or IRET instruction. Using the
assembler-generated prologue and epilogue code saves time and decreases the
number of repetitive lines of code in your procedures.
The generated prologue or epilogue code depends on the:
u Local variables defined.
u Arguments passed to the procedure.
u Current processor selected (affects epilogue code only).
u Current calling convention.
u Options passed in the prologuearg of the PROC directive.
u Registers being saved.
The prologuearg list contains options specifying how to generate the prologue
or epilogue code. The next section explains how to use these options, gives the
standard prologue and epilogue code, and explains the techniques for defining
your own prologue and epilogue code.
Using Automatic Prologue and Epilogue Code

The standard prologue and epilogue code handles parameters and local
variables. If a procedure does not have any parameters or local variables, the
prologue and epilogue code that sets up and restores a stack pointer is omitted,
unless
FORCEFRAME is included in the prologuearg list. (FORCEFRAME is
discussed later in this section.) Prologue and epilogue code also generates a push
and pop for each register in the register list.
The prologue code consists of three steps:
1. Point BP to top of stack.
2. Make space on stack for local variables.
3. Save registers the procedure must preserve.

The epilogue cancels these three steps in reverse order, then cleans the stack, if
necessary, with a RET num instruction. For example, the procedure declaration
myproc PROC NEAR PASCAL USES di si,
arg1:WORD, arg2:WORD, arg3:WORD
LOCAL local1:WORD, local2:WORD
generates the following prologue code:

push bp ; Step 1:
mov bp, sp ; point BP to stack top
sub sp, 4 ; Step 2: space for 2 local words
push di ; Step 3:
push si ; save registers listed in USES
The corresponding epilogue code looks like this:

pop si ; Undo Step 3
pop di
mov sp, bp ; Undo Step 2
pop bp ; Undo Step 1
ret 6 ; Clean stack of pushed arguments
Notice the RET 6 instruction cleans the stack of the three word-sized
arguments. The instruction appears in the epilogue because the procedure does
not use the C calling convention. If myproc used C conventions, the epilogue
would end with a RET instruction without an operand.
The assembler generates standard epilogue code when it encounters a RET
instruction without an operand. It does not generate an epilogue if RET has a
nonzero operand. To suppress generation of a standard epilogue, use RETN or
RETF with or without an operand, or use RET 0.
The standard prologue and epilogue code recognizes two operands passed in the
prologuearg list, LOADDS and FORCEFRAME. These operands modify the
prologue code. Specifying LOADDS saves and initializes DS. Specifying
FORCEFRAME as an argument generates a stack frame even if no arguments
are sent to the procedure and no local variables are declared. If your procedure
has any parameters or locals, you do not need to specify FORCEFRAME.

For example, adding LOADDS to the argument list for myproc creates this
prologue:
push bp ; Step 1:
mov bp, sp ; point BP to stack top
sub sp, 4 ; Step 2: space for 2 locals
push ds ; Save DS and point it
mov ax, DGROUP ; to DGROUP, as
mov ds, ax ; instructed by LOADDS
push di ; Step 3:
push si ; save registers listed in USES
The epilogue code restores DS:

pop si ; Undo Step 3
pop di
pop ds ; Restore DS
mov sp, bp ; Undo Step 2
pop bp ; Undo Step 1
ret 6 ; Clean stack of pushed arguments
User-Defined Prologue and Epilogue Code

If you want a different set of instructions for prologue and epilogue code in your
procedures, you can write macros that run in place of the standard prologue and
epilogue code. For example, while you are debugging your procedures, you may
want to include a stack check or track the number of times a procedure is called.
You can write your own prologue code to do these things whenever a procedure
executes. Different prologue code may also be necessary if you are writing
applications for Windows. User-defined prologue macros will respond correctly
if you specify FORCEFRAME in the prologuearg of a procedure.
To write your own prologue or epilogue code, the OPTION directive must
appear in your program. It disables automatic prologue and epilogue code
generation. When you specify
OPTION PROLOGUE : macroname
OPTION EPILOGUE : macroname
the assembler calls the macro specified in the OPTION directive instead of
generating the standard prologue and epilogue code. The prologue macro must
be a macro function, and the epilogue macro must be a macro procedure.

The assembler expects your prologue or epilogue macro to have this form:
macroname MACRO procname, \
flag, \
parmbytes, \
localbytes, \
<reglist>, \
userparms
Your macro must have formal parameters to match all the actual arguments
passed. The arguments passed to your macro include:
procname The name of the procedure.
flag A 16-bit flag containing the following information:
Bit = Value Description
Bit 0, 1, 2 For calling conventions (000=unspecified language type,
001=C, 010=SYSCALL, 011=STDCALL,
100=PASCAL, 101=FORTRAN, 110=BASIC).
Bit 3 Undefined (not necessarily zero).
Bit 4 Set if the caller restores the stack
(use RET, not RETn).
Bit 5 Set if procedure is FAR.
Bit 6 Set if procedure is PRIVATE.
Bit 7 Set if procedure is EXPORT.
Bit 8 Set if the epilogue is generated as a result of an IRET
instruction and cleared if the epilogue is generated as a
result of a RET instruction.
Bits 9–15 Undefined (not necessarily zero).
parmbytes The accumulated count in bytes of all parameters given in the PROC
statement.
localbytes The count in bytes of all locals defined with the LOCAL directive.
reglist A list of the registers following the USES operator in the procedure
declaration. Enclose this list with angle brackets (< >) and separate each item
with commas. Reverse the list for epilogues.
userparms Any argument you want to pass to the macro. The prologuearg (if there is
one) specified in the PROC directive is passed to this argument.
Your macro function must return the parmbytes parameter. However, if the
prologue places other values on the stack after pushing BP and these values are
not referenced by any of the local variables, the exit value must be the number
of bytes for procedure locals plus any space between BP and the locals.
Therefore, parmbytes is not always equal to the bytes occupied by the locals.

The following macro is an example of a user-defined prologue that counts the

number of times a procedure is called.
ProfilePro MACRO procname, \
flag, \
bytecount, \
numlocals, \
regs, \
macroargs
.DATA
procname&count WORD 0
.CODE
inc procname&count ; Accumulates count of times the
; procedure is called
push bp
mov bp, sp
; Other BP operations
IFNB <regs>
FOR r, regs
push r
ENDM
ENDIF
EXITM %bytecount
ENDM
Your program must also include this statement before calling any procedures
that use the prologue:
OPTION PROLOGUE:ProfilePro
If you define either a prologue or an epilogue macro, the assembler uses the
standard prologue or epilogue code for the one you do not define. The form of
the code generated depends on the .MODEL and PROC options used.
If you want to revert to the standard prologue or epilogue code, use
PROLOGUEDEF or EPILOGUEDEF as the macroname in the OPTION
statement.
OPTION EPILOGUE:EPILOGUEDEF
You can completely suppress prologue or epilogue generation with

OPTION PROLOGUE:None
OPTION EPILOGUE:None
In this case, no user-defined macro is called, and the assembler does not
generate a default code sequence. This state remains in effect until the next
OPTION
PROLOGUE or OPTION EPILOGUE is encountered.

For additional information about writing macros, see Chapter 9, “Using

Macros.” The PROLOGUE.INC file provided in the MASM 6.1 distribution
disks can create the prologue and epilogue sequences for the Microsoft C
professional development system.
MS-DOS Interrupts
In addition to jumps, loops, and procedures that alter program execution,
interrupt routines transfer execution to a different location. In this case, control
goes to an interrupt routine.
You can write your own interrupt routines, either to replace an existing routine
or to use an undefined interrupt number. For example, you may want to replace
an MS-DOS interrupt handler, such as the Critical Error (Interrup 24h) and
CONTROL+C (Interrupt 23h) handlers. The BOUND instruction checks array
bounds and calls Interrupt 5 when an error occurs. If you use this instruction,
you need to write an interrupt handler for it.
This section summarizes the following:
u How to call interrupts
u How the processor handles interrupts
u How to redefine an existing interrupt routine
The example routine in this section handles addition or multiplication overflow

and illustrates the steps necessary for writing an interrupt routine. For additional
information about MS-DOS and BIOS interrupts, see Chapter 11, “Writing
Memory-Resident Software.”
Calling MS-DOS and ROM-BIOS Interrupts

Interrupts provide a way to access MS-DOS and ROM-BIOS from assembly
language. They are called with the INT instruction, which takes an immediate
value between 0 and 255 as its only operand.
MS-DOS and ROM-BIOS interrupt routines accept data through registers. For
instance, most MS-DOS routines (and many BIOS routines) require a function
number in the AH register. Many handler routines also return values in registers.
To use an interrupt, you must know what data the handler routine expects and
what data, if any, it returns. For information, consult Help or one of the other
references mentioned in the Introduction.

The following fragment illustrates a simple call to MS-DOS Function 9, which

displays the string msg on the screen:
.DATA
msg BYTE "This writes to the screen$"
.CODE
mov ax, SEG msg ; Necessary only if DS does not
mov ds, ax ; already point to data segment
mov dx, offset msg ; DS:DX points to msg
mov ah, 09h ; Request Function 9
int 21h
When the INT instruction executes, the processor:

1. Looks up the address of the interrupt routine in the Interrupt Vector Table.
This table starts at the lowest point in memory (segment 0, offset 0) and
consists of a series of far pointers called vectors. Each vector comprises a 4-
byte address (segment:offset) pointing to an interrupt handler routine. The
table sequence implies the number of the interrupt the vector references: the
first vector points to the Interrupt 0 handler, the second vector to the
Interrupt 1 handler, and so forth. Thus, the vector at 0000:i*4 holds the
address of the handler routine for Interrupt i.
2. Clears the trap flag (TF) and interrupt enable flag (IF).
3. Pushes the flags register, the current code segment (CS), and the current
instruction pointer (IP), in that order. (The current instruction is the one
following the INT statement.) As with a CALL, this ensures control returns
to the next logical position in the program.
4. Jumps to the address of the interrupt routine, as specified in the Interrupt
Vector Table.
5. Executes the code of the interrupt routine until it encounters an IRET
instruction.
6. Pops the instruction pointer, code segment, and flags.

Figure 7.3 illustrates how interrupts work.
Figure 7.3 Operation of Interrupts
Replacing an Interrupt Routine

To replace an existing interrupt routine, your program must:
u Provide a new routine to handle the interrupt.
u Replace the old routine’s address in the Interrupt Vector Table with the
address of your new routine.
u Replace the old address back into the vector table before your program ends.
You can write an interrupt routine as a procedure by using the PROC and
ENDP directives. The routine should always be defined as FAR and should end
with an IRET instruction instead of a RET instruction.

Note You can use the full extended PROC syntax (described in “Declaring
Parameters with the PROC Directive,” earlier in this chapter) to write interrupt
procedures. However, you should not make interrupt procedures NEAR or
specify arguments for them. You can use the USES keyword, however, to
correctly generate code to save and restore a register list in interrupt procedures.
The IRET instruction in MASM 6.1 has two forms that suppress epilogue code.
This allows an interrupt to have local variables or use a user-defined prologue.
IRETF pops a FAR16 return address, and IRETFD pops a FAR32 return
address.
The following example shows how to replace the handler for Interrupt 4. Once
registered in the Interrupt Vector Table, the new routine takes control when the
processor encounters either an INT 4 instruction or its special variation INTO
(Interrupt on Overflow). INTO is a conditional instruction that acts only when
the overflow flag is set. With INTO after a numerical calculation, your code can
automatically route control to a handler routine if the calculation results in a
numerical overflow. By default, the routine for Interrupt 4 simply consists of an
IRET, so it returns without doing anything. Using INTO is an alternative to
using JO (Jump on Overflow) to jump to another set of instructions.
The following example program first executes INT 21h to invoke MS-DOS
Function 35h (Get Interrupt Vector). This function returns the existing vector
for Interrupt 4. The program stores the vector, then invokes MS-DOS Function
25h (Set Interrupt Vector) to place the address of the ovrflow procedure in the
Interrupt Vector Table. From this point on, ovrflow gains control whenever
the processor executes INTO while the overflow flag is set. The new routine
displays a message and returns with AX and DX set to 0.
.MODEL LARGE, C
FPFUNC TYPEDEF FAR PTR
.DATA
msg BYTE "Overflow - result set to 0",13,10,'$'
vector FPFUNC ?
.CODE
.STARTUP
mov ax, 3504h ; Load Interrupt 4 and call DOS

int 21h ; Get Interrupt Vector
mov WORD PTR vector[2],es ; Save segment
mov WORD PTR vector[0],bx ; and offset

push ds ; Save DS
mov ax, cs ; Load segment of new routine
mov ds, ax
mov dx, OFFSET ovrflow ; Load offset of new routine
mov ax, 2504h ; Load Interrupt 4 and call DOS
int 21h ; Set Interrupt Vector
pop ds ; Restore
.
.
.
add ax, bx ; Do arithmetic
into ; Call Interrupt 4 if overflow
.
.
.
lds dx, vector ; Load original address
mov ax, 2504h ; Restore it to vector table
int 21h ; with DOS set vector function
mov ax, 4C00h ; Terminate function
int 21h
ovrflow PROC FAR

sti ; Enable interrupts
; (turned off by INT)
mov ah, 09h ; Display string function
mov dx, OFFSET msg ; Load address
int 21h ; Call DOS
sub ax, ax ; Set AX to 0
cwd ; Set DX to 0
iret ; Return
ovrflow ENDP
END
Before the program ends, it again uses MS-DOS Function 25h to reset the
original Interrupt 4 vector back into the Interrupt Vector Table. This
reestablishes the original routine as the handler for Interrupt 4.
The first instruction of the ovrflow routine warrants further discussion. When
the processor encounters an INT instruction, it clears the interrupt flag before
branching to the specified interrupt handler routine. The interrupt flag serves a
crucial role in smoothing the processor’s tasks, but must not be abused. When
clear, the flag inhibits hardware interrupts such as the keyboard or system timer.
It should be left clear only briefly and only when absolutely necessary. Unless
you have a

compelling reason to leave the flag clear, always include an STI (Set Interrupt
Flag) instruction at the beginning of your interrupt handler routine to reenable
hardware interrupts.
CLI (Clear Interrupt Flag) and its corollary STI are designed to protect small
sections of time-dependent code from interruptions by the hardware. If you use
CLI in your program, be sure to include a matching STI instruction as well. The
sample interrupt handlers in Chapter 11, “Writing Memory-Resident Software,”
illustrate how to use these important instructions.


211
C H A P T E R 8
Sharing Data and Procedures

Among Modules and Libraries
To use symbols and procedures in more than one module, the assembler must
be able to recognize the shared data as global to all the modules where they are
used. MASM provides techniques to simplify data-sharing and give a high-level
interface to multiple-module programming. With these techniques, you can place
shared symbols in include files. This makes the data declarations in the file
available to all modules that use the include file.
This chapter explains the two data-sharing methods MASM 6.1 offers. The first
method simplifies data sharing between modules with include files. The second
does not involve include files. Instead, this method allows modules to share
procedures and data items using the PUBLIC and EXTERN directives.
The last section of this chapter explains how to create program libraries and
access their routines.
Selecting Data-Sharing Methods

If data defined in one module is to be used in other modules of a program, you
must declare the data public and external. MASM provides several ways to do
this:
u Declare a symbol public with the PUBLIC directive in the module where it is
defined. This makes the symbol available to other modules. You must also
place an EXTERN statement for that symbol in all other modules that refer
to the public symbol. This statement informs the assembler that the symbol is
external — that is, defined in another module.
u Declare the data communal with the COMM directive. However, communal
variables have limitations. You cannot depend on their location in memory
because they are allocated by the linker, and they cannot be initialized.
The EXTERNDEF directive declares a symbol either public or external, as
appropriate. EXTERNDEF simplifies the declarations for global (public and
external) variables and encourages the use of include files.

The next section provides further details on using include files. For more
information on PUBLIC and EXTERN, see “Using Alternatives to Include
Files,” page 219.
Sharing Symbols with Include Files

Include files can contain any valid MASM statement, but typically consist of
type and symbol declarations. The assembler inserts the contents of the include
file into a module at the location of the INCLUDE directive. Include files are
optional, but can simplify project organization by eliminating the need to insert
common declarations into all modules of a program. An alternative to using
include files is described in “Using Alternatives to Include Files,” page 219.
This section explains how to organize symbol definitions and the declarations
that make them global (available to all modules); how to make both variables
and procedures public with EXTERNDEF, PROTO, and COMM.; and where
to place these directives in the modules and include files.
Organizing Modules
This section summarizes the organization of declarations and definitions in
modules and include files and the use of the INCLUDE directive.
Include Files
Type declarations that need to be identical in every module should be placed in
an include file. This ensures consistency and saves time when you update
programs. Include files should contain only symbol declarations and any other
declarations that are resolved at assembly time. (For a list of assembly-time
operations, see “Generating and Running Executable Programs” in Chapter 1.)
If more than one module accesses the include file, the file cannot contain
statements that define and allocate memory for symbols. Otherwise, the
assembler would attempt to allocate the same symbol more than once.
Note An include file used in two or more modules should not allocate data
variables.
Modules
An INCLUDE statement is usually placed before data and code segments in
your modules. When the assembler encounters an INCLUDE directive, it opens
the specified file and assembles all its statements. The assembler then returns to
the original module and continues the assembly.

Chapter 8 Sharing Data and Procedures Among Modules and Libraries 213
The INCLUDE directive takes the form:

INCLUDE filename
where filename is the full name of the include file. For example, the following
declaration inserts the contents of the include file SCREEN.INC in your
program:
INCLUDE SCREEN.INC
The filename in the INCLUDE directive must be fully specified; no extensions

are assumed. If a full pathname is not given, the assembler first searches the
directory of the source file containing the INCLUDE directive.
If the include file is not in the source file directory, the assembler searches the
paths specified in the assembler’s command-line option /I, or in PWB’s Include
Paths field in the MASM Option dialog box (accessed from the Option menu).
The /I option takes this form:
/I path
You can include more than one /I option on the command line. The assembler
then searches for include files within each specified path in the order given. If
none of these directories contains the include file, the assembler finally searches
in the paths specified in the INCLUDE environment variable. If the include file
still cannot be found, an assembly error occurs. (The /x command-line option
tells the assembler to ignore the INCLUDE environment variable when
searching for include files.)
An include file may specify another include file. The assembler processes the
second include file before returning to the first. Your program can nest include
files this way as deeply as the amount of free memory allows.
Include Files or Modules

You can use the EQU directive to create named constants that cannot be
redefined in your program. (For information about the EQU directive, see
“Integer Constants and Constant Expressions,” page 11.) Placing a constant
defined with EQU in an include file makes it available to all modules that use
that include file.
Placing TYPEDEF, STRUCT, UNION, and RECORD definitions in an include
file guarantees consistency in type definitions. If required, the variable instances
derived from these definitions can be made public among the modules with
EXTERNDEF declarations (see the next section). Macros, including macros
defined with TEXTEQU, must be placed in include files to make them visible in
other modules.

If you elect to use full segment definitions with, or instead of, simplified
definitions, you can force a consistent segment order in all files by defining
segments in an include file. This technique is explained in “Controlling the
Segment Order,”
page 47.
Declaring Symbols Public and External

It is sometimes useful to make certain procedures and variables (such as status
flags) global to all program modules. Global variables are freely accessible within
all routines; you do not have to explicitly pass them to the routines that need
them. This section describes how to make variables and procedures global using
the EXTERNDEF, PROTO, or COMM declarations within include files.
When a procedure is defined in one module and called in another module, it
must be declared public in the defining module and external in the calling
module(s). MASM offers three ways to declare a procedure public and external:
u Use the PUBLIC directive in the defining module and EXTERN in all other
modules that reference the procedure. The PUBLIC and EXTERN
directives are explained on page 220.
u Declare the procedure with EXTERNDEF.
u Prototype the procedure with the PROTO directive.
Using EXTERNDEF
MASM treats EXTERNDEF as a public declaration in the defining module, and
as an external declaration in the referencing module(s). You can use the
EXTERNDEF statement in your include file to make a variable common to two
or more modules. EXTERNDEF works with all types of variables, including
arrays, structures, unions, and records. It also works with procedures.
As a result, a single include file can contain an EXTERNDEF declaration that
works in both the defining module and any referencing module. It is ignored in
modules that neither define nor reference the variable. Therefore, an include file
for a library which is used in multiple .EXE files does not force the definition of
a symbol as EXTERN does.
The EXTERNDEF statement takes this form:
EXTERNDEF [[langtype]] name:qualifiedtype
The name is the variable’s identifier. The qualifiedtype is explained in detail in
“Data Types,” page 14.
The optional langtype specifier sets the naming conventions for the name it
precedes. It overrides any language specified in the .MODEL directive. The
specifier can be C, SYSCALL, STDCALL, PASCAL, FORTRAN, or

BASIC. For information on selecting the appropriate langtype type, see

“Naming and Calling Conventions,” page 308.

The following diagram shows the statements that declare an array, make it
public, and use it in another module.
Figure 8.1 Using EXTERNDEF for Variables
The file position of EXTERNDEF directives is important. For more information,

see “Positioning External Declarations,” following.
You can also make procedures visible by using EXTERNDEF without PROTO
inside an include file. This method treats the procedure name as a simple
identifier, without the parameter list, so you forgo the assembler’s ability to
check for the correct parameters during assembly. Use EXTERNDEF with
procedures in the same way as variables:
EXTERNDEF MyProc:FAR ; Declare far procedure external
You can also use EXTERNDEF to make a code label global between modules
so that one module can reference a label in another module. Give the label
global scope with the double colon operator, like this:
EXTERNDEF codelabel:NEAR
.
.
.
codelabel::
Another module can reference codelabel like this:

EXTERNDEF codelabel:NEAR
.
.
.
jmp codelabel

Using PROTO
This section describes how to prototype a procedure with the PROTO directive.
PROTO automatically issues an EXTERNDEF for the procedure unless the
PROC statement declares the procedure PRIVATE. Defining a prototype
enables type-checking for the procedure arguments.
Follow these steps to create an interface for a procedure defined in one module
and called from other modules:
1. Place the PROTO declaration in the include file.
2. Define the procedure with PROC in one module. The PROC directive
declares the procedure PUBLIC by default.
3. Call the procedure with the INVOKE statement (or with CALL). Make sure
that all calling modules access the include file.
For descriptions, syntax, and examples of PROTO, PROC, and INVOKE, see
Chapter 7, “Controlling Program Flow.”
The following example illustrates these three steps. In the example, a PROTO
statement defines the far procedure CopyFile, which uses the C parameter-
passing and naming conventions, and takes the arguments filename and
numberlines. The diagram following the example shows the file placement
for these statements.
This definition goes into the include file:
CopyFile PROTO FAR C filename:BYTE, numberlines:WORD
The procedure definition for CopyFile is:

CopyFile PROC FAR C USES cx, filename:BYTE, numberlines:WORD
To call the CopyFile procedure, you can use this INVOKE statement:
INVOKE CopyFile, NameVar, 200

Figure 8.2 Using PROTO and INVOKE
Using COMM
Another way to share variables among modules is to add the COMM
(communal) declaration to your include file. Since communal variables are
allocated by the linker and cannot be initialized, you cannot depend on their
location or sequence.
Communal variables are supported by MASM primarily for compatibility with
communal variables in Microsoft C. Communal variables are not used in any
other Microsoft language, and they are not compatible with C++ and some other
languages.
COMM declares a data variable external and instructs the linker to allocate the
variable if it has not been explicitly defined in a module. The memory space for
communal variables may not be assigned until load time, so using communal
variables may reduce the size of your executable file.
The COMM declaration has the syntax:
COMM [[langtype]] [[NEAR | FAR]] label:type[[:count]]
The label is the name of the variable. The langtype sets the naming conventions
for the name it precedes. It overrides any language specified in the .MODEL
directive.

If NEAR or FAR is not specified, the variable determines the default from the
current memory model (NEAR for TINY, SMALL, COMPACT, and FLAT;
FAR for MEDIUM, LARGE, and HUGE). If you do not provide a memory
model with the .MODEL directive, you must specify a distance when accessing
a communal variable, like this:
mov ax, NEAR PTR CommNear
mov bx, FAR PTR CommFar
The type can be a constant expression, but it is usually a type such as BYTE,
WORD, or DWORD, or a structure, union, or record. If you first declare the
type with TYPEDEF, CodeView can provide type information. The count is the
number of elements. If no count is given, one element is assumed.
The following example creates the on far variable DataBlock, which is a
1,024-element array of uninitialized signed doublewords:
COMM FAR DataBlock:SDWORD:1024
Note C variables declared outside functions (except static variables) are

communal unless explicitly initialized; they are the same as assembly-language
communal variables. If you are writing assembly-language modules for C, you
can declare the same communal variables in both C and MASM include files.
However, communal variables in C do not have to be declared communal in
assembler. The linker will match the EXTERN, PUBLIC, and COMM
statements for the variable.
EXTERNDEF (explained in the previous section) is more flexible than COMM

because you can initialize variables defined with it, and your code can rely on
the position and sequence of the defined data.
Positioning External Declarations

Although LINK determines the actual address of an external symbol, the
assembler assumes a default segment for the symbol, based on the location of
the external directive in the source code. You should therefore position
EXTERN and
EXTERNDEF directives according to these rules:
u If you know which segment defines an external symbol, put the EXTERN
statement in that segment.

u If you know the group but not the segment, position the EXTERN statement
outside any segment and reference the variable with the group name. For
example, if var1 is in DGROUP, reference the variable as
mov DGROUP:var1, 10
u If you know nothing about the location of an external variable, put the
EXTERN statement outside any segment. You can use the SEG directive to
access the external variable like this:
mov ax, SEG var1
mov es, ax
mov ax, es:var1
u If the symbol is an absolute symbol or a far code label, you can declare it
external anywhere in the source code.
Always close any segments opened in include files so that external declarations
following an include statement are not incorrectly placed inside a segment. If
you want to be certain an external definition lies outside a segment, you can use
@CurSeg. The @CurSeg predefined symbol returns a blank if the definition is
not in a segment. For example,
.DATA
.
.
.
@CurSeg ENDS ; Close segment
EXTERNDEF var:WORD
For information about predefined symbols such as @CurSeg, see “Predefined

Symbols,” page 10.
Using Alternatives to Include Files

If your project uses only two modules (or if it is written with a version of
MASM prior to 6.0), you may want to continue using PUBLIC in the defining
module and EXTERN in the referencing module, and not create an include file
for the project. The EXTERN directive can be used in an include file, but the
include file containing EXTERN cannot be added to the module that contains
the corresponding PUBLIC directive for that symbol. This section assumes that
you are not using include files.

PUBLIC and EXTERN

The PUBLIC and EXTERN directives are less flexible than EXTERNDEF and
PROTO because they are module-specific: PUBLIC must appear in the
defining module and EXTERN must appear in the calling modules. This section
shows how to use PUBLIC and EXTERN. Information on where to place the
external declarations in your file is in “Positioning External Declarations,”
previous.
The PUBLIC directive makes a name visible outside the module in which it is
defined. This gives other program modules access to that identifier.
The EXTERN directive performs the complementary function. It tells the
assembler that a name referenced within a particular module is actually defined
and declared public in another module that will be specified at link time.
A PUBLIC directive can appear anywhere in a file. Its syntax is:
PUBLIC [[langtype]] name[[, [[langtype]] name]]...
The name must be the name of an identifier defined within the current source
file. Only code labels, data labels, procedures, and numeric equates can be
declared public.
If you specify the langtype field here, it overrides the language specified by
.MODEL. The langtype field can be C, SYSCALL, STDCALL, PASCAL,
FORTRAN, or BASIC. For more information on specifying langtype types, see
“Declaring Parameters with the PROC Directive,” page 184, and “Naming and
Calling Conventions,” page 308.
The EXTERN directive tells the assembler that an identifier is external —
defined in some other module that will be supplied at link time. Its syntax is:
EXTERN [[langtype]] name:{ABS | qualifiedtype}
“Data Types,” page 14, describes qualifiedtype. You can use the ABS
(absolute) keyword only with external numeric constants. ABS causes the
identifier to be imported as a relocatable unsized constant. This identifier can
then be used anywhere a constant can be used. If the identifier is not found in
another module at link time, the linker generates an error.

In the following example, the procedure BuildTable and the variable Var are
declared public. The procedure uses the Pascal naming and data-passing
conventions:
Figure 8.3 Using PUBLIC and EXTERN
Other Alternatives
You can also use the directives discussed earlier (EXTERNDEF, PROTO, and
COMM) without the include file. In this case, place the declarations to make a
symbol global in the same module where the symbol is defined. You might want
to use this technique if you are linking only a few modules that have very little
data in common.
Developing Libraries
As you create reusable procedures, you can place them in a library file for
convenient access. Although you can put any routine into a library, each library
file, recognizable by its .LIB extension, usually contains related routines. For
example, you might place string-manipulation functions in one library, matrix
calculations in another, and port communications in another. Do not place
communal variables (defined with the COMM directive) in a library.
A library consists of combined object modules, each created from a single
source file. The object module is the smallest independent unit in a library. If
you link with one symbol in a module, the linker adds the entire module to your
program, but not the entire library.

Associating Libraries with Modules

You can choose either of two methods for associating your libraries with the
modules that use them: you can use the INCLUDELIB directive inside your
source files, or link the modules from the command line.
To associate a specified library with your object code, use INCLUDELIB. You
can add this directive to the source file to specify the libraries you want linked,
rather than specifying them in the LINK command line. The INCLUDELIB
syntax is:
INCLUDELIB libraryname
The libraryname can be a file name or a complete path specification. If you do
not specify an extension, .LIB is assumed. The libraryname is placed in the
comment record of the object file. LINK reads this record and links with the
specified library file.
For example, the statement INCLUDELIB GRAPHICS passes a message from
the assembler to the linker telling LINK to use library routines from the file
GRAPHICS.LIB. If you place this statement in the source file DRAW.ASM and
GRAPHICS.LIB is in the same directory, you can assemble and link the
program with the following command:
ML DRAW.ASM
Without the INCLUDELIB directive, you must link the program DRAW.ASM
with either of the following commands:
ML DRAW.ASM GRAPHICS.LIB
ML DRAW /link GRAPHICS
If you want to assemble and link separately, type

ML /c DRAW.ASM
LINK DRAW,,,GRAPHICS
If you do not specify a complete path in the INCLUDELIB statement or at the

command line, LINK searches for the library file in the following order:
1. In the current directory.
2. In any directories in the library field of the LINK command line.
3. In any directories specified by the LIB environment variable.
The LIB.EXE utility helps you create, organize, and maintain run-time libraries.
Refer to Environment and Tools for instructions on LIB.EXE.

Using EXTERN with Library Routines

In some cases, EXTERN helps you limit the size of your executable file by
specifying in the syntax an alternative name for a procedure. You would use this
form of the EXTERN directive when declaring a procedure or symbol that may
not need to be used.
The syntax looks like this:
EXTERN [[langtype]] name [[ (altname) ]] :qualifiedtype
The addition of the altname to the syntax provides the name of an alternate
procedure that the linker uses to resolve the external reference if the procedure
given by name is not needed. Both name and altname must have the same
qualifiedtype.
When the linker encounters an external definition for a procedure that gives an
altname, the linker finishes processing that module before it links the object
module that contains the procedure given by name. If the program does not
reference any symbols in the name file’s object from any of the linked modules,
the linker uses altname to satisfy the external reference. This saves space
because the library object module is not brought in.
For example, assume that the contents of STARTUP.ASM include these
statements:
EXTERN init(dummy):PROC
.
.
.
dummy PROC
.
.
. ; A procedure definition containing no
ret ; executable code
dummy ENDP
.
.
.
call init ; Defined in FLOAT.OBJ
In this example, the reference to the routine init (defined in FLOAT.OBJ)

does not force the module FLOAT.OBJ to be linked into the executable file. If
another reference causes FLOAT.OBJ to be linked into the executable file, then
init will refer to the init label in FLOAT.OBJ. If there are no references that
force linkage with FLOAT.OBJ, the linker will use the alternate name for
init(dummy).


225
C H A P T E R 9
Using Macros
A “macro” is a symbolic name you give to a series of characters (a text macro)

or to one or more statements (a macro procedure or function). As the assembler
evaluates each line of your program, it scans the source code for names of
previously defined macros. When it finds one, it substitutes the macro text for
the macro name. In this way, you can avoid writing the same code several
places in your program.
This chapter describes the following types of macros:
u Text macros, which expand to text within a source statement.
u Macro procedures, which expand to one or more complete statements and
can optionally take parameters.
u Repeat blocks, which generate a group of statements a specified number of
times or until a specified condition becomes true.
u Macro functions, which look like macro procedures and can be used like text
macros but which also return a value.
u Predefined macro functions and string directives, which perform string
operations.
This chapter explains how to use macros for simple code substitutions and how
to write sophisticated macros with parameter lists and repeat loops. It also
describes how to use these features in conjunction with local symbols, macro
operators, and predefined macro functions.

Template: MSGRIDA1.DOT Author: rick debroux Last Saved By: Ruth L Silverio
Text Macros
You can give a sequence of characters a symbolic name and then use the name
in place of the text later in the source code. The named text is called a text
macro.
The TEXTEQU directive defines a text macro, as these examples show:
name TEXTEQU <text>
name TEXTEQU macroId | textmacro
name TEXTEQU %constExpr
In the previous lines, text is a sequence of characters enclosed in angle brackets,
macroId is a previously defined macro function, textmacro is a previously
defined text macro, and %constExpr is an expression that evaluates to text.
Here are some examples:
msg TEXTEQU <Some text> ; Text assigned to symbol
string TEXTEQU msg ; Text macro assigned to symbol
msg TEXTEQU <Some other text> ; New text assigned to symbol
value TEXTEQU %(3 + num) ; Text representation of resolved
; expression assigned to symbol
The first line assigns text to the symbol msg. The second line equates the text of
the msg text macro with a new text macro called string. The third line assigns
new text to msg. Although msg has new text, string retains its original text
value. The fourth line assigns 7 to value if num equals 4. If a text macro
expands to another text macro (or macro function, as discussed on page 248),
the resulting text macro will expand recursively.
Text macros are useful for naming strings of text that do not evaluate to
integers. For example, you might use a text macro to name a floating-point
constant or a bracketed expression. Here are some practical examples:
pi TEXTEQU <3.1416> ; Floating point constant
WPT TEXTEQU <WORD PTR> ; Sequence of key words
arg1 TEXTEQU <[bp+4]> ; Bracketed expression
Macro Procedures
If your program must perform the same task many times, you can avoid
repeatedly typing the same statements each time by writing a macro procedure.
Think of macro procedures (commonly called macros) as text-processing
mechanisms that automatically generate repeated text.

Chapter 9 Using Macros 227
This section uses the term “macro procedure” rather than “macro” when
necessary to distinguish between a macro procedure and a macro function.
Macro functions are described in “Returning Values with Macro Functions.”
Conforming to common usage, this chapter occasionally speaks of “calling” a
macro, a term that deserves further scrutiny. It’s natural to think of a program
calling a macro procedure in the same way it calls a normal subroutine
procedure, because they seem to perform identically. However, a macro is
simply a representative for real code. Wherever a macro name appears in your
program, so in reality does all the code the macro represents. A macro does not
cause the processor to vector off to a new location as does a normal procedure.
Thus, the expression “calling a macro” may imply the effect, but does not
accurately describe what actually occurs.
Creating Macro Procedures

You can define a macro procedure without parameters by placing the desired
statements between the MACRO and ENDM directives:
name MACRO
statements
ENDM
For example, suppose you want a program to beep when it encounters certain
errors. You could define a beep macro as follows:
beep MACRO
mov ah, 2 ;; Select DOS Print Char function
mov dl, 7 ;; Select ASCII 7 (bell)
int 21h ;; Call DOS
ENDM
The double semicolons mark the beginning of macro comments. Macro

comments appear in a listing file only at the macro’s initial definition, not at the
point where the macro is referenced and expanded. Listings are usually easier to
read if the comments aren’t repeatedly expanded. However, regular comments
(those with a single semicolon) are listed in macro expansions. See Appendix C
for listing files and examples of how macros are expanded in listings.
Once you define a macro, you can call it anywhere in the program by using the
macro’s name as a statement. The following example calls the beep macro two
times if an error flag has been set.
.IF error ; If error flag is true
beep ; execute macro two times
beep
.ENDIF

During assembly, the instructions in the macro replace the macro reference. The
listing file shows:
.IF error
0017 80 3E 0000 R 00 * cmp error, 000h
001C 74 0C * je @C0001
beep
001E B4 02 1 mov ah, 2
0020 B2 07 1 mov dl, 7
0022 CD 21 1 int 21h
beep
0024 B4 02 1 mov ah, 2
0026 B2 07 1 mov dl, 7
0028 CD 21 1 int 21h
.ENDIF
002A *@C0001:
Contrast this with the results of defining beep as a procedure using the PROC
directive and then calling it with the CALL instruction.
Many such tasks can be handled as either a macro or a procedure. In deciding
which method to use, you must choose between speed and size. For repetitive
tasks, a procedure produces smaller code, because the instructions physically
appear only once in the assembled program. However, each call to the
procedure involves the additional overhead of a CALL and RET instruction.
Macros do not require a change in program flow and so execute faster, but
generate the same code multiple times rather than just once.
Passing Arguments to Macros

By defining parameters for macros, you can define a general task and then
execute variations of it by passing different arguments each time you call the
macro. The complete syntax for a macro procedure includes a parameter list:
name MACRO parameterlist
statements
ENDM
The parameterlist can contain any number of parameters. Use commas to
separate each parameter in the list. You cannot use reserved words as parameter
names unless you disable the keyword with OPTION NOKEYWORD. You
must also set the compatibility mode with OPTION M510 or the /Zm
command-line option.
To pass arguments to a macro, place the arguments after the macro name when
you call the macro:
macroname arglist

The assembler treats as one item all text between matching quotation marks in
an arglist.
The beep macro introduced in the previous section used the MS-DOS interrupt
to write only the bell character (ASCII 7). We can rewrite the macro with a
parameter that accepts any character:
writechar MACRO char
mov dl, char ;; Select ASCII char
int 21h ;; Call DOS
ENDM
Whenever it expands the macro, the assembler replaces each instance of char
with the given argument value. The rewritten macro now writes any character to
the screen, not just ASCII 7:
writechar 7 ; Causes computer to beep
writechar ‘A’ ; Writes A to screen
If you pass more arguments than there are parameters, the additional arguments
generate a warning (unless you use the VARARG keyword; see page 242). If
you pass fewer arguments than the macro procedure expects, the assembler
assigns empty strings to the remaining parameters (unless you have specified
default values). This may cause errors. For example, a reference to the
writechar macro with no argument results in the following line:
mov dl,
The assembler generates an error for the expanded statement but not for the
macro definition or the macro call.
You can make macros more flexible by leaving off arguments or adding
additional arguments. The next section tells some of the ways your macros can
handle missing or extra arguments.
Specifying Required and Default Parameters

Macro parameters can have special attributes to make them more flexible and
improve error handling. You can make parameters required, give them default
values, or vary their number. Variable parameters are used almost exclusively
with the FOR directive, so are covered in “FOR Loops and Variable-Length
Parameters,” later in this chapter.

The syntax for a required parameter is:

parameter:REQ
For example, you can rewrite the writechar macro to require the char
parameter:
writechar MACRO char:REQ
int 21h ;; Call DOS
ENDM
If the call does not include a matching argument, the assembler reports the error
in the line that contains the macro reference. REQ can thus improve error
reporting.
You can also accommodate missing parameters by specifying a default value,
like this:
parameter:=textvalue
Suppose that you often use writechar to beep by printing ASCII 7. The
following macro definition uses an equal sign to tell the assembler to assume the
parameter char is 7 unless you specify otherwise:
writechar MACRO char:=<7>
int 21h ;; Call DOS
ENDM
If a reference to this macro does not include the argument char, the assembler
fills in the blank with the default value of 7 and the macro beeps when called.
Enclose the default parameter value in angle brackets so the assembler
recognizes the supplied value as a text value. This is explained in detail in “Text
Delimiters and the Literal-Character Operator,” later in this chapter.
Missing arguments can also be handled with the IFB, IFNB, .ERRB, and
.ERRNB directives. They are described in the section “Conditional Directives”
in chapter 1 and in Help. Here is a slightly more complex macro that uses some
of these techniques:

Scroll MACRO distance:REQ, attrib:=<7>, tcol, trow, bcol, brow

IFNB <tcol> ;; Ignore arguments if blank
mov cl, tcol
ENDIF
IFNB <trow>
mov ch, trow
ENDIF
IFNB <bcol>
mov dl, bcol
ENDIF
IFNB <brow>
mov dh, brow
ENDIF
IFDIFI <attrib>, <bh> ;; Don’t move BH onto itself
mov bh, attrib
ENDIF
IF distance LE 0 ;; Negative scrolls up, positive down
mov ax, 0600h + (-(distance) AND 0FFh)
ELSE
mov ax, 0700h + (distance AND 0FFh)
ENDIF
int 10h
ENDM
In this macro, the distance parameter is required. The attrib parameter has
a default value of 7 (white on black), but the macro also tests to make sure the
corresponding argument isn’t BH, since it would be inefficient (though legal) to
load a register onto itself. The IFNB directive is used to test for blank
arguments. These are ignored to allow the user to manipulate rows and columns
directly in registers CX and DX at run time.
The following shows two valid ways to call the macro:
; Assume DL and CL already loaded
dec dh ; Decrement top row
inc ch ; Increment bottom row
Scroll -3 ; Scroll white on black dynamic
; window up three lines
Scroll 5, 17h, 2, 2, 14, 12 ; Scroll white on blue constant
; window down five lines
This macro can generate completely different code, depending on its arguments.
In this sense, it is not comparable to a procedure, which always has the same
code regardless of arguments.

Defining Local Symbols in Macros

You can make a symbol local to a macro by identifying it at the start of the
macro with the LOCAL directive. Any identifier may be declared local.
You can choose whether you want numeric equates and text macros to be local
or global. If a symbol will be used only inside a particular macro, you can
declare it local so that the name will be available for other declarations outside
the macro.
You must declare as local any labels within a macro, since a label can occur
only once in the source. The LOCAL directive makes a special instance of the
label each time the macro appears. This prevents redefinition of the label when
expanding the macro. It also allows you to reuse the label elsewhere in your
code.
You must declare all local symbols immediately following the MACRO
statement (although blank lines and comments may precede the local symbol).
Separate each symbol with a comma. You can attach comments to the LOCAL
statement and list multiple LOCAL statements in the macro. Here is an example
macro that declares local labels:
power MACRO factor:REQ, exponent:REQ
LOCAL again, gotzero ;; Local symbols
sub dx, dx ;; Clear top
mov ax, 1 ;; Multiply by one on first loop
mov cx, exponent ;; Load count
jcxz gotzero ;; Done if zero exponent
mov bx, factor ;; Load factor
again:
mul bx ;; Multiply factor times exponent
loop again ;; Result in AX
gotzero:
ENDM
If the labels again and gotzero were not declared local, the macro would
work the first time it is called, but it would generate redefinition errors on
subsequent calls. MASM implements local labels by generating different names
for them each time the macro is called. You can see this in listing files. The
labels in the power macro might be expanded to ??0000 and ??0001 on the
first call and to ??0002 and ??0003 on the second.

You should avoid using anonymous labels in macros (see “Anonymous Labels”
in Chapter 7). Although legal, they can produce unwanted results if you expand
a macro near another anonymous label. For example, consider what happens in
the following:
Update MACRO arg1
@@: .
.
.
loop @B
ENDM
.
.
.
jcxz @F
Update ax
@@:
Expanding Update places another anonymous label between the jump and its
target. The line
jcxz @F
consequently jumps to the start of the loop rather than over the loop — exactly
the opposite of what the programmer intended.
Assembly-Time Variables and Macro Operators

In writing macros, you will often assign and modify values assigned to symbols.
Think of these symbols as assembly-time variables. Like memory variables,
they are symbols that represent values. But since macros are processed at
assembly time, any symbol modified in a macro must be resolved as a constant
by the end of assembly.
The three kinds of assembly-time variables are:
u Macro parameters
u Text macros
u Macro functions
When the assembler expands a macro, it processes the symbols in the order
shown here. MASM first replaces macro parameters with the text of their actual
arguments, then expands text macros.

Macro parameters are similar to procedure parameters in some ways, but they
also have important differences. In a procedure, a parameter has a type and a
memory location. Its value can be modified within the procedure. In a macro, a
parameter is a placeholder for the argument text. The value can only be assigned
to another symbol or used directly; it cannot be modified. The macro may
interpret the argument text it receives either as a numeric value or as a text
value.
It is important to understand the difference between text values and numeric
values. Numeric values can be processed with arithmetic operators and assigned
to numeric equates. Text values can be processed with macro functions and
assigned to text macros.
Macro operators are often helpful when processing assembly-time variables.
Table 9.1 shows the macro operators that MASM provides.
Table 9.1 MASM Macro Operators
Symbol Name Description
<> Text Delimiters Opens and closes a literal string.
! Literal-Character Operator Treats the next character as a literal character,
even if it would normally have another meaning.
% Expansion Operator Causes the assembler to expand a constant
expression or text macro.
& Substitution Operator Tells the assembler to replace a macro
parameter or text macro name with its
actual value.
The next sections explain these operators in detail.
Text Delimiters and the Literal-Character Operator

The angle brackets (< >) are text delimiters. A text value is usually delimited
when assigning a text macro. You can do this with TEXTEQU, as previously
shown, or with the SUBSTR and CATSTR directives discussed in “String
Directives and Predefined Functions,” later in this chapter.
By delimiting the text of macro arguments, you can pass text that includes
spaces, commas, semicolons, and other special characters. The following
example expands a macro called work in two different ways:
work <1, 2, 3, 4, 5> ; Passes one argument with 13 chars,
; including commas and spaces
work 1, 2, 3, 4, 5 ; Passes five arguments, each
; with 1 character

The literal-character operator (!) lets you include angle brackets as part of a
delimited text value, so the assembler does not interpret them as delimiters. The
assembler treats the character following ! literally rather than as a special
character, like this:
errstr TEXTEQU <Expression !> 255> ; errstr = “Expression > 255”
Text delimiters also have a special use with the FOR directive, as explained in
“FOR Loops and Variable-Length Parameters,” later in this chapter.
Expansion Operator
The expansion operator (%) expands text macros or converts constant
expressions into their text representations. It performs these tasks differently in
different contexts, as discussed in the following.
Converting Numeric Expressions to Text

The expansion operator can convert numbers to text. The operator forces
immediate evaluation of a constant expression and replaces it with a text value
consisting of the digits of the result. The digits are generated in the current radix
(default decimal).
This application of the expansion operator is useful when defining a text macro,
as the following lines show. Notice how you can enclose expressions with
parentheses to make them more readable:
a TEXTEQU <3 + 4> ; a = “3 + 4”
b TEXTEQU %3 + 4 ; b = “7”
c TEXTEQU %(3 + 4) ; c = “7”
When assigning text macros, you can use numeric equates in the constant
expressions, but not text macros:
num EQU 4 ; num = 4
numstr TEXTEQU <4> ; numstr = <4>
a TEXTEQU %3 + num ; a = <7>
b TEXTEQU %3 + numstr ; b = <7>
The expansion operator gives you flexibility when passing arguments to macros.
It lets you pass a computed value rather than the literal text of an expression.
The following example illustrates by defining a macro
work MACRO arg
mov ax, arg * 4
ENDM

which accepts different arguments:

work 2 + 3 ; Passes “2 + 3”
; Code: mov ax, 2 + (3 * 4)
work %2 + 3 ; Passes 5
; Code: mov ax, 5 * 4
work 2 + num ; Passes “2 + num”
work %2 + num ; Passes “6”
work 2 + numstr ; Passes “2 + numstr”
work %2 + numstr ; Passes “6”
You must consider operator precedence when using the expansion operator.
Parentheses inside the macro can force evaluation in a desired order:
work MACRO arg
mov ax, (arg) * 4
ENDM
work 2 + 3 ; Code: mov ax, (2 + 3) * 4

work %2 + 3 ; Code: mov ax, (5) * 4
Several other uses for the expansion operator are reviewed in “Returning Values
with Macro Functions,” later in this chapter.
Expansion Operator as First Character on a Line

The expansion operator has a different meaning when used as the first character
on a line. In this case, it instructs the assembler to expand any text macros and
macro functions it finds on the rest of the line.
This feature makes it possible to use text macros with directives such as ECHO,
TITLE, and SUBTITLE, which take an argument consisting of a single text
value. For instance, ECHO displays its argument to the standard output device
during assembly. Such expansion can be useful for debugging macros and
expressions, but the requirement that its argument be a single text value may
have unexpected results. Consider this example:
ECHO Bytes per element: %(SIZEOF array / LENGTHOF array)
Instead of evaluating the expression, this line echoes it:

Bytes per element: %(SIZEOF array / LENGTHOF array)
However, you can achieve the desired result by assigning the text of the
expression to a text macro and then using the expansion operator at the
beginning of the line to force expansion of the text macro.

temp TEXTEQU %(SIZEOF array / LENGTHOF array)

% ECHO Bytes per element: temp
Note that you cannot get the same results simply by putting the % at the
beginning of the first echo line, because % expands only text macros, not
numeric equates or constant expressions.
Here are more examples of the expansion operator at the start of a line:
; Assume memmod, lang, and os specified with /D option
% SUBTITLE Model: memmod Language: lang Operating System: os
; Assume num defined earlier

tnum TEXTEQU %num
% .ERRE num LE 255, <Failed because tnum !> 255>
Substitution Operator
References to a parameter within a macro can sometimes be ambiguous. In such
cases, the assembler may not expand the argument as you intend. The
substitution operator (&) lets you identify unambiguously any parameter within a
macro.
As an example, consider the following macro:
errgen MACRO num, msg
PUBLIC errnum
errnum BYTE “Error num: msg”
ENDM
This macro is open to several interpretations:

u Is errnum a distinct word or the word err next to the parameter num?
u Should num and msg within the string be treated literally as part of the string
or as arguments?
In each case, the assembler chooses the most literal interpretation. That is, it
treats errnum as a distinct word, and num and msg as literal parts of the string.
The substitution operator can force different interpretations. If we rewrite the
macro with the & operator, it looks like this:
errgen MACRO num, msg
PUBLIC err&num
err&num BYTE “Error &num: &msg”
ENDM

When called with the following arguments,

errgen 5, <Unreadable disk>
the macro now generates this code:

PUBLIC err5
err5 BYTE “Error 5: Unreadable disk”
When it encounters the & operator, the assembler interprets subsequent text as a
parameter name until the next & or until the next separator character (such as a
space, tab, or comma). Thus, the assembler correctly parses the expression
err&num because num is delimited by & and a space. The expression could also
be written as err&num&, which again unambiguously identifies num as a
parameter.
The rule also works in reverse. You can delimit a parameter reference with & at
the end rather than at the beginning. For example, if num is 5, the expression
num&12 resolves to “512.”
The assembler processes substitution operators from left to right. This can have
unexpected results when you are pasting together two macro parameters. For
example, if arg1 has the value var and arg2 has the value 3, you could paste
them together with this statement:
&arg1&&arg2& BYTE “Text”
Eliminating extra substitution operators, you might expect the following to be

equivalent:
&arg1&arg2 BYTE “Text”
However, this actually produces the symbol vararg2, because in processing

from left to right, the assembler associates both the first and the second &
symbols with the first parameter. The assembler replaces &arg1& by var,
producing vararg2. The arg2 is never evaluated. The correct abbreviation is:
arg1&&arg2 BYTE “Text”
which produces the desired symbol var3. The symbol arg1&&arg2 is replaced
by var&arg2, which is replaced by var3.
The substitution operator is also necessary if you want to substitute a text macro
inside quotes. For example,

arg TEXTEQU <hello>

%echo This is a string “&arg” ; Produces: This is a string “hello”
%echo This is a string “arg” ; Produces: This is a string “arg”
You can also use the substitution operator in lines beginning with the expansion
operator (%) symbol, even outside macros (see page 236). It may be necessary
to use the substitution operator to paste text macro names to adjacent characters
or symbol names, as shown here:
text TEXTEQU <var>
value TEXTEQU %5
% ECHO textvalue is text&&value
This echoes the message

textvalue is var5
Macro substitution always occurs before evaluation of the high-level control

structures. The assembler may therefore mistake a bit-test operator (&) in your
macro for a substitution operator. You can guarantee the assembler correctly
recognizes a bit-test operator by enclosing its operands in parentheses, as shown
here:
test MACRO x
.IF ax==&x ; &x substituted with parameter value
mov ax, 10
.ELSEIF ax&(x) ; & is bitwise AND
mov ax, 20
.ENDIF
ENDM
The rules for using the substitution operator have changed significantly since
MASM 5.1, making macro behavior more consistent and flexible. If you have
macros written for MASM 5.1 or earlier, you can specify the old behavior by
using OLDMACROS or M510 with the OPTION directive (see page 24).
Defining Repeat Blocks with Loop Directives

A “repeat block” is an unnamed macro defined with a loop directive. The loop
directive generates the statements inside the repeat block a specified number of
times or until a given condition becomes true.
MASM provides several loop directives, which let you specify the number of
loop iterations in different ways. Some loop directives can also accept arguments
for each iteration. Although the number of iterations is usually specified in the
directive, you can use the EXITM directive to exit the loop early.

Repeat blocks can be used outside macros, but they frequently appear inside
macro definitions to perform some repeated operation in the macro. Since repeat
blocks are macros themselves, they end with the ENDM directive.

This section explains the following four loop directives: REPEAT, WHILE,
FOR, and FORC. In versions of MASM prior to 6.0, REPEAT was called
REPT, FOR was called IRP, and FORC was called IRPC. MASM 6.1
recognizes the old names.
The assembler evaluates repeat blocks on the first pass only. You should
therefore avoid using address spans as loop counters, as in this example:
REPEAT (OFFSET label1 - OFFSET label2) ; Don't do this!
Since the distance between two labels may change on subsequent assembly
passes as the assembler optimizes code, you should not assume that address
spans remain constant between passes.
Note The REPEAT and WHILE directives should not be confused with the
REPEAT and WHILE directives (see “Loop-Generating Directives” in Chapter
7), which generate loop and jump instructions for run-time program control.
REPEAT Loops
REPEAT is the simplest loop directive. It specifies the number of times to
generate the statements inside the macro. The syntax is:
REPEAT constexpr
statements
ENDM
The constexpr can be a constant or a constant expression, and must contain no
forward references. Since the repeat block expands at assembly time, the
number of iterations must be known then.
Here is an example of a repeat block used to generate data. It initializes an array
containing sequential ASCII values for all uppercase letters.
alpha LABEL BYTE ; Name the data generated
letter = ‘A’ ; Initialize counter
REPEAT 26 ;; Repeat for each letter
BYTE letter ;; Allocate ASCII code for letter
letter = letter + 1 ;; Increment counter
ENDM

Here is another use of REPEAT, this time inside a macro:

beep MACRO iter:=<3>
mov ah, 2 ;; Character output function
mov dl, 7 ;; Bell character
REPEAT iter ;; Repeat number specified by macro
int 21h ;; Call DOS
ENDM
ENDM
WHILE Loops
The WHILE directive is similar to REPEAT, but the loop continues as long as a
given condition is true. The syntax is:
WHILE expression
statements
ENDM
The expression must be a value that can be calculated at assembly time.
Normally, the expression uses relational operators, but it can be any expression
that evaluates to zero (false) or nonzero (true). Usually, the condition changes
during the evaluation of the macro so that the loop won’t attempt to generate an
infinite amount of code. However, you can use the EXITM directive to break
out of the loop.
The following repeat block uses the WHILE directive to allocate variables
initialized to calculated values. This is a common technique for generating
lookup tables. (A lookup table is any list of precalculated results, such as a table
of interest payments or trigonometric values or logarithms. Programs optimized
for speed often use lookup tables, since calculating a value often takes more
time than looking it up in a table.)
cubes LABEL BYTE ;; Name the data generated
root = 1 ;; Initialize root
cube = root * root * root ;; Calculate first cube
WHILE cube LE 32767 ;; Repeat until result too large
WORD cube ;; Allocate cube
root = root + 1 ;; Calculate next root and cube
cube = root * root * root
ENDM

FOR Loops and Variable-Length Parameters

With the FOR directive you can iterate through a list of arguments, working on
each of them in turn. It has the following syntax:
FOR parameter, <argumentlist>
statements
ENDM
The parameter is a placeholder that represents the name of each argument
inside the FOR block. The argument list must contain comma-separated
arguments and must always be enclosed in angle brackets. Here’s an example of
a FOR block:
series LABEL BYTE
FOR arg, <1,2,3,4,5,6,7,8,9,10>
BYTE arg DUP (arg)
ENDM
On the first iteration, the arg parameter is replaced with the first argument, the
value 1. On the second iteration, arg is replaced with 2. The result is an array
with the first byte initialized to 1, the next 2 bytes initialized to 2, the next 3
bytes initialized to 3, and so on.
The argument list is given specifically in this example, but in some cases the list
must be generated as a text macro. The value of the text macro must include the
angle brackets.
arglist TEXTEQU <!<3,6,9!>> ; Generate list as text macro
%FOR arg, arglist
. ; Do something to arg
.
.
ENDM
Note the use of the literal character operator (!) to identify angle brackets as
characters, not delimiters. See “Text Delimiters (< >) and the Literal-Character
Operator,” earlier in this chapter.
The FOR directive also provides a convenient way to process macros with a
variable number of arguments. To do this, add VARARG to the last parameter
to indicate that a single named parameter will have the actual value of all
additional arguments. For example, the following macro definition includes the
three possible parameter attributes — required, default, and variable.
work MACRO rarg:REQ, darg:=<5>, varg:VARARG

The variable argument must always be last. If this macro is called with the
statement
work 4, , 6, 7, a, b
the first argument is received as the value 4, the second is replaced by the
default value 5, and the last four are received as the single argument <6, 7,
a, b>. This is the same format expected by the FOR directive. The FOR
directive discards leading spaces but recognizes trailing spaces.
The following macro illustrates variable arguments:
show MACRO chr:VARARG
mov ah, 02h
FOR arg, <chr>
mov dl, arg
int 21h
ENDM
ENDM
When called with

show ‘O’, ‘K’, 13, 10
the macro displays each of the specified characters one at a time.

The parameter in a FOR loop can have the required or default attribute. You
can modify the show macro to make blank arguments generate errors:
mov ah, 02h
FOR arg:REQ, <chr>
mov dl, arg
int 21h
ENDM
ENDM
The macro now generates an error if called with

show ‘O’,, ‘K’, 13, 10

Another approach would be to use a default argument:

mov ah, 02h
FOR arg:=<‘ ’>, <chr>
mov dl, arg
int 21h
ENDM
ENDM
Now calling the macro with

show ‘O’,, ‘K’, 13, 10
inserts the default character, a space, for the blank argument.
FORC Loops
The FORC directive is similar to FOR, but takes a string of text rather than a
list of arguments. The statements are assembled once for each character
(including spaces) in the string, substituting a different character for the
parameter each time through.
The syntax looks like this:
FORC parameter, < text>
statements
ENDM
The text must be enclosed in angle brackets. The following example illustrates
FORC:
FORC arg, <ABCDEFGHIJKLMNOPQRSTUVWXYZ>
BYTE ‘&arg’ ;; Allocate uppercase letter
BYTE ‘&arg’ + 20h ;; Allocate lowercase letter
BYTE ‘&arg’ - 40h ;; Allocate ordinal of letter
ENDM
Notice that the substitution operator must be used inside the quotation marks to
make sure that arg is expanded to a character rather than treated as a literal
string.
With versions of MASM earlier than 6.0, FORC is often used for complex
parsing tasks. A long sentence can be examined character by character. Each
character is then either thrown away or pasted onto a token string, depending on
whether it is a separator character. The new predefined macro functions and
string processing directives discussed in the following section are usually more
efficient for these tasks.

String Directives and Predefined Functions

The assembler provides four directives for manipulating text:
Directive Description
SUBSTR Assigns part of string to a new symbol.
INSTR Searches for one string within another.
SIZESTR Determines the size of a string.
CATSTR Concatenates one or more strings to a single string.
These directives assign a processed value to a text macro or numeric equate. For
example, the following lines
num = 7
newstr CATSTR <3 + >, %num, < = > , %3 + num ; "3 + 7 = 10"
assign the string "3 + 7 = 10" to newstr. CATSTR and SUBSTR assign
text in the same way as the TEXTEQU directive. SIZESTR and INSTR assign
a number in the same way as the = operator. The four string directives take only
text values as arguments. Use the expansion operator (%) when you need to
make sure that constants and numeric equates expand to text, as shown in the
preceding lines.
Each of the string directives has a corresponding predefined macro function
version: @SubStr, @InStr, @SizeStr, and @CatStr. Macro functions are
similar to the string directives, but you must enclose their arguments in
parentheses. Macro functions return text values and can appear in any context
where text is expected. The following section, “Returning Values with Macro
Functions,” tells how to write your own macro functions. The following
example is equivalent to the previous CATSTR example:
num = 7
newstr TEXTEQU @CatStr( <3 + >, %num, < = > , %3 + num )
Macro functions are often more convenient than their directive counterparts
because you can use a macro function as an argument to a string directive or to
another macro function. Unlike string directives, predefined macro function
names are case sensitive when you use the /Cp command-line option.
Each string directive and predefined function acts on a string, which can be any
textItem. The textItem can be text enclosed in angle brackets (< >), the name of
a text macro, or a constant expression preceded by % (as in %constExpr).
Refer to Appendix B, “BNF Grammar,” for a list of types that textItem can
represent.

The following sections summarize the syntax for each of the string directives
and functions. The explanations focus on the directives, but the functions work
the same except where noted.
SUBSTR
name SUBSTR string, start[[, length]]
@SubStr( string, start[[, length]] )
The SUBSTR directive assigns a substring from a given string to the symbol
name. The start parameter specifies the position in string, beginning with 1, to
start the substring. The length gives the length of the substring. If you do not
specify length, SUBSTR returns the remainder of the string, including the start
character.
INSTR
name INSTR [[start,]] string, substring
@InStr( [[start]], string, substring )
The INSTR directive searches a specified string for an occurrence of substring
and assigns its position number to name. The search is case sensitive. The start
parameter is the position in string to start the search for substring. If you do not
specify start, it is assumed to be position 1, the start of the string. If INSTR
does not find substring, it assigns position 0 to name.
The INSTR directive assigns the position value name as if it were a numeric
equate. In contrast, the @InStr returns the value as a string of digits in the
current radix.
The @InStr function has a slightly different syntax than the INSTR directive.
You can omit the first argument and its associated comma from the directive.
You can leave the first argument blank with the function, but a blank function
argument must still have a comma. For example,
pos INSTR <person>, <son>
is the same as
pos = @InStr( , <person>, <son> )
You can also assign the return value to a text macro, like this:

strpos TEXTEQU @InStr( , <person>, <son> )
SIZESTR
name SIZESTR string
@SizeStr( string )
The SIZESTR directive assigns the number of characters in string to name. An
empty string returns a length of zero. The SIZESTR directive assigns the size
value to a name as if it were a numeric equate. The @SizeStr function returns
the value as a string of digits in the current radix.
CATSTR
name CATSTR string[, string]...
@CatStr( string[, string]... )
The CATSTR directive concatenates a list of text values into a single text value
and assigns it to name. TEXTEQU is technically a synonym for CATSTR.
TEXTEQU is normally used for single-string assignments, while CATSTR is
used for multistring concatenations.
The following example pushes and pops one set of registers, illustrating several
uses of string directives and functions:

; SaveRegs - Macro to generate a push instruction for each

; register in argument list. Saves each register name in the
; regpushed text macro.
regpushed TEXTEQU <> ;; Initialize empty string
SaveRegs MACRO regs:VARARG

LOCAL reg
FOR reg, <regs> ;; Push each register
push reg ;; and add it to the list
regpushed CATSTR <reg>, <,>, regpushed
ENDM ;; Strip off last comma
regpushed CATSTR <!<>, regpushed ;; Mark start of list with <
regpushed SUBSTR regpushed, 1, @SizeStr( regpushed )
regpushed CATSTR regpushed, <!>> ;; Mark end with >
ENDM
; RestoreRegs - Macro to generate a pop instruction for registers

; saved by the SaveRegs macro. Restores one group of registers.
RestoreRegs MACRO
LOCAL reg
%FOR reg, regpushed ;; Pop each register
pop reg
ENDM
ENDM
Notice how the SaveRegs macro saves its result in the regpushed text macro
for later use by the RestoreRegs macro. In this case, a text macro is used as a
global variable. By contrast, the reg text macro is used only in RestoreRegs.
It is declared LOCAL so it won’t take the name reg from the global name
space. The MACROS.INC file provided with MASM 6.1 includes expanded
versions of these same two macros.
Returning Values with Macro Functions

A macro function is a named group of statements that returns a value. When
calling a macro function, you must enclose its argument list in parentheses, even
if the list is empty. The function always returns text.
MASM 6.1 provides several predefined macro functions for common tasks. The
predefined macros include @Environ (see page 10) and the string functions
@SizeStr, @CatStr, @SubStr, and @InStr (discussed in the preceding
section).
You define macro functions in exactly the same way as macro procedures,
except that a macro function always returns a value through the EXITM
directive. Here is an example:

DEFINED MACRO symbol:REQ

IFDEF symbol
EXITM <-1> ;; True
ELSE
EXITM <0> ;; False
ENDIF
ENDM
This macro works like the defined operator in the C language. You can use it to
test the defined state of several different symbols with a single statement, as
shown here:
IF DEFINED( DOS ) AND NOT DEFINED( XENIX )
;; Do something
ENDIF
Notice that the macro returns integer values as strings of digits, but the IF
statement evaluates numeric values or expressions. There is no conflict because
the assembler sees the value returned by the macro function exactly as if the
user had typed the values directly into the program:
IF -1 AND NOT 0
Returning Values with EXITM

The return value must be text, a text equate name, or the result of another
macro function. A macro function must first convert a numeric value — such as
a constant, a numeric equate, or the result of a numeric expression — before
returning it. The macro function can use angle brackets or the expansion
operator (%) to convert numbers to text. The DEFINED macro, for instance,
could have returned its value as
EXITM %-1
Here is another example of a macro function that uses the WHILE directive to
calculate factorials:

factorial MACRO num:REQ

LOCAL i, factor
factor = num
i = 1
WHILE factor GT 1
i = i * factor
factor = factor - 1
ENDM
EXITM %i
ENDM
The integer result of the calculation is changed to a text string with the expansion
operator (%). The factorial macro can define data, as shown here:
var WORD factorial( 4 )
This statement initializes var with the number 24 (the factorial of 4).
Using Macro Functions with Variable-Length Parameter Lists

You can use the FOR directive to handle macro parameters with the VARARG
attribute. “FOR Loops and Variable-Length Parameters,” page 242, explains
how to do this in simple cases where the variable parameters are handled
sequentially, from first to last. However, you may sometimes need to process
the parameters in reverse order or nonsequentially. Macro functions make these
techniques possible.
For example, the following macro function determines the number of arguments
in a VARARG parameter:
@ArgCount MACRO arglist:VARARG
LOCAL count
count = 0
FOR arg, <arglist>
count = count + 1 ;; Count the arguments
ENDM
EXITM %count
ENDM

You can use @ArgCount inside a macro that has a VARARG parameter, as
shown here:
work MACRO args:VARARG
% ECHO Number of arguments is: @ArgCount( args )
ENDM
Another useful task might be to select an item from an argument list using an
index to indicate the item. The following macro simplifies this.
@ArgI MACRO index:REQ, arglist:VARARG
LOCAL count, retstr
retstr TEXTEQU <> ;; Initialize count
count = 0 ;; Initialize return string
FOR arg, <arglist>
count = count + 1
IF count EQ index ;; Item is found
retstr TEXTEQU <arg> ;; Set return string
EXITM ;; and exit IF
ENDIF
ENDM
EXITM retstr ;; Exit function
ENDM
You can use @ArgI like this:

% ECHO Third argument is: @ArgI( 3, args )
ENDM
Finally, you might need to process arguments in reverse order. The following
macro returns a new argument list in reverse order.
@ArgRev MACRO arglist:REQ
LOCAL txt, arg
txt TEXTEQU <>
% FOR arg, <arglist>
txt CATSTR <arg>, <,>, txt ;; Paste each onto list
ENDM
;; Remove terminating comma
txt SUBSTR txt, 1, @SizeStr( %txt ) - 1
txt CATSTR <!<>, txt, <!>> ;; Add angle brackets
EXITM txt
ENDM

Here is an example showing @ArgRev in use:

% FOR arg, @ArgRev( <args> ) ;; Process in reverse order
ECHO arg
ENDM
ENDM
These three macro functions appear in the MACROS.INC include file, located
on one of the MASM distribution disks.
Expansion Operator in Macro Functions

This list summarizes the behavior of the expansion operator (%) with macro
functions.
u If a macro function is preceded by a %, it will be expanded. However, if it
expands to a text macro or a macro function call, it will not expand further.
u If you use a macro function call as an argument for another macro function
call, a % is not needed.
u If a macro function is called inside angle brackets and is preceded by %, it
will be expanded.
Advanced Macro Techniques

The concept of replacing macro names with predefined macro text is simple in
theory, but it has many implications and complications. Here is a brief summary
of some advanced techniques you can use in macros.
Defining Macros within Macros

Macros can define other macros, a technique called “nesting macros.” MASM
expands macros as it encounters them, so nested macros are always processed
in nesting order. You cannot reference a nested macro directly in your program,
since the assembler begins expansion from the outer macro. In effect, a nested
macro is local to the macro that defines it. Only the amount of available memory
limits the number of macros a program can nest.

The following example demonstrates how one macro can define another. The
macro takes as an argument the name of a shift or rotate instruction, then
creates another macro that simplifies the instruction for 8088/86 processors.
shifts MACRO opname ;; Macro generates macros
opname&s MACRO operand:REQ, rotates:=<1>
IF rotates LE 2 ;; One at a time is faster
REPEAT rotate ;; for 2 or less
opname operand, 1
ENDM
ELSE ;; Using CL is faster for
mov cl, rotates ;; more than 2
opname operand, cl
ENDIF
ENDM
ENDM
Recall that the 8086 processor allows only 1 or CL as an operand for shift and
rotate instructions. Expanding shifts generates a macro for the shift
instruction that uses whichever operand is more efficient. You create the entire
series of macros, one for each shift instruction, like this:
; Call macro repeatedly to make new macros
shifts ror ; Generates rors
shifts rol ; Generates rols
shifts shr ; Generates shrs
shifts shl ; Generates shls
shifts rcl ; Generates rcls
shifts rcr ; Generates rcrs
shifts sal ; Generates sals
shifts sar ; Generates sars
Then use the new macros as replacements for shift instructions, like this:
shrs ax, 5
rols bx, 3
Testing for Argument Type and Environment

Macros can expand conditional blocks of code by testing for argument type with
the OPATTR operator. OPATTR returns a single word constant that indicates
the type and scope of an expression, like this:
OPATTR expression
If expression is not valid or is forward-referenced, OPATTR returns a 0.
Otherwise, the return value incorporates the bit flags shown in the table below.

OPATTR serves as an enhanced version of the .TYPE operator, which returns

only the low byte (bits 0 – 7) shown in the table. Bits 11 – 15 of the return value
are undefined.
Bit Set If expression
0 References a code label
1 Is a memory variable or has a relocatable data label
2 Is an immediate value
3 Uses direct memory addressing
4 Is a register value
5 References no undefined symbols and is without error
6 Is relative to SS
7 References an external label
8 – 10 Has the following language type:
u 000 — No language type
u 001 — C
u 010 — SYSCALL
u 011 — STDCALL
u 100 — Pascal
u 101 — FORTRAN
u 110 — Basic
A macro can use OPATTR to determine if an argument is a constant, a register,

or a memory operand. With this information, the macro can conditionally
generate the most efficient code depending on argument type.
For example, given a constant argument, a macro can test it for 0. Depending on
the argument’s value, the code can select the most effective method to load the
value into a register:
IF CONST
mov bx, CONST ; If CONST > 0, move into BX
ELSE
sub bx, bx ; More efficient if CONST = 0
ENDIF

The second method is faster than the first, yet has the same result (with the
byproduct of changing the processor flags).
The following macro illustrates some techniques using OPATTR by loading an
address into a specified offset register:
load MACRO reg:REQ, adr:REQ
IF (OPATTR (adr)) AND 00010000y ;; Register
IFDIFI reg, adr ;; Don’t load register
mov reg, adr ;; onto itself
ENDIF
ELSEIF (OPATTR (adr)) AND 00000100y
mov reg, adr ;; Constant
ELSEIF (TYPE (adr) EQ BYTE) OR (TYPE (adr) EQ SBYTE)
mov reg, OFFSET adr ;; Bytes
ELSEIF (SIZE (TYPE (adr)) EQ 2
mov reg, adr ;; Near pointer
ELSEIF (SIZE (TYPE (adr)) EQ 4
mov reg, WORD PTR adr[0] ;; Far pointer
mov ds, WORD PTR adr[2]
ELSE
.ERR <Illegal argument>
ENDIF
ENDM
A macro also can generate different code depending on the assembly

environment. The predefined text macro @Cpu returns a flag for processor
type. The following example uses the more efficient constant variation of the
PUSH instruction if the processor is an 80186 or higher.
IF @Cpu AND 00000010y
pushc MACRO op ;; 80186 or higher
push op
ENDM
ELSE
pushc MACRO op ;; 8088/8086
mov ax, op
push ax
ENDM
ENDIF
Another macro can now use pushc rather than conditionally testing for
processor type itself. Although either case produces the same code, using pushc
assembles faster because the environment is checked only once.

You can test the language and operating system using the @Interface text
macro. The memory model can be tested with the @Model, @DataSize, or
@CodeSize text macros.
You can save the contexts inside macros with PUSHCONTEXT and
POPCONTEXT. The options for these keywords are:
Option Description
ASSUMES Saves segment register information
RADIX Saves current default radix
LISTING Saves listing and CREF information
CPU Saves current CPU and processor
ALL All of the above
Using Recursive Macros

Macros can call themselves. In MASM 5.1 and earlier, recursion is an important
technique for handling variable arguments. MASM 6.1 handles variable
arguments much more cleanly with the FOR directive and the VARARG
attribute, as described in “FOR Loops and Variable-Length Parameters,” earlier
in this chapter. However, recursion is still available and may be useful for some
macros.


257
C H A P T E R 1 0
Writing a Dynamic-Link Library

For Windows
The Windows operating system relies heavily on service routines and data
contained in special libraries called “dynamic-link libraries,” or DLLs for short.
Most of what Windows comprises, from the collections of screen fonts to the
routines that handle the graphical interface, is provided by DLLs. MASM 6.1
contains tools that you can use to write DLLs in assembly language. This
chapter shows you how.
DLLs do not run under MS-DOS. The information in this chapter applies only
to Windows, drawing in part on the chapter “Writing a Module-Definition File”
in Environment and Tools. The acronym API, which appears throughout this
chapter, refers to the application programming interface that Windows provides
for programs. For documentation of API functions, see the Programmer’s
Reference, Volume 2 of the Windows Software Development Kit (SDK).
The first section of this chapter gives an overview of DLLs and their similarities
to normal libraries. The next section explores the parts of a DLL and the rules
you must follow to create one. The third section applies this information to an
example DLL.
Overview of DLLs
A dynamic-link library is similar to a normal run-time library. Both types of
libraries contain a collection of compiled procedures, which serve one or more
calling modules. To link a normal library, the linker copies the required functions
from the library file (which usually has a .LIB extension) and combines them
with other modules to form an executable program in .EXE format. This
process is called static linking.
In dynamic linking, the library functions are not copied to an .EXE file. Instead,
they reside in a separate file in executable form, ready to serve any calling
program, called a “client.” When the first client requires the library, Windows
takes care of loading the functions into memory and establishing linkage. If

Template: MSGRIDA1.DOT Author: Mike Eddy Last Saved By: Ruth L Silverio
subsequent clients also need the library, Windows dynamically links them with
the proper library functions already in memory.
Loading a DLL
How Windows loads a DLL affects the client rather than the DLL itself.
Accordingly, this section focuses on how to set up a client program to use a
DLL. Since the client can itself be a DLL, this is information a DLL
programmer should know. However, MASM 6.1 does not provide all the tools
required to create a stand-alone program for Windows. To create such a
program, called an “application,” you must use tools in the Windows SDK.
Windows provides two methods for loading a dynamic-link library into memory:
Method Description
Implicit loading Windows loads the DLL along with the first client program and links it
before the client begins execution.
Explicit loading Windows does not load the DLL until the first client explicitly requests it
during execution.
When you write a DLL, you do not need to know beforehand which of the two
methods will be used to load the library. The loading method is determined by
how the client is written, not the DLL.
Implicit Loading
The implicit method of loading a DLL offers the advantage of simplicity. The
client requires no extra programming effort and can call the library functions as
if they were normal run-time functions. However, implicit loading carries two
constraints:
u The name of the library file must have a .DLL extension.
u You must either list all DLL functions the client calls in the IMPORTS
section of the client’s module-definition file, or link the client with an import
library.
An import library contains no executable code. It consists of only the names and
locations of exported functions in a DLL. The linker uses the locations in the
import library to resolve references to DLL functions in the client and to build
an executable header. For example, the file LIBW.LIB provided with MASM
6.1 is the import library for the DLL files that contain the Windows API
functions.
The IMPLIB utility described in Environment and Tools creates an import
library. Run IMPLIB from the MS-DOS command line like this:

Chapter 10 Writing a Dynamic-Link Library for Windows 259
IMPLIB implibfile dllfile

where implibfile is the name of the import library you want to create from the
DLL file dllfile. Once you have created an import library from a DLL, link it
with a client program that relies on implicit loading, but does not list imported
functions in its module-definition file. Continuing the preceding example, here’s
the link step for a client program that calls library procedures in the DLL dllfile:
LINK client.OBJ, client.EXE, , implibfile, client.DEF
This simplified example creates the client program client.EXE, linking it with the
import library implibfile, which in turn was created from the DLL file dllfile.
To summarize implicit loading, a client program must either
u List DLL functions in the IMPORTS section of its module-definition file, or
u Link with an import library created from the DLL.
Implicit loading is best when a client always requires at least one procedure in
the library, since Windows automatically loads the library with the client. If the
client does not always require the library service, or if the client must choose at
run time between several libraries, you should use explicit loading, discussed
next.
Explicit Loading
To explicitly load a DLL, the client does not require linking with an import
library, nor must the DLL file have an extension of .DLL. Explicit loading
involves three steps in which the client calls Windows API functions:
1. The client calls LoadLibrary to load the DLL.
2. The client calls GetProcAddress to obtain the address of each DLL function
it requires.
3. When finished with the DLL, the client calls FreeLibrary to unload the DLL
from memory.
The following example fragment shows how a client written in assembly

language explicitly loads a DLL called SYSINFO.DLL and calls the DLL
function GetSysDate.
INCLUDE windows.inc
.DATA
hInstance HINSTANCE 0
szDLL BYTE 'SYSINFO.DLL', 0
szDate BYTE 'GetSysDate', 0
lpProc DWORD 0

.CODE
.
.
.
INVOKE LoadLibrary, ADDR szDLL ; Load SYSINFO.DLL
mov hInstance, ax ; Save instance count
INVOKE GetProcAddress, ax, ADDR szDate ; Get and save
mov lpProc, ax ; far address of
mov lpProc[2], dx ; GetSysDate
call lpProc ; Call GetSysDate
.
.
.
INVOKE FreeLibrary, hInstance ; Unload SYSINFO.DLL
For simplicity, the above example contains no error-checking code. An actual

program should check all values returned from the API functions.
The explicit method of loading a DLL requires more programming effort in the
client program. However, the method allows the client to control which (if any)
dynamic-link libraries to load at run time.
Searching for a DLL File

To load a DLL, whether implicitly or explicitly, Windows searches for the DLL
file in the following directories in the order shown:
1. The current directory
2. The Windows directory, which contains WIN.COM
3. The Windows system directory, which contains system files such as
GDI.EXE
4. The directory where the client program resides (except Windows 3.0 and
earlier)
5. Directories listed in the PATH environment string
6. Directories mapped in a network
If Windows does not locate the DLL in any of these directories, it prompts the
user with a message box.
Building a DLL
A DLL has additional programming requirements beyond those for a normal
run-time library. This section describes the requirements pertaining to the
library’s code, data, and stack. It also discusses the effects of the library’s
extension name.

DLL Code
The code in a DLL consists of exported and nonexported functions. Exported
functions, listed in the EXPORTS section of the module-definition file, are
public routines serving clients. Nonexported functions provide private, internal
support for the exported procedures. They are not visible to a client.
Under Windows, an exported library routine must appear to the caller as a far
procedure. Your DLL routines can use any calling convention you wish,
provided the caller assumes the same convention. You can think of dynamic-
link code as code for a normal run-time library with the following additions:
u An entry procedure
u A termination procedure
u Special prologue and epilogue code
Entry Procedure
A DLL, like any Windows-based program, must have an entry procedure.
Windows calls the entry procedure only once when it first loads the DLL,
passing the following information in registers:
u DS contains the library’s data segment address.
u DI holds the library’s instance handle.
u CX holds the library’s heap size in bytes.
Note Windows API functions destroy all registers except DI, SI, BP, DS, and
the stack pointer. To preserve the contents of other registers, your program
must save the registers before an API call and restore them afterwards.
This information corresponds to the data provided to an application. Since a

DLL has only one occurrence in memory, called an “instance,” the value in DI
is not usually important. However, a DLL can use its instance handle to obtain
resources from its own executable file.
The entry procedure does not need to record the address of the data segment.
Windows automatically ensures that each exported routine in the DLL has
access to the library’s data segment, as explained in “Prologue and Epilogue
Code,” on page 264.
The heap size contained in CX reflects the value provided in the HEAPSIZE
statement of the module-definition file. You need not make an accurate guess in
the HEAPSIZE statement about the library’s heap requirements, provided you
specify a moveable data segment. With a moveable segment, Windows
automatically allocates more heap when needed. However, Windows can

provide no more heap in a fixed data segment than the amount specified in the
HEAPSIZE statement. In any case, a library’s total heap cannot exceed 64K,
less the amount of static data. Static data and heap reside in the same segment.
Windows does not automatically deallocate unneeded heap while the DLL is in
memory. Therefore, you should not set an unnecessarily large value in the
HEAPSIZE statement, since doing so wastes memory.
The entry procedure calls the Windows API function LocalInit to allocate the
heap. The library must create a heap before its routines call any heap functions,
such as LocalAlloc. The following example illustrates these steps:
DLLEntry PROC FAR PASCAL PUBLIC ; Entry point for DLL
jcxz @F ; If no heap, skip

INVOKE LocalInit, ds, 0, cx ; Else set up the heap
.IF ( ax ) ; If successful,
INVOKE UnlockSegment, -1 ; unlock the data segment
@@: call LibMain ; Call DLL's data init routine
mov ax, TRUE ; Return AX = 1 if okay,
.ENDIF ; else if LocalInit error,
ret ; return AX = 0
DLLEntry ENDP
This example code is taken from the DLLENTRY.ASM module, contained in

the LIB subdirectory on one of the MASM 6.1 distribution disks. After
allocating the heap, the procedure calls the library’s initialization procedure —
called LibMain in this case. LibMain initializes the library’s static data (if
required), then returns to DLLEntry, which returns to Windows. If Windows
receives a return value of 0 (FALSE) from DLLEntry, it unloads the library and
displays an error message.
The process is similar to the way MS-DOS loads a terminate-and-stay-resident
program (TSR), described in the next chapter. Both the DLL and TSR return
control immediately to the operating system, then wait passively in memory to
be called.
The following section explains how a DLL gains control when Windows unloads
it from memory.
Termination Procedure
Windows maintains a DLL in memory until the last client program terminates or
explicitly unloads the library. When unloading a DLL, Windows first calls the
library’s termination procedure. This allows the DLL to return resources and do
any necessary cleanup operations before Windows unloads the library from
memory.

Libraries that have registered window procedures with RegisterClass need not
call UnregisterClass to remove the class registration. Windows does this
automatically when it unloads the library.
You must name the library’s termination procedure WEP (for Windows Exit
Procedure) and list it in the EXPORTS section of the library’s module-definition
file. To ensure immediate operation, provide an ordinal number and use the
RESIDENTNAME keyword, as described in the chapter “Creating Module-
Definition Files” in Environment and Tools. This keeps the name “WEP” in the
Windows-resident name table at all times.
Besides its name, the code for WEP should also remain constantly in memory.
To ensure this, place WEP in its own code segment and set the segment’s
attributes as PRELOAD FIXED in the SEGMENTS statement of the module-
definition file. Thus, your DLL code should use a memory model that allows
multiple code segments, such as medium model. Since a termination procedure
is usually short, keeping it resident in memory does not burden the operating
system.
The termination procedure accepts a single parameter, which can have one of
two values. These values are assigned to the following symbolic constants in the
WINDOWS.INC file located in the LIB subdirectory:
u WEP_SYSTEM_EXIT (value 1) indicates Windows is shutting down.
u WEP_FREE_DLL (value 0) indicates the library’s last client has terminated or
has called FreeLibrary, and Windows is unloading the DLL.
The following fragment provides an outline for a typical termination procedure:

WEP PROC FAR PASCAL EXPORT
wExitCode:WORD
Prolog ; Prologue macro,

.IF wExitCode == WEP_FREE_DLL ; discussed below
. ; Get ready to
. ; unload
.
ELSEIF wExitCode == WEP_SYSTEM_EXIT
. ; Windows is
. ; shutting down
.
. ENDIF ; If neither value,
; take no action
mov ax, TRUE ; Always return AX = 1
Epilog ; Epilogue code,
ret ; discussed below
WEP ENDP

Usually, the WEP procedure takes the same actions regardless of the parameter
value, since in either case Windows will unload the DLL.
Under Windows 3.0, the WEP procedure receives stack space of about 256
bytes. This allows the procedure to unhook interrupts, but little else. Any other
action, such as calling an API function, usually results in an unrecoverable
application error because of stack overflow. Later versions of Windows provide
at least 4K of stack to the WEP procedure, allowing it to call many API
functions.
However, WEP should not send or post a message to a client, because the client
may already be terminated. The WEP procedure should also not attempt file I/O,
since only application processes — not DLLs — can own files. When control
reaches WEP, the client may no longer exist and its files are closed.
Prologue and Epilogue Code

Exported procedures in a Windows-based program require special epilogue and
prologue code. (For a definition of these terms, see “Generating Prologue and
Epilogue Code” in Chapter 7.) The SAMPLES subdirectory on one of the
MASM 6.1 distribution disks contains macros you can use for far procedures in
your Windows-based programs. Here’s a listing of the prologue macro:
Prolog MACRO
mov ax, ds ; Must be 1st, since Windows overwrites
nop ; Placeholder for 3rd byte
inc bp ; Push odd BP. Not required, but
push bp ; allows CodeView to recognize frame
mov bp, sp ; Set up stack frame to access params
push ds ; Save DS
mov ds, ax ; Point DS to DLL's data segment
ENDM
The instruction
inc bp
marks the beginning of the stack frame with an odd number. This allows real-
mode Windows to locate segment addresses on the stack and update the
addresses when it moves or discards the corresponding segments. In protected
mode, selector values do not change when segments are moved, so marking the
stack frame is not required. However, certain debugging applications, such as
Microsoft Codeview for Windows and the Microsoft Windows 80386 Debugger
(both documented in Programming Tools of the SDK), search for a marked
frame to determine if the frame belongs to a far procedure. Without the mark,
these debuggers give meaningless information when backtracing through the
stack. Therefore, you should include the INC BP instruction for Windows-

based programs that may run in real mode or that require debugging with a
Microsoft debugger.
Another characteristic of the prologue macro may seem puzzling at first glance.
The macro moves DS into AX, then AX back into DS. This sequence of
instructions lets Windows selectively overwrite the prologue code in far
procedures. When Windows loads a program, it compares the names of far
procedures with the list of exported procedures in the module-definition file. For
procedures that do not appear on the list, Windows leaves their prologue code
untouched. However, Windows overwrites the first 3 bytes of all exported
procedures with
mov ax, DGROUP
where DGROUP represents the selector value for the library’s data segment.
This explains why the prologue macro reserves the third byte with a NOP
instruction. The 1-byte instruction serves as padding to provide a 3-byte area for
Windows to overwrite.
The epilogue code returns BP to normal, like this:
Epilog MACRO
pop ds ; Recover original DS
pop bp ; and BP+1
dec bp ; Reset to original BP
ENDM
DLL Data
A DLL can have its own local data segment up to 64K. Besides static data, the
segment contains the heap from which a procedure can allocate memory
through the LocalAlloc API function. You should minimize static data in a
DLL to reserve as much memory as possible for temporary allocations.
Furthermore, all procedures in the DLL draw from the same heap space. If
more than one procedure in the library accesses the heap, a procedure should
not hold allocated space unnecessarily at the expense of the other procedures.
A Windows-based program must reserve a “task header” in the first 16 bytes of
its data segment. If you link your program with a C run-time function, the C
startup code automatically allocates the task header. Otherwise, you must
explicitly reserve and initialize the header with zeros. The sample program
described in “Example of a DLL:SYSINFO,” page 267, shows how to allocate a
task header.

DLL Stack
A DLL does not declare a stack segment and does not allocate stack space. A
client program calls a library’s exported procedure through a simple far call, and
the stack does not change. The procedure is, in effect, part of the calling
program, and therefore uses the caller’s stack.
This simple arrangement differs from that used in small and medium models, in
which many C run-time functions accept near pointers as arguments. Such
functions assume the pointer is relative to the current data segment. In
applications, the call works even if the argument points to a local variable on the
stack, since DS and SS contain the same segment address.
However, in a DLL, DS and SS point to different segments. Under small and
medium models, a library procedure must always pass pointers to static
variables located in the data segment, not to local variables on the stack.
When you write a DLL, include the FARSTACK keyword with the .MODEL
directive, like this:
.MODEL small, pascal, farstack
This informs the assembler that SS points to a segment other than DGROUP.
With full segment definitions, also add the line:
ASSUME DS:DGROUP, SS:NOTHING
DLL Extension Names

You can name an explicitly-loaded DLL file with any extension. The many files
in your Windows directory with extensions such as .DRV and .FON are almost
certainly DLLs. Many DLLs have an .EXE extension, though they are not true
executable files.
A library with an .EXE extension should always include stub code, specified by
the STUB statement in the module-definition file. The stub code activates when
run under MS-DOS, usually displaying a message to inform the user that the
program requires Windows. Without the stub code, the system hangs if a user
attempts to run a DLL with an .EXE extension.
Do not name a DLL with a .COM extension, since MS-DOS will give control to
the first byte of the program header. The header does not contain executable
instructions, and the system will hang even if the DLL has stub code.
Summary
Following is a summary of the previous information in this chapter.

u A dynamic-link library has only one instance — that is, it can load only once
during a Windows session.
u A single DLL can service calls from many client programs. Windows takes
care of linkage between the DLL and each client.
u Windows loads a DLL either implicitly (along with the first client) or
explicitly (when the first client calls LoadLibrary). It unloads the DLL when
the last client either terminates or calls FreeLibrary.
u A client calls a DLL routine as a simple far procedure. The routine can use
any calling convention.
u Windows ensures that the first instruction in a DLL procedure moves the
address of the library’s data segment into AX. You must provide the proper
prologue code to allow space for this 3-byte instruction and to copy AX to
DS.
u All procedures in a DLL have access to a single common data segment. The
segment contains both static variables and heap space, and cannot exceed
64K.
u A DLL procedure uses the caller’s stack.
u All exported procedures in a DLL must appear in the EXPORTS list in the
library’s module-definition file.
Example of a DLL: SYSINFO

Like any library, a DLL should be as small and fast as possible — a good
argument for writing it in assembly language. This section describes an example
library called SYSINFO, written entirely in assembly language. The following
text applies previous information in this chapter to an actual DLL.
SYSINFO contains three callable procedures. The acronym ASCIIZ refers to a
string of ASCII characters terminated with a zero. The callable procedures are:
Procedure Description
GetSysTime Returns a far pointer to a 12-byte ASCIIZ string containing the current time
in hh:mm:ss format.
GetSysDate Returns a far pointer to an ASCIIZ string containing the current date in any
of six languages.
GetSysInfo Returns a far pointer to a structure containing the following system data:
u ASCIIZ string of Windows version
u ASCIIZ string of MS-DOS version
u Current keyboard status
u Current video mode

u Math coprocessor flag

u Processor type
u ASCIIZ string of ROM-BIOS release date
To see SYSINFO in action, follow the steps below. The file SYSDATA.EXE
resides in the SAMPLES\WINDLL subdirectory of MASM if you requested
example files when installing MASM. Otherwise, you must first install the file
with the MASM 6.1 SETUP utility.
u Create SYSINFO.DLL as described in the following section and place it in
the SAMPLES\WINDLL subdirectory for MASM 6.1.
u From the Windows File Manager, make the SAMPLES\WINDLL
subdirectory the current directory.
u In the Program Manager, choose Run from the File menu and type
SYSDATA
to run the example program SYSDATA.EXE. This program calls the
routines in SYSINFO.DLL and displays the returned data.
Entry Routine for SYSINFO

SYSINFO links with the DLLENTRY module, which serves as the library’s
entry point when Windows first loads the program. For a listing and description
of DLLENTRY.ASM, see the previous section, “Entry Procedure.”
DLLENTRY replaces the LIBENTRY module provided with the Windows
SDK, but unlocks the data segment after calling the API function InitTask.
LIBENTRY does not unlock the segment. DLLENTRY saves some space over
LIBENTRY, because it does not pass any arguments to LibMain.
The LibMain procedure handles the library’s initialization tasks. You can name
the procedure whatever you want, provided you make the same change in
DLLENTRY.ASM and reassemble both modules. You can even combine
DLLENTRY with LibMain to form one procedure, like this:

DLLInit PROC FAR PASCAL PUBLIC ; Entry point for DLL
jcxz @F ; If no heap, skip

INVOKE LocalInit, ds, 0, cx ; Else set up the heap
.IF ( ax ) ; If successful,
INVOKE UnlockSegment, -1 ; unlock the data segment
@@: . ; Initialize DLL data. This
. ; replaces the call to the
. ; LibMain procedure.
mov ax, TRUE ; Return AX = 1 if okay,
.ENDIF ; else if LocalInit error,
ret ; return AX = 0
DLLInit ENDP
END DLLInit
Whatever you call your combined procedure (DLLInit in the preceding

example), place the name on the END statement as shown. This identifies the
procedure as the one that first executes when Windows loads the DLL.
SYSINFO accommodates several international languages. Currently, SYSINFO
recognizes English, French, Spanish, German, Italian, and Swedish, but you can
easily extend the code to include other languages. LibMain calls
GetProfileString to determine the current language, then initializes the variable
indx accordingly. The variable indirectly points to an array of strings containing
days and months in different languages. The GetSysDate procedure uses these
strings to create a full date in the correct language.
Static Data
SYSINFO stores the strings in its static data segment. This data remains in
memory along with the library’s code. All procedures have equal access to the
data segment.
Because the library does not call any C run-time functions, it explicitly allocates
the low paragraph of the data segment with the variable TaskHead. This 16-
byte area serves as the required Windows task header, described in “DLL
Data,” earlier in this chapter.
Module-Definition File
The library’s module-definition file, named SYSINFO.DEF, looks like this:

LIBRARY SYSINFO
DESCRIPTION 'Sample assembly-language DLL'
EXETYPE WINDOWS
CODE PRELOAD MOVEABLE DISCARDABLE
DATA PRELOAD MOVEABLE SINGLE
SEGMENTS CODE2 PRELOAD FIXED
EXPORTS WEP @1 RESIDENTNAME
GetSysTime @2
GetSysDate @3
GetSysInfo @4
Note the following points about the module-definition file:

u The LIBRARY statement identifies SYSINFO as a dynamic-link library.
u SYSINFO places its termination procedure WEP in a separate code segment,
called CODE2, which the SEGMENTS statement declares as FIXED. This
keeps the WEP routine fixed in memory, while all other code remains
moveable.
u The EXPORTS section lists all procedures the library exports, including WEP.
u None of the library’s procedures require heap space, so SYSINFO.DEF
includes no HEAPSIZE statement.
Assembling and Linking SYSINFO

The following listing shows the description file for SYSINFO:
sysinfo.obj: sysinfo.asm dll.inc
ML /c /W3 sysinfo.asm
dllentry.obj: dllentry.asm dll.inc.
ML /c /W3 dllentry.asm
sysinfo.dll: dllentry.obj sysinfo.obj
LINK dllentry sysinfo, sysinfo.dll,, libw.lib mnocrtdw.lib,
sysinfo.def
To create SYSINFO.DLL, run the NMAKE utility described in Chapter 16 of

Environments and Tools. Or assemble and link SYSINFO with the three
command lines shown in the preceding listing. This does not require running
NMAKE.
SYSINFO links with the library modules MNOCRTDW.LIB and LIBW.LIB.
The first supplies the required Windows startup code for a medium-model DLL
that does not use any C run-time functions. LIBW.LIB is the Windows import
library, which contains no executable code. The import library provides linkage
information for the Windows API functions referenced in the DLL. Windows
establishes the final links when it loads the program.

Expanding SYSINFO
SYSINFO is an example of how to write an assembly-language DLL without
overwhelming detail. It has plenty of room for expansion and improvements.
The following list may give you some ideas:
u To create a heap area for the library, add the line
HEAPSIZE value
to the module-definition file, where value is an approximate guess for the

amount of heap required in bytes. The DLLEntry procedure automatically
allocates the indicated amount of heap. Keep the data segment moveable,
because Windows then provides more heap space if required by the DLL
procedures.
u If you want to add a procedure that calls C run-time functions, you must
replace MNOCRTDW.LIB with MDLLCW.LIB, which is supplied with the
Windows SDK. The MDLLCW.LIB library contains the run-time functions
for medium-model DLLs.
u Each time the GetSysInfo procedure is called, it retrieves the version
number of MS-DOS and Windows, gets the processor type, checks for a
coprocessor, and reads the ROM-BIOS release date. Since this information
does not change throughout a Windows session, it would be handled more
efficiently in the LibMain procedure, which executes only once. The code is
currently placed in GetSysInfo for the sake of clarity at the expense of
efficiency.
u SYSINFO is not a true international program. You can easily add more
languages, extending the days and months arrays accordingly. Moreover,
for the sake of simplicity, the GetSysDate procedure arranges the date with
an American bias. For example, in many parts of the world, the date numeral
appears before the month rather than after. If you use SYSINFO in your
own applications, you should include code in LibMain to determine the
correct date format with additional calls to GetProfileString. You can find
more information on how to do this in Chapter 18 of the Microsoft Windows
Programmer’s Reference, Volume 1, supplied with the Windows SDK.


273
C H A P T E R 1 1
Writing Memory-Resident Software
Through its memory-management system, MS-DOS allows a program to remain

resident in memory after terminating. The resident program can later regain
control of the processor to perform tasks such as background printing or
“popping up” a calculator on the screen. Such a program is commonly called a
TSR, from the terminate-and-stay-resident function it uses to return to MS-
DOS.
This chapter explains the techniques of writing memory-resident software. The
first two sections present introductory material. Following sections describe
important MS-DOS and BIOS interrupts and focus on how to write safe,
compatible, memory-resident software. Two example programs illustrate the
techniques described in the chapter. The MASM 6.1 disks contain complete
source code for the two example TSR programs.
Terminate-and-Stay-Resident Programs
MS-DOS maintains a pointer to the beginning of unused memory. Programs
load into memory at this position and terminate execution by returning control to
MS-DOS. Normally, the pointer remains unchanged, allowing MS-DOS to reuse
the same memory when loading other programs.
A terminating program can, however, prevent other programs from loading on
top of it. These programs exit to MS-DOS through the terminate-and-stay-
resident function, which resets the free-memory pointer to a higher position.
This leaves the program resident in a protected block of memory, even though it
is no longer running.

The terminate-and-stay-resident function (Function 31h) is one of the MS-DOS

services invoked through Interrupt 21h. The following fragment shows how a
TSR program terminates through Function 31h and remains resident in a 1000h-
byte block of memory:
mov ah, 31h ; Request DOS Function 31h
mov al, err ; Set return code
mov dx, 100h ; Reserve 100h paragraphs
; (1000h bytes)
int 21h ; Terminate-and-stay-resident
Note In current versions of MS-DOS, Interrupt 27h also provides a terminate-

and-stay-resident service. However, Microsoft cannot guarantee future support
for Interrupt 27h and does not recommend its use.
Structure of a TSR
TSRs consist of two distinct parts that execute at different times. The first part
is the installation section, which executes only once, when MS-DOS loads the
program. The installation code performs any initialization tasks required by the
TSR and then exits through the terminate-and-stay-resident function.
The second part of the TSR, called the resident section, consists of code and
data left in memory after termination. Though often identified with the TSR
itself, the resident section makes up only part of the entire program.
The TSR’s resident code must be able to regain control of the processor and
execute after the program has terminated. Methods of executing a TSR are
classified as either passive or active.
Passive TSRs
The simplest way to execute a TSR is to transfer control to it explicitly from
another program. Because the TSR in this case does not solicit processor
control, it is said to be passive. If the calling program can determine the TSR’s
memory address, it can grant control via a far jump or call. More commonly, a
program activates a passive TSR through a software interrupt. The installation
section of the TSR writes the address of its resident code to the proper position
in the interrupt vector table (see “MS-DOS Interrupts” in Chapter 7). Any
subsequent program can then execute the TSR by calling the interrupt.
Passive TSRs often replace existing software interrupts. For example, a passive
TSR might replace Interrupt 10h, the BIOS video service. By intercepting calls
that read or write to the screen, the TSR can access the video buffer directly,
increasing display speed.

Chapter 11 Writing Memory-Resident Software 275
Passive TSRs allow limited access since they can be invoked only from another
program. They have the advantage of executing within the context of the calling
program, and thus run no risk of interfering with another process. Such a risk
does exist with active TSRs.
Active TSRs
The second method of executing a TSR involves signaling it through some
hardware event, such as a predetermined sequence of keystrokes. This type of
TSR is “active” because it must continually search for its startup signal. The
advantage of active TSRs lies in their accessibility. They can take control from
any running application, execute, and return, all on demand.
An active TSR, however, must not seize processor control blindly. It must
contain additional code that determines the proper moment at which to execute.
The extra code consists of one or more routines called “interrupt handlers,”
described in the following section.
Interrupt Handlers in Active TSRs

The memory-resident portion of an active TSR consists of two parts. One part
contains the body of the TSR — the code and data that perform the program’s
main tasks. The other part contains the TSR’s interrupt handlers.
An interrupt handler is a routine that takes control when a specific interrupt
occurs. Although sometimes called an “interrupt service routine,” a TSR’s
handler usually does not service the interrupt. Instead, it passes control to the
original interrupt routine, which does the actual interrupt servicing. (See the
section “Replacing an Interrupt Routine” in Chapter 7 for information on how to
write an interrupt handler.)
Collectively, interrupt handlers ensure that a TSR operates compatibly with the
rest of the system. Individually, each handler fulfills one or more of the
following functions:
u Auditing hardware events that may signal a request for the TSR
u Monitoring system status
u Determining whether a request for the TSR should be honored, based on
current system status
Auditing Hardware Events for TSR Requests

Active TSRs commonly use a special keystroke sequence or the timer as a
request signal. A TSR invoked through one of these channels must be equipped
with handlers that audit keyboard or timer events.

A keyboard handler receives control at every keystroke. It examines each key,

searching for the proper signal or “hot key.” Generally, a keyboard handler
should not attempt to call the TSR directly when it detects the hot key. If the
TSR cannot safely interrupt the current process at that moment, the keyboard
handler is forced to exit to allow the process to continue. Since the handler
cannot regain control until the next keystroke, the user has to press the hot key
repeatedly until the handler can comply with the request.
Instead, the handler should merely set a request flag when it detects a hot-key
signal and then exit normally. Examples in the following paragraphs illustrate this
technique.
For computers other than MCA (IBM PS/2 and compatible), an active TSR
audits keystrokes through a handler for Interrupt 09, the keyboard interrupt:
Keybrd PROC FAR
sti ; Interrupts are okay
push ax ; Save AX register
in al, 60h ; AL = key scan code
call CheckHotKey ; Check for hot key
.IF carry? ; If hot key pressed,
mov cs:TsrRequestFlag, TRUE ; raise flag and
. ; set up for exit
.
.
A TSR running on a PS/2 computer cannot reliably read key scan codes using
this method. Instead, the TSR must search for its hot key through a handler for
Interrupt 15h (Miscellaneous System Services). The handler determines the
current keypress from the AL register when AH equals 4Fh, as shown here:
MiscServ PROC FAR
sti ; Interrupts okay
.IF ah == 4Fh ; If Keyboard Intercept Service:
call CheckHotKey ; Check for hot key
.IF carry? ; If hot key pressed,
mov cs:TsrRequestFlag, TRUE ; raise flag and
. ; set up for exit
.
.
The example program on page 293 shows how a TSR tests for a PS/2 machine
and then sets up a handler for either Interrupt 09 or Interrupt 15h to audit
keystrokes.
Setting a request flag in the keyboard handler allows other code, such as the
timer handler (Interrupt 08), to recognize a request for the TSR. The timer
handler gains control at every timer interrupt, which occurs an average of 18.2
times per second.

The following fragment shows how a timer handler tests the request flag and
continually polls until it can safely execute the TSR.
NewTimer PROC FAR
.
.
.
cmp TsrRequestFlag, FALSE ; Has TSR been requested?
.IF !zero? ; If so, can system be
call CheckSystem ; interrupted safely?
.IF carry? ; If so,
call ActivateTsr ; activate TSR
.
.
.
Monitoring System Status

A TSR that uses a hardware device such as the video or disk must not interrupt
while the device is active. A TSR monitors a device by handling the device’s
interrupt. Each interrupt handler simply sets a flag to indicate the device is in
use, and then clears the flag when the interrupt finishes.
The following shows a typical monitor handler:
NewHandler PROC FAR
mov cs:ActiveFlag, TRUE ; Set active flag
pushf ; Simulate interrupt by
; pushing flags, then
call OldHandler ; far-calling original routine
mov cs:ActiveFlag, FALSE ; Clear active flag
iret ; Return from interrupt
NewHandler ENDP
Only hardware used by the TSR requires monitoring. For example, a TSR that
performs disk input/output (I/O) must monitor disk use through Interrupt 13h.
The disk handler sets an active flag that prevents the TSR from executing during
a read or write operation. Otherwise, the TSR’s own I/O would move the disk
head. This would cause the suspended disk operation to continue with the head
incorrectly positioned when the TSR returned control to the interrupted
program.
In the same way, an active TSR that displays to the screen must monitor calls to
Interrupt 10h. The Interrupt 10h BIOS routine does not protect critical sections
of code that program the video controller. The TSR must therefore ensure it
does not interrupt such nonreentrant operations.
The activities of the operating system also affect the system status. With few
exceptions, MS-DOS functions are not reentrant and must not be interrupted.

However, monitoring MS-DOS is somewhat more complicated than monitoring

hardware. This subject is discussed in “Using MS-DOS in Active TSRs,” later
in this chapter.
Figure 11.1 illustrates the process described so far. It shows a time line for a
typical TSR signaled from the keyboard. When the keyboard handler detects the
proper hot key, it sets a request flag called TsrRequestFlag. Thereafter, the
timer handler continually checks the system status until it can safely call the
TSR.
Figure 11.1 Time Line of Interactions Between Interrupt Handlers for a Typical TSR
The following comments describe the chain of events depicted in Figure 11.1.
Each comment refers to one of the numbered pointers in the figure.
1. At time = t, the timer handler activates. It finds the flag TsrRequestFlag
clear, indicating the user has not requested the TSR. The handler terminates
without taking further action. Notice that Interrupt 13h is currently
processing a disk I/O operation.
2. Before the next timer interrupt, the keyboard handler detects the hot key,
signaling a request for the TSR. The keyboard handler sets
TsrRequestFlag and returns.
3. At time = t + 1/18 second, the timer handler again activates and finds
TsrRequestFlag set. The handler checks other active flags to determine if
the TSR can safely execute. Since Interrupt 13h has not yet completed its
disk operation, the timer handler finds DiskActiveFlag set. The handler
therefore terminates without activating the TSR.

4. At time = t + 2/18 second, the timer handler again finds TsrRequestFlag

set and repeats its scan of the active flags. DiskActiveFlag is now clear,
but in the interim, Interrupt 10h has activated as indicated by the flag
VideoActiveFlag. The timer handler accordingly terminates without
activating the TSR.
5. At time = t + 3/18 second, the timer handler repeats the process. This time it
finds all active flags clear, indicating the TSR can safely execute. The timer
handler calls the TSR, which sets its own active flag to ensure it will not
interrupt itself if requested again.
6. The timer and other interrupts continue to function normally while the TSR
executes.
The timer itself can serve as the startup signal if the TSR executes periodically.
Screen clocks that continuously show seconds and minutes are examples of
TSRs that use the timer this way. ALARM.ASM, a program described in the
next section, shows another example of a timer-driven TSR.
Determining Whether to Invoke the TSR

Once a handler receives a request signal for the TSR, it checks the various
active flags maintained by the handlers that monitor system status. If any of the
flags are set, the handler ignores the request and exits. If the flags are clear, the
handler invokes the TSR, usually through a near or far call. Figure 11.1
illustrates how a timer handler detects a request and then periodically scans
various active flags until all the flags are clear.
A TSR that changes stacks must not interrupt itself. Otherwise, the second
execution would overwrite the stack data belonging to the first. A TSR prevents
this by setting its own active flag before executing, as shown in Figure 11.1. A
handler must check this flag along with the other active flags when determining
whether the TSR can safely execute.
Example of a Simple TSR: ALARM

This section presents a simple alarm clock TSR that demonstrates some of the
material covered so far. The program accepts an argument from the command
line that specifies the alarm setting in military form, such as 1635 for 4:35 P.M.
For simplicity, the argument must consist of four digits, including leading zeros.
To set the alarm at 7:45 A.M. , for example, enter the command:
ALARM 0745
The installation section of the program begins with the Install procedure.
Install computes the number of five-second intervals that must elapse before
the alarm sounds and stores this number in the word CountDown. The

procedure then obtains the vector for Interrupt 08 (timer) through MS-DOS
Function 35h and stores it in the far pointer OldTimer. Function 25h replaces
the vector with the far address of the new timer handler NewTimer. Once
installed, the new timer handler executes at every timer interrupt. These
interrupts occur 18.2 times per second or 91 times every five seconds.
Each time it executes, NewTimer subtracts one from a secondary counter called
Tick91. By counting 91 timer ticks, Tick91 accurately measures a period of
five seconds. When Tick91 reaches zero, it’s reset to 91 and CountDown is
decremented by one. When CountDown reaches zero, the alarm sounds.
;* ALARM.ASM - A simple memory-resident program that beeps the speaker
;* at a prearranged time. Can be loaded more than once for multiple
;* alarm settings. During installation, ALARM establishes a handler
;* for the timer interrupt (Interrupt 08). It then terminates through
;* the terminate-and-stay-resident function (Function 31h). After the
;* alarm sounds, the resident portion of the program retires by setting
;* a flag that prevents further processing in the handler.
.MODEL tiny ; Create ALARM.COM

.STACK
.CODE
ORG 5Dh ; Location of time argument in PSP,
CountDown LABEL WORD ; converted to number of 5-second
; intervals to elapse
.STARTUP
jmp Install ; Jump over data and resident code
; Data must be in code segment so it won’t be thrown away with Install code.
OldTimer DWORD ? ; Address of original timer routine
tick_91 BYTE 91 ; Counts 91 clock ticks (5 seconds)
TimerActiveFlag BYTE 0 ; Active flag for timer handler
;* NewTimer - Handler routine for timer interrupt (Interrupt 08).

;* Decrements CountDown every 5 seconds. No other action is taken
;* until CountDown reaches 0, at which time the speaker sounds.
NewTimer PROC FAR

.IF cs:TimerActiveFlag != 0 ; If timer busy or retired,
jmp cs:OldTimer ; jump to original timer routine
.ENDIF
inc cs:TimerActiveFlag ; Set active flag
pushf ; Simulate interrupt by pushing flags,
call cs:OldTimer ; then far-calling original routine
push ds ; Preserve DS register
push cs ; Point DS to current segment for
pop ds ; further memory access
dec tick_91 ; Count down for 91 ticks

.IF zero? ; If 91 ticks have elapsed,

mov tick_91, 91 ; reset secondary counter and
dec CountDown ; subtract one 5-second interval
.IF zero? ; If CountDown drained,
call Sound ; sound speaker
inc TimerActiveFlag ; Alarm has sounded--inc flag
.ENDIF ; again so it remains set
.ENDIF
dec TimerActiveFlag ; Decrement active flag

pop ds ; Recover DS
iret ; Return from interrupt handler
NewTimer ENDP
;* Sound - Sounds speaker with the following tone and duration:
BEEP_TONE EQU 440 ; Beep tone in hertz

BEEP_DURATION EQU 6 ; Number of clocks during beep,
; where 18 clocks = approx 1 second
Sound PROC USES ax bx cx dx es ; Save registers used in this routine

mov al, 0B6h ; Initialize channel 2 of
out 43h, al ; timer chip
mov dx, 12h ; Divide 1,193,180 hertz
mov ax, 34DCh ; (clock frequency) by
mov bx, BEEP_TONE ; desired frequency
div bx ; Result is timer clock count
out 42h, al ; Low byte of count to timer
mov al, ah
out 42h, al ; High byte of count to timer
in al, 61h ; Read value from port 61h
or al, 3 ; Set first two bits
out 61h, al ; Turn speaker on
; Pause for specified number of clock ticks
mov dx, BEEP_DURATION ; Beep duration in clock ticks

sub cx, cx ; CX:DX = tick count for pause
mov es, cx ; Point ES to low memory data
add dx, es:[46Ch] ; Add current tick count to CX:DX
adc cx, es:[46Eh] ; Result is target count in CX:DX
.REPEAT
mov bx, es:[46Ch] ; Now repeatedly poll clock
mov ax, es:[46Eh] ; count until the target
sub bx, dx ; time is reached
sbb ax, cx
.UNTIL !carry?

in al, 61h ; When time elapses, get port value

xor al, 3 ; Kill bits 0-1 to turn
out 61h, al ; speaker off
ret
Sound ENDP
;* Install - Converts ASCII argument to valid binary number, replaces

;* NewTimer as the interrupt handler for the timer, then makes program
;* memory-resident by exiting through Function 31h.
;*
;* This procedure marks the end of the TSR's resident section and the
;* beginning of the installation section. When ALARM terminates through
;* Function 31h, the above code and data remain resident in memory. The
;* memory occupied by the following code is returned to DOS.
Install PROC
; Time argument is in hhmm military format. Converts ASCII digits to

; number of minutes since midnight, then converts current time to number
; of minutes since midnight. Difference is number of minutes to elapse
; until alarm sounds. Converts to seconds-to-elapse, divides by 5 seconds,
; and stores result in word CountDown.
DEFAULT_TIME EQU 3600 ; Default alarm setting = 1 hour
; (in seconds) from present time
mov ax, DEFAULT_TIME
cwd ; DX:AX = default time in seconds
.IF BYTE PTR CountDown != ' ' ; If not blank argument,
xor CountDown[0], '00' ; convert 4 bytes of ASCII
xor CountDown[2], '00' ; argument to binary
mov al, 10 ; Multiply 1st hour digit by 10

mul BYTE PTR CountDown[0] ; and add to 2nd hour digit
add al, BYTE PTR CountDown[1]
mov bh, al ; BH = hour for alarm to go off
mov al, 10 ; Repeat procedure for minutes
mul BYTE PTR CountDown[2] ; Multiply 1st minute digit by 10
add al, BYTE PTR CountDown[3] ; and add to 2nd minute digit
mov bl, al ; BL = minute for alarm to go off
mov ah, 2Ch ; Request Function 2Ch
int 21h ; Get Time (CX = current hour/min)

mov dl, dh
sub dh, dh
push dx ; Save DX = current seconds
mov al, 60 ; Multiply current hour by 60
mul ch ; to convert to minutes
sub ch, ch
add cx, ax ; Add current minutes to result
; CX = minutes since midnight
mov al, 60 ; Multiply alarm hour by 60
mul bh ; to convert to minutes
sub bh, bh
add ax, bx ; AX = number of minutes since
; midnight for alarm setting
sub ax, cx ; AX = time in minutes to elapse
; before alarm sounds
.IF carry? ; If alarm time is tomorrow,
add ax, 24 * 60 ; add minutes in a day
.ENDIF
mov bx, 60
mul bx ; DX:AX = minutes-to-elapse-times-60
pop bx ; Recover current seconds
sub ax, bx ; DX:AX = seconds to elapse before
sbb dx, 0 ; alarm activates
.IF carry? ; If negative,
mov ax, 5 ; assume 5 seconds
cwd
.ENDIF
.ENDIF
mov bx, 5 ; Divide result by 5 seconds
div bx ; AX = number of 5-second intervals
mov CountDown, ax ; to elapse before alarm sounds
mov ax, 3508h ; Request Function 35h

int 21h ; Get Vector for timer (Interrupt 08)
mov WORD PTR OldTimer[0], bx ; Store address of original
mov WORD PTR OldTimer[2], es ; timer interrupt
mov ax, 2508h ; Request Function 25h
mov dx, OFFSET NewTimer ; DS:DX points to new timer handler
int 21h ; Set Vector with address of NewTimer
mov dx, OFFSET Install ; DX = bytes in resident section

mov cl, 4
shr dx, cl ; Convert to number of paragraphs
inc dx ; plus one
mov ax, 3100h ; Request Function 31h, error code=0
int 21h ; Terminate-and-stay-resident
Install ENDP
END

Note the following points about ALARM:

u The constant BEEP_TONE specifies the alarm tone. Practical values for the
tone range from approximately 100 to 4,000 hertz.
u The Install procedure marks the beginning of the installation section of
the program. Execution begins here when ALARM.COM is loaded. A TSR
generally places its installation code after the resident section. This allows the
terminating TSR to include the installation code with the rest of the memory
it returns to MS-DOS. Since the installation section executes only once, the
TSR can discard it after becoming resident.
u You can install ALARM any number of times in quick succession, each time
with a new alarm setting. The timer handler does not restore the original
vector for Interrupt 08 after the alarm sounds. In effect, the multiple
installations remain daisy-chained in memory. The address in OldTimer for
one installation is the address of NewTimer in the preceding installation.
u Until a system reboot, NewTimer remains in place as the Interrupt 08
handler, even after the alarm sounds. To save unnecessary activity, the byte
TimerActiveFlag remains set after the alarm sounds. This forces an
immediate jump to the original handler for all subsequent executions of
NewTimer.
u NewTimer and Sound alter registers DS, AX, BX, CX, DX, and ES. To
preserve the original values in these registers, the procedures first push them
onto the stack and then restore the original values before exiting. This
ensures that the process interrupted by NewTimer continues with valid
registers after NewTimer returns.
u ALARM requires little stack space. It assumes the current stack is adequate
and makes no attempt to set up a new one. More sophisticated TSRs,
however, should as a matter of course provide their own stacks to ensure
adequate stack depth. The example program presented in “Example of an
Advanced TSR: SNAP,” later in this chapter, demonstrates this safety
measure.

Using MS-DOS in Active TSRs

This section explains how to write active TSRs that can safely call MS-DOS
functions. The material explores the problems imposed by the nonreentrant
nature of MS-DOS and explains how a TSR can resolve those problems. The
solution consists of four parts:
u Understanding how MS-DOS uses stacks
u Determining when MS-DOS is active
u Determining whether a TSR can safely interrupt an active MS-DOS function
u Monitoring the Critical Error flag
Understanding MS-DOS Stacks

MS-DOS functions set up their own stacks, which makes them nonreentrant. If
a TSR interrupts an MS-DOS function and then executes another function that
sets up the same stack, the second function will overwrite everything placed on
the stack by the first function. The problem occurs when the second function
returns and the first is left with unusable stack data. A TSR that calls an MS-
DOS function must not interrupt any function that uses the same stack.
MS-DOS versions 2.0 and later use three internal stacks: an I/O stack, a disk
stack, and an auxiliary stack. The current stack depends on the MS-DOS
function. Functions 01 through 0Ch set up the I/O stack. Functions higher than
0Ch (with few exceptions) use the disk stack, as do Interrupts 25h and 26h.
MS-DOS normally uses the auxiliary stack only when it executes Interrupt 24h
(Critical Error Handler).
Determining MS-DOS Activity

A TSR’s handlers can determine when MS-DOS is active by consulting a 1-byte
flag called the InDos flag. Every MS-DOS function sets this flag upon entry and
clears it upon termination. During installation, a TSR locates the flag through
Function 34h (Get Address of InDos Flag), which returns the address as ES:BX.
The installation portion then stores the address so the handlers can later find the
flag without again calling Function 34h.
Theoretically, a TSR can wait to execute until the InDos flag is clear, thus
sidestepping the entire issue of interrupting MS-DOS. However, several low-
order functions — such as Function 0Ah (Get Buffered Keyboard Input) — wait
idly for an expected keystroke before they terminate. If a TSR were allowed to
execute only after MS-DOS returned, it too would have to wait for the
terminating event.

The solution lies in determining when the low-order functions 01 through 0Ch
are active. MS-DOS provides another service for this purpose: Interrupt 28h,
the Idle Interrupt.
Interrupting MS-DOS Functions

MS-DOS continually calls Interrupt 28h from the low-order polling functions as
they wait for keyboard input. This signal says that MS-DOS is idle and that a
TSR may interrupt provided it does not overwrite the I/O stack. Put another
way, a TSR can safely interrupt MS-DOS Functions 01 through 0Ch provided it
does not call them.
An active TSR that calls MS-DOS must monitor Interrupt 28h with a handler.
When the handler gains control, it checks the TSR request flag. If the flag
indicates the TSR has been requested and if system hardware is inactive, the
handler executes the TSR. Since control must eventually return to the idle MS-
DOS function which has stored data on the I/O stack, the TSR in this case must
not call any MS-DOS function that also uses the I/O stack. Table 11.1 shows
which functions set up the I/O stack for various versions of MS-DOS.
Table 11.1 MS-DOS Internal Stacks
Critical MS-DOS MS-DOS MS-DOS
Function Error flag 2.x 3.0 3.1+
01–0Ch Clear I/O* I/O I/O
Set Aux* Aux Aux
33h Clear Disk* Disk Caller*
Set Disk Disk Caller
50h–51h Clear I/O Caller Caller
Set Aux Caller Caller
59h Clear n/a* I/O Disk
Set n/a Aux Disk
5D0Ah Clear n/a n/a Disk
Set n/a n/a Disk
62h Clear n/a Caller Caller
Set n/a Caller Caller
All others Clear Disk Disk Disk
Set Disk Disk Disk
* I/O=I/O stack, Aux = auxiliary stack, Disk = disk stack, Caller = caller’s stack, n/a = function not
available.

TSRs that perform tasks of long or indefinite duration should themselves call
Interrupt 28h. For example, a TSR that polls for keyboard input should include
an INT 28h instruction in the polling loop, as shown here:
poll: int 28h ; Signal idle state
mov ah, 1
int 16h ; Key waiting?
jnz poll ; If not, repeat polling loop
sub ah, ah
int 16h ; Otherwise, get key
This courtesy gives other TSRs a chance to execute if the InDos flag happens to
be set.
Monitoring the Critical Error Flag

MS-DOS sets the Critical Error flag to a nonzero value when it detects a critical
error. It then invokes Interrupt 24h (Critical Error Handler) and clears the flag
when Interrupt 24h returns. MS-DOS functions higher than 0Ch are illegal
during critical error processing. Therefore, a TSR that calls MS-DOS must not
execute while the Critical Error flag is set.
MS-DOS versions 3.1 and later locate the Critical Error flag in the byte
preceding the InDos flag. A single call to Function 34h (Get Address of InDos
Flag) thus effectively returns the addresses of both flags. For earlier versions of
MS-DOS or for the compatibility version of MS-DOS in OS/2 1.x, a TSR must
call Function 34h and then scan the segment returned in the ES register for one
of the two following sequences of instructions:
; Sequence of instructions in DOS Versions 2.0 - 3.0
cmp ss:[CriticalErrorFlag], 0
jne @F
int 28h
; Sequence of instructions in DOS compatibility version for OS/2 1.x

test [CriticalErrorFlag], 0FFh
jnz @F
push ss:[ ? ]
int 28h
The question mark inside brackets in the preceding PUSH statement indicates
that the operand for the PUSH instruction can be any legal operand.
In either version of MS-DOS, the operand field in the first instruction gives the
flag’s offset. The value in ES determines the segment address. “Example of an
Advanced TSR: SNAP,” later in the chapter, presents a program that shows
how to locate the Critical Error flag with this technique.

Preventing Interference
This section describes how an active TSR can avoid interfering with the process
it interrupts. Interference occurs when a TSR commits an error or performs an
action that affects the interrupted process after the TSR returns. Examples of
interference range from relatively harmless, such as moving the cursor, to
serious, such as overrunning a stack.
Although a TSR can interfere with another process in many different ways,
protection against interference involves only three steps:
1. Recording a current configuration
2. Changing the configuration so it applies to the TSR
3. Restoring the original configuration before terminating
The example program described on page 293 demonstrates all the

noninterference safeguards described in this section. These safeguards by no
means exhaust the subject of noninterference. More sophisticated TSRs may
require more sophisticated methods. However, noninterference methods
generally fall into one of the following categories:
u Trapping errors
u Preserving an existing condition
u Preserving existing data
Trapping Errors
A TSR committing an error that triggers an interrupt must handle the interrupt to
trap the error. Otherwise, the existing interrupt routine, which belongs to the
underlying process, would attempt to service an error the underlying process did
not commit.
For example, a TSR that accepts keyboard input should include handlers for
Interrupts 23h and 1Bh to trap keyboard break signals. When MS-DOS detects
CTRL+C from the keyboard or input stream, it transfers control to Interrupt 23h
(CTRL+C Handler). Similarly, the BIOS keyboard routine calls Interrupt 1Bh
(CTRL+BREAK Handler) when it detects a CTRL+BREAK key combination. Both
routines normally terminate the current process.
A TSR that calls MS-DOS should also trap critical errors through Interrupt 24h
(Critical Error Handler). MS-DOS functions call Interrupt 24h when they
encounter certain hardware errors. The TSR must not allow the existing
interrupt routine to service the error, since the routine might allow the user to
abort service and return control to MS-DOS. This would terminate both the

TSR and the underlying process. By handling Interrupt 24h, the TSR retains
control if a critical error occurs.
An error-trapping handler differs in two ways from a TSR’s other handlers:
1. It is temporary, in service only while the TSR executes. At startup, the TSR
copies the handler’s address to the interrupt vector table; it then restores the
original vector before returning.
2. It provides complete service for the interrupt; it does not pass control on to
the original routine.
Error-trapping handlers often set a flag to let the TSR know the error has
occurred. For example, a handler for Interrupt 1Bh might set a flag when the
user presses CTRL+BREAK. The TSR can check the flag as it polls for keyboard
input, as shown here:
BrkHandler PROC FAR ; Handler for Interrupt 1Bh
.
.
.
mov cs:BreakFlag, TRUE ; Raise break flag
iret ; Terminate interrupt
BrkHandler ENDP
.
.
.
mov BreakFlag, FALSE ; Initialize break flag
poll: .
.
.
cmp BreakFlag, TRUE ; Keyboard break pressed?
je exit ; If so, break polling loop
mov ah, 1
int 16h ; Key waiting?
jnz poll ; If not, repeat polling loop
Preserving an Existing Condition

A TSR and its interrupt handlers must preserve register values so that all
registers are returned intact to the interrupted process. This is usually done by
pushing the registers onto the stack before changing them, then popping the
original values before returning.
Setting up a new stack is another important safeguard against interference. A
TSR should usually provide its own stack to avoid the possibility of overrunning
the current stack. Exceptions to this rule are simple TSRs such as the sample
program ALARM that make minimal stack demands.

A TSR that alters the video configuration should return the configuration to its
original state upon return. Video configuration includes cursor position, cursor
shape, and video mode. The services provided through Interrupt 10h enable a
TSR to determine the existing configuration and alter it if necessary.
However, some applications set video parameters by directly programming the
video controller. When this happens, BIOS remains unaware of the new
configuration and consequently returns inaccurate information to the TSR.
Unfortunately, there is no solution to this problem if the controller’s data
registers provide write-only access and thus cannot be queried directly. For
more information on video controllers, refer to Richard Wilton, Programmer’s
Guide to the PC & PS/2 Video Systems. (See “Books for Further Reading” in
the Introduction.)
Preserving Existing Data

A TSR requires its own disk transfer area (DTA) if it calls MS-DOS functions
that access the DTA. These include file control block functions and Functions
11h, 12h, 4Eh, and 4Fh. The TSR must switch to a new DTA to avoid
overwriting the one belonging to the interrupted process. On becoming active,
the TSR calls Function 2Fh to obtain the address of the current DTA. The TSR
stores the address and then calls Function 1Ah to establish a new DTA. Before
returning, the TSR again calls Function 1Ah to restore the address of the original
DTA.
MS-DOS versions 3.1 and later allow a TSR to preserve extended error
information. This prevents the TSR from destroying the original information if it
commits an MS-DOS error. The TSR retrieves the current extended error data
by calling MS-DOS Function 59h. It then copies registers AX, BX, CX, DX, SI,
DI, DS, and ES to an 11-word data structure in the order given. MS-DOS
reserves the last three words of the structure, which should each be set to zero.
Before returning, the TSR calls Function 5Dh with AL = 0Ah and DS:DX
pointing to the data structure. This call restores the extended error data to their
original state.
Communicating Through the Multiplex Interrupt

The Multiplex interrupt (Interrupt 2Fh) provides the Microsoft-approved way
for a program to verify the presence of an installed TSR and to exchange
information with it. MS-DOS version 2.x uses Interrupt 2Fh only as an interface
for the resident print spooler utility PRINT.COM. Later MS-DOS versions
standardize calling conventions so that multiple TSRs can share the interrupt.
A TSR chains to the Multiplex interrupt by setting up a handler. The TSR’s
installation code records the Interrupt 2Fh vector and then replaces it with the
address of the new multiplex handler.

The Multiplex Handler

A program communicates with a multiplex handler by calling Interrupt 2Fh with
an identity number in the AH register. As each handler in the chain gains
control, it compares the value in AH with its own identity number. If the handler
finds that it is not the intended recipient of the call, it passes control to the
previous handler. The process continues until control reaches the target handler.
When the target handler finishes its tasks, it returns via an IRET instruction to
terminate the interrupt.
The target handler determines its tasks from the function number in AL.
Convention reserves Function 0 as a request for installation status. A multiplex
handler must respond to Function 0 by setting AL to 0FFh, to inform the caller
of the handler’s presence in memory. The handler should also return other
information to provide a completely reliable identification. For example, it might
return in ES:BX a far pointer to the TSR’s copyright notice. This assures the
caller it has located the intended TSR and not another TSR that has already
claimed the identity number in AH.
Identity numbers range from 192 to 255, since MS-DOS reserves lesser values
for its own use. During installation, a TSR must verify the uniqueness of its
number. It must not set up a multiplex handler identified by a number already in
use. A TSR usually obtains its identity number through one of the following
methods:
u The programmer assigns the number in the program.
u The user chooses the number by entering it as an argument in the command
line, placing it into an environment variable, or by altering the contents of an
initialization file.
u The TSR selects its own number through a process of trial and error.
The last method offers the most flexibility. It finds an identity number not
currently in use among the installed multiplex handlers and does not require
intervention from the user.
To use this method, a TSR calls Interrupt 2Fh during installation with AH = 192
and AL = 0. If the call returns AL = 0FFh, the program tests other registers to
determine if it has found a prior installation of itself. If the test fails, the program
resets AL to zero, increments AH to 193, and again calls Interrupt 2Fh. The
process repeats with incrementing values in AH until the TSR locates a prior
installation of itself — in which case it should abort with an appropriate message
to the user — or until AL returns as zero. The TSR can then use the value in AH
as its identity number and proceed with installation.
The SNAP.ASM program in this chapter demonstrates how a TSR can use this
trial-and-error method to select a unique identity number. During installation, the
program calls Interrupt 2Fh to verify that SNAP is not already installed. When

deinstalling, the program again calls Interrupt 2Fh to locate the resident TSR in
memory. SNAP’s multiplex handler services the call and returns the address of
the resident code’s program-segment prefix. The calling program can then locate
the resident code and deinstall it, as explained in “Deinstalling a TSR,”
following.
Using the Multiplex Interrupt Under MS-DOS Version 2.x

A TSR can use the Multiplex interrupt under MS-DOS version 2.x, with certain
limitations. Under version 2.x, only MS-DOS’s print spooler PRINT, itself a
TSR program, provides an Interrupt 2Fh service. The Interrupt 2Fh vector
remains null until PRINT or another TSR is installed that sets up a multiplex
handler.
Therefore, a TSR running under version 2.x must first check the existing
Interrupt 2Fh vector before installing a multiplex handler. The TSR locates the
current Interrupt 2Fh handler through Function 35h (Get Interrupt Vector). If
the function returns a null vector, the TSR’s handler will be last in the chain of
Interrupt 2Fh handlers. The handler must terminate with an IRET instruction
rather than pass control to a nonexistent routine.
PRINT in MS-DOS version 2.x does not pass control to the previous handler. If
you intend to run PRINT under version 2.x, the program must be installed
before other TSRs that also handle Interrupt 2Fh. This places PRINT’s
multiplex handler last in the chain of handlers.
Deinstalling a TSR
A TSR should provide a means for the user to remove or “deinstall” it from
memory. Deinstallation returns occupied memory to the system, offering these
benefits:
u The freed memory becomes available to subsequent programs that may
require additional memory space.
u Deinstallation restores the system to a normal state. Thus, sensitive programs
that may be incompatible with TSRs can execute without the presence of
installed routines.
A deinstallation program must first locate the TSR in memory, usually by

requesting an address from the TSR’s multiplex handler. When it has located the
TSR, the deinstallation program should then compare addresses in the vector
table with the addresses of the TSR’s handlers. A mismatch indicates that
another TSR has chained a handler to the interrupt routine. In this case, the
deinstallation program should deny the request to deinstall. If the addresses of

the TSR’s handlers match those in the vector table, deinstallation can safely
continue.

You can deinstall the TSR with these three steps:

1. Restore to the vector table the original interrupt vectors replaced by the
handler addresses.
2. Read the segment address stored at offset 2Ch of the resident TSR’s
program segment prefix (PSP). This address points to the TSR’s
“environment block,” a list of environment variables that MS-DOS copies
into memory when it loads a program. Place the block’s address in the ES
register and call MS-DOS Function 49h (Release Memory Block) to return
the block’s memory to the operating system.
3. Place the resident PSP segment address in ES and again call Function 49h.
This call releases the block of memory occupied by the TSR’s code and
data.
The example program in the next section demonstrates how to locate a resident
TSR through its multiplex handler, and deinstall it from memory.
Example of an Advanced TSR: SNAP

This section presents SNAP, a memory-resident program that demonstrates
most of the techniques discussed in this chapter. SNAP takes a snapshot of the
current screen and copies the text to a specified file. SNAP accommodates
screens with various column and line counts, such as CGA’s 40-column mode
or VGA’s 50-line mode. The program ignores graphics screens.
Once installed, SNAP occupies approximately 7.5K of memory. When it detects
the ALT+LEFT SHIFT+S key combination, SNAP displays a prompt for a file
specification. The user can type a new filename, accept the previous filename
by pressing ENTER, or cancel the request by pressing ESC.
SNAP reads text directly from the video buffer and copies it to the specified file.
The program sets the file pointer to the end of the file so that text is appended
without overwriting previous data. SNAP copies each line only to the last
character, ignoring trailing spaces. The program adds a carriage return–linefeed
sequence (0D0Ah) to the end of each line. This makes the file accessible to any
text editor that can read ASCII files.
To demonstrate how a program accesses resident data through the Multiplex
interrupt, SNAP can reset the display attribute of its prompt box. After installing
SNAP, run the main program with the /C option to change box colors:
SNAP /Cxx
The argument xx specifies the desired attribute as a two-digit hexadecimal

number — for example, 7C for red on white, or 0F for monochrome high

intensity. For a list of color and monochrome display attributes, refer to the
“Tables” section of the Reference.
SNAP can deinstall itself, provided another TSR has not been loaded after it.
Deinstall SNAP by executing the main program with the /D option:
SNAP /D
If SNAP successfully deinstalls, it displays the following message:

TSR deinstalled
Building SNAP.EXE
SNAP combines four modules: SNAP.ASM, COMMON.ASM,
HANDLERS.ASM, and INSTALL.ASM. Source files are located on one of
your distribution disks. Each module stores temporary code and data in the
segments INSTALLCODE and INSTALLDATA. These segments apply only to
SNAP’s installation phase; MS-DOS recovers the memory they occupy when
the program exits through the terminate-and-stay-resident function. The
following briefly describes each module:
u SNAP.ASM contains the TSR’s main code and data.
u COMMON.ASM contains procedures used by other example programs.
u HANDLERS.ASM contains interrupt handler routines for Interrupts 08, 09,
10h, 13h, 15h, 28h, and 2Fh. It also provides simple error-trapping handlers
for Interrupts 1Bh, 23h, and 24h. Additional routines set up and deinstall the
handlers.
u INSTALL.ASM contains an exit routine that calls the terminate-and-stay-
resident function and a deinstallation routine that removes the program from
memory. The module includes error-checking services and a command-line
parser.
This building-block approach allows you to create other TSRs by replacing

SNAP.ASM and linking with the HANDLERS and INSTALL object modules.
The library of routines accommodates both keyboard-activated and time-
activated TSRs. A time-activated TSR is a program that activates at a
predetermined time of day, similar to the example program ALARM introduced
earlier in this chapter. The header comments for the Install procedure in
HANDLERS.ASM explain how to install a time-activated TSR.
You can write new TSRs in assembly language or any high-level language that
conforms to the Microsoft conventions for ordering segments. Regardless of the
language, the new code must not invoke an MS-DOS function that sets up the
I/O stack (see “Interrupting MS-DOS Functions,” earlier in this chapter). Code

in Microsoft C, for example, must not call getche or kbhit, since these
functions in turn call MS-DOS Functions 01 and 0Bh.
Code written in a high-level language must not check for stack overflows.
Compiler-generated stack probes do not recognize the new stack setup when the
TSR executes, and therefore must be disabled. The example program BELL.C,
included on disk with the TSR library routines, demonstrates how to disable
stack checking in Microsoft C using the check_stack pragma.
Outline of SNAP
The following sections outline in detail how SNAP works. Each part of the
outline covers a specific portion of SNAP’s code. Headings refer to earlier
sections of this chapter, providing cross-references to SNAP’s key procedures.
For example, the part of the outline that describes how SNAP searches for its
startup signal refers to the section “Auditing Hardware Events for TSR
Requests,” earlier in this chapter.
Figures 11.2 through 11.4 are flowcharts of the SNAP program. Each chart
illustrates a separate phase of SNAP’s operation, from installation through
memory-residency to deinstallation.

Figure 11.2 Flowchart for SNAP.EXE: Installation Phase

Figure 11.3 Flowchart for SNAP.EXE: Resident Phase

Figure 11.4 Flowchart for SNAP.EXE: Deinstallation Phase
Refer to the flowcharts as you read the following outline. They will help you
maintain perspective while exploring the details of SNAP’s operation. Text in
the outline cross-references the charts.
Note that information in both the outline and the flowcharts is generic. Except
for references to the SNAP procedure, all descriptions in the outline and the

flowcharts apply to any TSR created with the HANDLERS and INSTALL
modules.
Auditing Hardware Events for TSR Requests

To search for its startup signal, SNAP audits the keyboard with an interrupt
handler for either Interrupt 09 (keyboard) or Interrupt 15h (Miscellaneous
System Services). The Install procedure determines which of the two
interrupts to handle based on the following code:
.IF HotScan == 0 ; If valid scan code given:
mov ah, HotShift ; AH = hour to activate
mov al, HotMask ; AL = minute to activate
call GetTimeToElapse ; Get number of 5-second intervals
mov CountDown, ax ; to elapse before activation
.ELSE ; Force use of KeybrdMonitor as

; keyboard handler
cmp Version, 031Eh ; DOS Version 3.3 or higher?
jb setup ; No? Skip next step
; Test for IBM PS/2 series. If not PS/2, use Keybrd and
; SkipMiscServ as handlers for Interrupts 09 and 15h
; respectively. If PS/2 system, set up KeybrdMonitor as the
; Interrupt 09 handler. Audit keystrokes with MiscServ
; handler, which searches for the hot key by handling calls
; to Interrupt 15h (Miscellaneous System Services). Refer to
; Section 11.2.1 for more information about keyboard handlers.
mov ax, 0C00h ; Function 0Ch (Get System

int 15h ; Configuration Parameters)
sti ; Compaq ROM may leave disabled
jc setup ; If carry set,

or ah, ah ; or if AH not 0,
jnz setup ; services are not supported
; Test bit 4 to see if Intercept is implemented

test BYTE PTR es:[bx+5], 00010000y
jz setup
; If so, set up MiscServ as Interrupt 15h handler

mov ax, OFFSET MiscServ
mov WORD PTR intMisc.NewHand, ax
.ENDIF
; Set up KeybrdMonitor as Interrupt 09 handler

mov ax, OFFSET KeybrdMonitor
mov WORD PTR intKeybrd.NewHand, ax
The following describes the code’s logic:

u If the program is running under MS-DOS version 3.3 or higher and if

Interrupt 15h supports Function 4Fh, set up handler MiscServ to search for
the hot key. Handle Interrupt 09 with KeybrdMonitor only to maintain the
keyboard active flag.
u Otherwise, set up a handler for Interrupt 09 to search for the hot key. Handle
calls to Interrupt 15h with the routine SkipMiscServ, which contains this
single instruction:
jmp cs:intMisc.OldHand
The jump immediately passes control to the original Interrupt 15h routine;
thus, SkipMiscServ has no effect. It serves only to simplify coding in
other parts of the program.
At each keystroke, the keyboard interrupt handler (either Keybrd or

MiscServ) calls the procedure CheckHotKey with the scan code of the
current key. CheckHotKey compares the scan code and shift status with the
bytes HotScan and HotShift. If the current key matches, CheckHotKey
returns the carry flag clear to indicate that the user has pressed the hot key.
If the keyboard handler finds the carry flag clear, it sets the flag
TsrRequestFlag and exits. Otherwise, the handler transfers control to the
original interrupt routine to service the interrupt.
The timer handler Clock reads the request flag at every occurrence of the timer
interrupt. Clock takes no action if it finds a zero value in TsrRequestFlag.
Figures 11.1 and 11.3 depict the relationship between the keyboard and timer
handlers.
Monitoring System Status

Because SNAP produces output to both video and disk, it avoids interrupting
either video or disk operations. The program uses interrupt handlers Video and
DiskIO to monitor Interrupts 10h (video) and 13h (disk). SNAP also avoids
interrupting keyboard use. The instructions at the far label KeybrdMonitor
serve as the monitor handler for Interrupt 09 (keyboard).
The three handlers perform similar functions. Each sets an active flag and then
calls the original routine to service the interrupt. When the service routine
returns, the handler clears the active flag to indicate that the device is no longer
in use.

The BIOS Interrupt 13h routine clears or sets the carry flag to indicate the
operation’s success or failure. DiskIO therefore preserves the flags register
when returning, as shown here:
DiskIO PROC FAR
mov cs:intDiskIO.Flag, TRUE ; Set active flag
; Simulate interrupt by pushing flags and far-calling old
; Int 13h routine
pushf
call cs:intDiskIO.OldHand
; Clear active flag without disturbing flags register
mov cs:intDiskIO.Flag, FALSE
; Simulate IRET without popping flags (since services use
; carry flag)
ret 2
DiskIO ENDP
The terminating RET 2 instruction discards the original flags from the stack
when the handler returns.
Determining Whether to Invoke the TSR

The procedure CheckRequest determines whether the TSR:
u Has been requested.
u Can safely interrupt the system.
Each time it executes, the timer handler Clock calls CheckRequest to read
the flag TsrRequestFlag. If CheckRequest finds the flag set, it scans other
flags maintained by the TSR’s interrupt handlers and by MS-DOS. These flags
indicate the current system status. As the flowchart in Figure 11.3 shows,
CheckRequest calls CheckDos (described following) to determine the status
of the operating system. CheckRequest then calls CheckHardware to check
hardware status.
CheckHardware queries the interrupt controller to determine if any device is
currently being serviced. It also reads the active flags maintained by the
KeybrdMonitor, Video, and DiskIO handlers. If the controller, keyboard,
video, and disk are all inactive, CheckHardware clears the carry flag and
returns.
CheckRequest indicates system status with the carry flag. If the procedure
returns the carry flag set, the caller exits without invoking the TSR. A clear
carry signals that the caller can safely execute the TSR.

Determining MS-DOS Activity

As Figure 11.2 shows, the procedure GetDosFlags locates the InDos flag
during SNAP’s installation phase. GetDosFlags calls Function 34h (Get
Address of InDos Flag) and then stores the flag’s address in the far pointer
InDosAddr.
When called from the CheckRequest procedure, CheckDos reads InDos to

determine whether the operating system is active. Note that CheckDos reads
the flag directly from the address in InDosAddr. It does not call Function 34h
to locate the flag, since it has not yet established whether MS-DOS is active.
This follows from the general rule that interrupt handlers must not call any MS-
DOS function.
The next two sections more fully describe the procedure CheckDos.
Interrupting MS-DOS Functions

Figure 11.3 shows that the call to CheckDos can initiate either from Clock
(timer handler) or Idle (Interrupt 28h handler). If CheckDos finds the InDos
flag set, it reacts in different ways, depending on the caller:
u If called from Clock, CheckDos cannot know which MS-DOS function is
active. In this case, it returns the carry flag set, indicating that Clock must
deny the request for the TSR.
u If called from Idle, CheckDos assumes that one of the low-order polling
functions is active. It therefore clears the carry flag to let the caller know the
TSR can safely interrupt the function.
For more information on this topic, see the section “Interrupting MS-DOS
Functions,” earlier in this chapter.
Monitoring the Critical Error Flag

The procedure GetDosFlags (Figure 11.2) determines the address of the
Critical Error flag. The procedure stores the flag’s address in the far pointer
CritErrAddr.
When called from either the Clock or Idle handlers, CheckDos reads the
Critical Error flag. A nonzero value in the flag indicates that the Critical Error
Handler (Interrupt 24h) is processing a critical error and the TSR must not
interrupt. In this case, CheckDos sets the carry flag and returns, causing the
caller to exit without executing the TSR.
Trapping Errors
As Figure 11.3 shows, Clock and Idle invoke the TSR by calling the
procedure Activate. Before calling the main body of the TSR, Activate
sets up the following handlers:

Handler Name For Interrupt Receives Control When

CtrlBreak 1Bh (CTRL+BREAK Handler) CTRL+BREAK sequence entered at
keyboard
CtrlC 23h (CTRL+C Handler) MS-DOS detects a CTRL+C sequence
from the keyboard or input stream
CritError 24h (Critical Error Handler) MS-DOS encounters a critical error
These handlers trap keyboard break signals and critical errors that would
otherwise trigger the original handler routines. The CtrlBreak and CtrlC
handlers contain a single IRET instruction, thus rendering a keyboard break
ineffective. The CritError handler contains the following instructions:
CritError PROC FAR
sti
sub al, al ; Assume DOS 2.x
; Set AL = 0 for ignore error
.IF cs:major != 2 ; If DOS 3.x, set AL = 3
mov al, 3 ; DOS call fails
.ENDIF
iret
CritError ENDP
The return code in AL stops MS-DOS from taking further action when it
encounters a critical error.
As an added precaution, Activate also calls Function 33h (Get or Set
CTRL+BREAK Flag)to determine the current setting of the checking flag.
Activate stores the setting, then calls Function 33h again to turn off break
checking.
When the TSR’s main procedure finishes its work, it returns to Activate,
which restores the original setting for the checking flag. It also replaces the
original vectors for Interrupts 1Bh, 23h, and 24h.
SNAP’s error-trapping safeguards enable the TSR to retain control in the event
of an error. Pressing CTRL+BREAK or CTRL+C at SNAP’s prompt has no effect. If
the user specifies a nonexistent drive — a critical error — SNAP merely beeps the
speaker and returns normally.
Preserving an Existing Condition

Activate records the stack pointer SS:SP in the doubleword OldStackAddr.
The procedure then resets the pointer to the address of a new stack before
calling the TSR. Switching stacks ensures that SNAP has adequate stack depth
while it executes.
The label NewStack points to the top of the new stack buffer, located in the
code segment of the HANDLERS.ASM module. The equate constant

STACK_SIZ determines the size of the stack. The include file TSR.INC
contains the declaration for STACK_SIZ.
Activate preserves the values in all registers by pushing them onto the new
stack. It does not push DS, since that register is already preserved in the Clock
or Idle handler.
SNAP does not alter the application’s video configuration other than by moving
the cursor. Figure 11.3 shows that Activate calls the procedure Snap, which
executes Interrupt 10h to determine the current cursor position. Snap stores the
row and column in the word OldPos. The procedure restores the cursor to its
original location before returning to Activate.
Preserving Existing Data

Because SNAP does not call an MS-DOS function that writes to the DTA, it
does not need to preserve the DTA belonging to the interrupted process.
However, the code for switching and restoring the DTA is included within
IFDEF blocks in the procedure Activate. The equate constant DTA_SIZ,
declared in the TSR.INC file, governs the assembly of the blocks as well as the
size of the new DTA.
It is possible for SNAP to overwrite existing extended error information by
committing a file error. The program does not attempt to preserve the original
information by calling Functions 59h and 5Dh. In certain rare instances, this
may confuse the interrupted process after SNAP returns.
Communicating Through the Multiplex Interrupt

The program uses the Multiplex interrupt (Interrupt 2Fh) to
u Verify that SNAP is installed.
u Select a unique multiplex identity number.
u Locate resident data.
For more information about Interrupt 2Fh, see the section “Communicating
through the Multiplex Interrupt,” earlier in this chapter.
SNAP accesses Interrupt 2Fh through the procedure CallMultiplex, as
shown in Figures 11.2 and 11.4. By searching for a prior installation,
CallMultiplex ensures that SNAP is not installed more than once. During
deinstallation, CallMultiplex locates data required to deinstall the resident
TSR.
The procedure Multiplex serves as SNAP’s multiplex handler. When it
recognizes its identity number in AH, Multiplex determines its tasks from the
function number in the AL register. The handler responds to Function 0 by

returning AL equalling 0FFh and ES:DI pointing to an identifier string unique to

SNAP.
CallMultiplex searches for the handler by invoking Interrupt 2Fh in a loop,
beginning with a trial identity number of 192 in AH. At the start of each iteration
of the loop, the procedure sets AL to zero to request presence verification from
the multiplex handler. If the handler returns 0FFh in AL, CallMultiplex
compares its copy of SNAP’s identifier string with the text at memory location
ES:DI. A failed match indicates that the multiplex handler servicing the call is
not SNAP’s handler. In this case, CallMultiplex increments AH and cycles
back to the beginning of the loop.
The process repeats until the call to Interrupt 2Fh returns a matching identifier
string at ES:DI, or until AL returns as zero. A matching string verifies that
SNAP is installed, since its multiplex handler has serviced the call. A return
value of zero indicates that SNAP is not installed and that no multiplex handler
claims the trial identity number in AH. In this case, SNAP assigns the number to
its own handler.
Deinstalling a TSR
During deinstallation, CallMultiplex locates SNAP’s multiplex handler as
described previously. The handler Multiplex receives the verification request
and returns in ES the code segment of the resident program.
Deinstall reads the addresses of the following interrupt handlers from the
data structure in the resident code segment:
Handler Name Description
Clock Timer handler
Keybrd Keyboard handler (non-PS/2)
KeybrdMonitor Keyboard monitor handler (PS/2)
Video Video monitor handler
DiskIO Disk monitor handler
SkipMiscServ Miscellaneous Systems Services handler (non-PS/2)
MiscServ Miscellaneous Systems Services handler (PS/2)
Idle MS-DOS Idle handler
Multiplex Multiplex handler
Deinstall calls MS-DOS Function 35h (Get Interrupt Vector) to retrieve the
current vectors for each of the listed interrupts. By comparing each handler
address with the corresponding vector, Deinstall ensures that SNAP can be
safely deinstalled. Failure in any of the comparisons indicates that another TSR
has been installed after SNAP and has set up a handler for the same interrupt. In

this case, Deinstall returns an error code, stopping the program with the
following message:
Can’t deinstall TSR
If all addresses match, Deinstall calls Interrupt 2Fh with SNAP’s identity
number in AH and AL set to 1. The handler Multiplex responds by returning
in ES the address of the resident code’s PSP. Deinstall then calls MS-DOS
Function 25h (Set Interrupt Vector) to restore the vectors for the original service
routines. This is called “unhooking” or “unchaining” the interrupt handlers.
After unhooking all of SNAP’s interrupt handlers, Deinstall returns with AX
pointing to the resident code’s PSP. The procedure FreeTsr then calls MS-
DOS Function 49h (Release Memory) to return SNAP’s memory to the
operating system. The program ends with the message
TSR deinstalled
to indicate a successful deinstallation.

Deinstalling SNAP does not guarantee more available memory space for the
next program. If another TSR loads after SNAP but handles interrupts other
than 08, 09, 10h, 13h, 15h, 28h, or 2Fh, SNAP still deinstalls properly. The
result is a harmless gap of deallocated memory formerly occupied by SNAP.
MS-DOS can use the free memory to store the next program’s environment
block. However, MS-DOS loads the program itself above the still-resident TSR.

307
C H A P T E R 1 2
Mixed-Language Programming
Mixed-language programming allows you to combine the unique strengths of

Microsoft Basic, C, C++, and FORTRAN with your assembly-language
routines. Any one of these languages can call MASM routines, and you can call
any of these languages from within your assembly-language programs. This
makes virtually all the routines from high-level–language libraries available to a
mixed-language program.
MASM 6.1 provides mixed-language features similar to those in high-level
languages. For example, you can use the INVOKE directive to call high-level-
language procedures, and the assembler handles the argument-passing details for
you. You can also use H2INC to translate C header files to MASM include files,
as explained in Chapter 20 of Environment and Tools.
The mixed-language features of MASM 6.1 do not make older methods of
defining mixed-language interfaces obsolete. In most cases, mixed-language
programs written with earlier versions of MASM will assemble and link correctly
under MASM 6.1. (For more information, see Appendix A.)
This chapter explains how to write assembly routines that can be called from
high-level–language modules and how to call high-level language routines from
MASM. You should already understand the languages you want to combine and
should know how to write, compile, and link multiple-module programs with
these languages.
This chapter covers only assembly-language interface with C, C++, Basic, and
FORTRAN; it does not cover mixed-language programming between high-level
languages. The focus here is the Microsoft versions of C, C++, Basic, and
FORTRAN, but the same principles apply to other languages and compilers.
Many of the techniques used in this chapter are explained in the material in
Chapter 7 on writing procedures in assembly language, and in Chapter 8 on
multiple-module programming.
The first section of this chapter discusses naming and calling conventions. The
next section, “Writing an Assembly Procedure for a Mixed-Language Program,”
provides a template for writing an assembly-language procedure that can be

called from another module written in a high-level language. This represents the
essence of mixed-language programming. Assembly language is often used for
creating fast secondary routines in a large program written in a high-level
language.
The third section describes specific conventions for linking assembly-language
procedures with modules in C, C++, Basic, and FORTRAN. These language-
specific sections also provide details on how the language manages various data
structures so that your MASM programs are compatible with the data from the
high-level language.
Naming and Calling Conventions

Each language has its own set of conventions, which fall into two categories:
u The “naming convention” specifies how or if the compiler or assembler alters
the name of an identifier before placing it into an object file.
u The “calling convention” determines how a language implements a call to a
procedure and how the procedure returns to the caller.
MASM supports several different conventions. The assembler uses C

convention when you specify a language type (langtype) of C, and Pascal
convention for language types PASCAL, BASIC, or FORTRAN. To the
assembler, the keywords BASIC, PASCAL, and FORTRAN are synonymous.
MASM also supports the SYSCALL and STDCALL conventions, which mix
elements of the C and Pascal conventions.
MASM gives you several ways to set the naming and calling conventions in your
assembly-language program. Using .MODEL with a langtype sets the default
for the module. This can also be done with the OPTION directive. This is
equivalent to the /Gc or /Gd option from the command line. Procedure
prototypes and declarations can specify a langtype to override the default.
When you write mixed-language routines, the easiest way to ensure convention
compatibility is to adopt the conventions of the called procedure’s language.
However, Microsoft languages can change the naming and calling conventions
for different procedures. If your program must call a procedure that uses an
argument-passing method different from that of the default language, prototype
the procedure first with the desired language type. This tells the assembler to
override the conventions of the default language and assume the proper
conventions for the prototyped procedure. “The MASM/High-Level–Language
Interface” section in this chapter explains how to change the default
conventions. The following sections provide more detail on the information
summarized in Table 12.1.

Chapter 12 Mixed-Language Programming 309
Table 12.1 Naming and Calling Conventions

Convention C SYSCALL STDCALL BASIC FORTRAN PASCAL
Leading X X
underscore
Capitalize all X X X
Arguments X X X
pushed left to
right
Arguments X X X
pushed right to
left
Caller stack X X *
cleanup
:VARARG X X X
allowed
* The STDCALL language type uses caller stack cleanup if the :VARARG parameter is used. Otherwise, the called routine must
clean up the stack.
Naming Conventions
“Naming convention” refers to the way a compiler or assembler stores the
names of identifiers. The first two rows of Table 12.1 show how each language
type affects symbol names. SYSCALL leaves symbol names as they appear in
the source code, but C and STDCALL add an underscore prefix. PASCAL,
BASIC, and FORTRAN change symbols to all uppercase.
The following list describes how these naming conventions affect a variable
called Big Time in your source code:
Langtype Specified Characteristics
SYSCALL Leaves the name unmodified. The linker sees the variable as
Big Time.
C, STDCALL The assembler (or compiler) adds a leading underscore to the
name, but does not change case. The linker sees the variable as
_Big Time.
PASCAL, FORTRAN, Converts all names to uppercase. The linker sees the variable as
BASIC Big Time.
The C Calling Convention

Specify the C language type for assembly-language procedures called from
programs that assume the C calling convention. Note that such programs are not
necessarily written in C, since other languages can mimic C conventions.

Argument Passing
With the C calling convention, the caller pushes arguments from right to left as
they appear in the caller’s argument list. The called procedure returns without
removing the arguments from the stack. It is the caller’s responsibility to clean
the stack after the call, either by popping the arguments or by adding an
appropriate value to the stack pointer SP.
Register Preservation
The called routine must return with the original values in BP, SI, DI, DS, and
SS. It must also preserve the direction flag.
Varying Number of Arguments

The additional overhead of cleaning the stack after each call has compensations.
It frees the caller from having to pass a set number of arguments to the called
procedure each time. Because the first argument in the list is always the last one
pushed, it is always on the top of the stack. Thus, it has the same address
relative to the frame pointer, regardless of how many arguments were actually
passed.
For example, consider the C library function printf, which accepts different
numbers of arguments. A C program calls the function like this:
printf( "Numbers: %f %f %.2f\n", n1, n2, n3 );
printf( "Also: %f", n4 );
The first line passes four arguments (including the string in quotes) and the
second line passes only two arguments. Notice that printf has no reliable way of
determining how many arguments the caller has pushed. Therefore, the function
returns without adjusting the stack. The C calling convention requires the caller
to take responsibility for removing the arguments from the stack, since only the
caller knows how many arguments it passed.
Use INVOKE to call a C-callable function from your assembly-language
program, since INVOKE automatically generates the necessary stack-cleaning
code after the call. You must also prototype the function with the VARARG
keyword if appropriate, as explained in “Procedures,” Chapter 7. Similarly,
when you write a C-callable procedure that accepts a varying number of
arguments, include VARARG in the procedure’s PROC statement.
The Pascal Calling Convention

By default, the langtype for FORTRAN, BASIC, and PASCAL selects the
Pascal calling convention. This convention pushes arguments left to right so that
the last argument is lowest on the stack, and it requires that the called routine
remove arguments from the stack.

Argument Passing
Arguments are placed on the stack in the same order in which they appear in the
source code. The first argument is highest in memory (because it is also the first
argument to be placed on the stack), and the stack grows downward.
A routine that uses the Pascal calling convention must preserve SI, DI, BP, DS,
and SS. For 32-bit code, the EBX, ES, FS, and GS registers must be preserved
as well as EBP, ESI, and EDI. The direction flag is also cleared upon entry and
must be preserved.

Passing a variable number of arguments is not possible with the Pascal calling
convention.
The STDCALL and SYSCALL Calling Conventions

A STDCALL procedure adopts the C name and calling conventions when
prototyped with the VARARG keyword. Refer to the section “Declaring
Parameters with the PROC Directive” in Chapter 7. Without VARARG, the
procedure uses the C naming and Pascal calling conventions. STDCALL
provides compatibility with 32-bit versions of Microsoft compilers.
As Table 12.1 shows, SYSCALL is identical to the C calling convention, but
does not add an underscore prefix to symbols.
Argument Passing
Argument passing order for both STDCALL and SYSCALL is the same as the
C calling convention. The caller pushes the arguments from right to left and
must remove the parameters from the stack after the call. However, STDCALL
requires the called procedure to clean the stack if the procedure does not accept
a variable number of arguments.
Both conventions require the called procedure to preserve the registers BP, SI,
DI, DS, and SS. Under STDCALL, the direction flag is clear on entry and must
be returned clear.

SYSCALL allows a variable number of arguments in the same way as the C
calling convention. STDCALL also mimics the C convention when VARARG
appears in the called procedure’s declaration or definition. It allows a varying
number of arguments and requires the caller to clean the stack. If not declared

or defined with VARARG, the called procedure does not accept a variable
argument list and must clean the stack before it returns.
Writing an Assembly Procedure

For a Mixed-Language Program
MASM 6.1 simplifies the coding required for linking MASM routines to high-
level– language routines. You can use the PROTO directive to write procedure
prototypes, and the INVOKE directive to call external routines. MASM
simplifies procedure-related tasks in the following ways:
u The PROTO directive improves error checking on argument types.
u INVOKE pushes arguments onto the stack and converts argument types to
types expected when possible. These arguments can be referenced by their
parameter label, rather than as offsets of the stack pointer.
u The LOCAL directive following the PROC statement saves places on the
stack for local variables. These variables can also be referenced by name,
rather than as offsets of the stack pointer.
u PROC sets up the appropriate stack frame according to the processor mode.
u The USES keyword preserves registers given as arguments.
u The C calling conventions specified in the PROC syntax allow for a variable
number of arguments to be passed to the procedure.
u The RET keyword adjusts the stack upward by the number of bytes in the
argument list, removes local variables from the stack, and pops saved
registers.
u The PROC statement lists parameter names and types. The parameters can
be referenced by name inside the procedure.
The complete syntax and parameter descriptions for these procedure directives
are explained in “Procedures” in Chapter 7. This section provides a template
that you can use for writing a MASM routine to be called from a high-level
language.
The template looks like this:
Label PROC [[distance langtype visibility <prologueargs> USES reglist
parmlist]]
LOCAL varlist
.
.
.

RET
Label ENDP
Replace the italicized words with appropriate keywords, registers, or variables as
defined by the syntax in “Declaring Parameters with the PROC Directive” in
Chapter 7.
The distance (NEAR or FAR) and visibility (PUBLIC, PRIVATE, or
EXPORT) that you give in the procedure declaration override the current
defaults. In some languages, the model can also be specified with command-line
options.
The langtype determines the calling convention for accessing arguments and
restoring the stack. For information on calling conventions, see “Naming and
Calling Conventions” earlier in this chapter.
The types for the parameters listed in the parmlist must be given. Also, if any of
the parameters are pointers, the assembler does not generate code to get the
value of the pointer references. You must write this code yourself. An example
of how to write such code is provided in “Declaring Parameters with the PROC
Directive” in Chapter 7.
If you need to code your own stack-frame setup manually, or if you do not
want the assembler to generate the standard stack setup and cleanup, see
“Passing Arguments on the Stack” and “User-Defined Prologue and Epilogue
Code” in Chapter 7.
The MASM/High-Level–Language Interface

Since high-level–language programs require initialization, you must write the
main routine of a mixed-language program in the high-level language, or link
with the startup code supplied by the high-level–language compiler. This gives
the assembly code access to high-level routines or library functions. The next
section explains how to link an assembly-language program with C-language
startup code.
For procedures with prototypes, INVOKE makes calls from MASM to high-
level–
language programs, much like procedure or function calls in the high-level
language. INVOKE calls procedures and generates the code to push arguments
in the order specified by the procedure’s calling convention, and to remove
arguments from the stack at the end of the procedure.
INVOKE can also do type checking and data conversion for the argument types
so that the procedure receives compatible data. For explanations of how to write
procedure prototypes and several examples of procedure declarations and the
corresponding prototypes, see “Declaring Procedure Prototypes” in Chapter 7.

For programs that mix assembly language and C, the H2INC utility makes it
easy to write prototypes and data declarations for the C procedures you want to
call from MASM. H2INC translates the C prototypes and declarations into the
corresponding MASM prototypes and declarations, which INVOKE can use to
call the procedure. The use of H2INC is explained in Chapter 20 in
Environment and Tools.
Mixed-language programming also allows the main program or a routine to use
external data — data defined in the other module. External data is the data that is
stored in a set place in memory (unlike dynamic and local data, which is
allocated on the stack and heap) and is visible to other modules.
External data is shared by all routines. One of the modules must define the static
data, which causes the compiler to allocate storage for the data. The other
modules that access the data must declare the data as external.
Argument Passing
Each language has its own convention for how an argument is actually passed. If
the argument-passing conventions of your routines do not agree, then a called
routine receives bad data. Microsoft languages support three different methods
for passing an argument:
u Near reference. Passes a variable’s near (offset) address, expressed as an
offset from the default data segment. This method gives the called routine
direct access to the variable itself. Any change the routine makes to the
parameter is reflected in the calling routine.
u Far reference. Passes a variable’s far (segmented) address. Though slower
than passing a near reference, this method is necessary for passing data that
lies outside the default data segment. (This is not an issue in Basic unless you
have specifically requested far memory.)
u Value. Passes only a copy of the variable, not its address. With this method,
the called routine gets a copy of the argument on the stack, but has no access
to the original variable. The copy is discarded when the routine returns, and
the variable retains its original value.
When you pass arguments between routines written in different languages, you
must ensure that the caller and the called routine use the same conventions for
passing and receiving arguments. In most cases, you should check the argument-
passing defaults used by each language and make any necessary adjustments.
Most languages have features that allow you to change argument-passing
methods.

A procedure called from any high-level language should preserve the direction
flag and the values of BP, SI, DI, SS, and DS. Routines called from MASM
must not alter SI, DI, SS, DS, or BP.
Pushing Addresses
Microsoft high-level languages push segment addresses before offsets. This lets
the called routine use the LES and LDS instructions to read far addresses from
the stack. Furthermore, each word of an argument is placed on the stack in
order of significance. Thus, the high word of a long integer is pushed first,
followed by the low word.
Array Storage
Most high-level-language compilers store arrays in row-major order. This means
that all elements of a row are stored consecutively. The first five elements of an
array with four rows and three columns are stored in row-major order as
A[1, 1], A[1, 2], A[1, 3], A[2, 1], A[2, 2]
In column-major order, the column elements are stored consecutively. For

example, this same array would be stored in column-major order as
A[1, 1], A[2, 1], A[3, 1], A[4, 1], A[1, 2], A[2, 2]
The C/MASM Interface

This section summarizes the characteristics of the interface between MASM and
Microsoft C and QuickC compilers. With the default naming and calling
convention, the assembler (or compiler) pushes arguments right to left and adds
a leading underscore to routine names.
Compatible Data Types

This list shows the 16-bit C data types and equivalent data types in MASM 6.1.
For 32-bit C compilers, int and unsigned int are equivalent to the MASM types
SDWORD and DWORD, respectively.
C Type Equivalent MASM Type
unsigned char BYTE
char SBYTE
unsigned short, unsigned int WORD
int, short SWORD
unsigned long DWORD
long SDWORD

float REAL4
double REAL8
long double REAL10
Naming Restrictions
C is case-sensitive and does not convert names to uppercase. Since C normally
links with the /NOI command-line option, you should assemble MASM modules
with the /Cx or /Cp option to prevent the assembler from converting names to
uppercase.

Argument-Passing Defaults
C always passes arrays by reference and all other variables (including structures)
by value. C programs in tiny, small, and medium model pass near addresses for
arrays, unless another distance is specified. Compact-, large-, and huge-model
programs pass far addresses by default. To pass by reference a variable type
other than array, use the C-language address-of operator (&).
If you need to pass an array by value, declare the array as a structure member
and pass a copy of the entire structure. However, this practice is rarely
necessary and usually impractical except for very small arrays, since it can make
substantial demands on stack space. If your program must maintain an array
through a procedure call, create a temporary copy of the array in heap and
provide the copy to the procedure by reference.
Changing the Calling Convention

Put _pascal or _fortran in the C function declaration to specify the Pascal
calling convention.
Array Storage
Array declarations give the number of elements. A1[a][b] declares a two-
dimensional array in C with a rows and b columns. By default, the array’s
lower bound is zero. Arrays are stored by the compiler in row-major order. By
default, passing arrays from C passes a pointer to the first element of the array.
String Format
C stores strings as arrays of bytes and uses a null character as the end-of-string
delimiter. For example, consider the string declared as follows:
char msg[] = "string of text"
The string occupies 15 bytes of memory as:
Figure 12.1 C String Format
Since msg is an array of characters, it is passed by reference.
External Data
In C, the extern keyword tells the compiler that the data or function is external.
You can define a static data object in a C module by defining a data object
outside all functions and subroutines. Do not use the static keyword in C with a
data object that you want to be public.

Structure Alignment
By default, C uses word alignment (unpacked storage) for all data objects longer
than 1 byte. This storage method specifies that occasional bytes may be added
as padding, so that word and doubleword objects start on an even boundary. In
addition, all nested structures and records start on a word boundary. MASM
aligns on byte boundaries by default.
When converting .H files with H2INC, you can use the /Zp command-line
option to specify structure alignment. If you do not specify the /Zp option,
H2INC uses word-alignment. Without H2INC, set the alignment to 2 when
declaring the MASM structure, compile the C module with /Zp1, or assemble
the MASM module with /Zp2.
Compiling and Linking

Use the same memory model for both C and MASM.
Returning Values
The assembler returns simple data types in registers. Table 12.2 shows the
register conventions for returning simple data types to a C program.
Table 12.2 Register Conventions for Simple Return Values
Data Type Registers
char AL
short, near, int (16-bit) AX
short, near, int (32-bit) EAX
long, far (16-bit) High-order portion (or segment address) in DX;
low-order portion (or offset address) in AX
long, far (32-bit) High-order portion (or segment address) in EDX;
low-order portion (or offset address) in EAX
Procedures using the C calling convention and returning type float or type
double store their return values into static variables. In multi-threaded
programs, this could mean that the return value may be overwritten. You can
avoid this by using the Pascal calling convention for multi-threaded programs so
float or double values are passed on the stack.
Structures less than 4 bytes long are returned in DX:AX. To return a longer
structure from a procedure that uses the C calling convention, you must copy
the structure to a global variable and then return a pointer to that variable in the
AX register (DX:AX, if you compiled in compact, large, or huge model or if the
variable is declared as a far pointer).

Structures, Records, and User-Defined Data Types

You can pass structures, records, and user-defined types as arguments by value
or by reference.
Writing Procedure Prototypes

The H2INC utility simplifies the task of writing prototypes for the C functions
you want to call from MASM. The C prototype converted by H2INC into a
MASM prototype allows INVOKE to correctly call the C function. Here are
some examples of C functions and the MASM prototypes created with H2INC.
/* Function Prototype Declarations to Convert with H2INC */
long checktypes (
char *name,
unsigned char a,
int b,
float d,
unsigned int *num );
my_func (float fNum, unsigned int x);
extern my_func1 (char *argv[]);
struct videoconfig _far * _far pascal my_func2 (int, scri );
For these C prototypes, H2INC generates this code:

@proto_0 TYPEDEF PROTO C :PTR SBYTE, :BYTE,
:SWORD, :REAL4, :PTR WORD
checktypes PROTO @proto_0
@proto_1 TYPEDEF PROTO C :REAL4, :WORD

my_func PROTO @proto_1
@proto_2 TYPEDEF PROTO C :PTR PTR SBYTE

my_func1 PROTO @proto_2
@proto_3 TYPEDEF PROTO FAR PASCAL :SWORD, :scri

my_func2 PROTO @proto_3

Example
As shown in the following short example, the main module (written in C) calls
an assembly routine, Power2.
#include <stdio.h>
extern int Power2( int factor, int power );
void main()
{
printf( "3 times 2 to the power of 5 is %d\n", Power2( 3, 5 ) );
}

Figure 12.2 shows how functions that observe the C calling convention use the
stack frame.
Figure 12.2 C Stack Frame

The MASM module that contains the Power2 routine looks like this:
.MODEL small, c
Power2 PROTO C factor:SWORD, power:SWORD

.CODE
Power2 PROC C factor:SWORD, power:SWORD

mov ax, factor ; Load Arg1 into AX
mov cx, power ; Load Arg2 into CX
shl ax, cl ; AX = AX * (2 to power of CX)
; Leave return value in AX
ret
Power2 ENDP
END
The MASM procedure declaration for the Power2 routine specifies the C
langtype and the parameters expected by the procedure. The langtype specifies
the calling and naming conventions for the interface between MASM and C.
The routine is public by default. When the C module calls Power2, it passes
two arguments, 3 and 5 by value.
Using the C Startup Code

This section explains how to write an assembly-language program that can call C
library functions. It links with the C startup module, which performs the
necessary initialization required by the library functions.
You must follow these steps when writing such a program:
1. Specify the C convention in the .MODEL statement.
2. Include the following (optional) statement to note linkage with the C startup
module:
EXTERN _acrtused:abs
3. Prototype or declare as external all C functions the program references.

4. Include a public procedure called main in your assembly-language module.
The C startup code calls _main (which is why all C programs begin with a
main function). This procedure serves as the effective entry point for your
program.
5. Omit an entry point in the program’s END directive. The C startup code
serves as the true entry point when the program runs.
6. Assemble with ML’s /Cx switch to preserve the case of nonlocal names.

The following example serves as a template for these steps. The program calls
the C run-time function printf to display two variables.
.MODEL small, c ; Step 1: declare C conventions
EXTERN _acrtused:abs ; Step 2: bring in C startup
.
.
.
printf PROTO NEAR, ; Step 3: prototype
pstring:NEAR PTR BYTE, ; external C
num1:WORD, num2:VARARG ; routines
.DATA
format BYTE '%i %i', 13, 0
.CODE
main PROC PUBLIC ; Step 4: C startup calls here

.
.
.
INVOKE printf, OFFSET format, ax, bx
.
.
.
END ; Step 5: no label on END
The C++/MASM Interface

C++ can apply a protocol called a “linkage specification” to mixed-language
procedures. This lets you link C++ code in the same way as C code. All
information in the preceding section applies when linking assembly-language and
C++ routines through the C linkage specification.
The C linkage specification forces the C++ compiler to adopt C conventions —
which are not the same as C++ conventions — for listed routines. Since MASM
does not specifically support C++ conventions, set the C linkage specification in
your C++ code for all mixed-language routines, as shown here:
extern “C” declaration
where declaration is the prototype of an exported C++ function or an imported
assembly-language procedure. You can bracket a list of declarations:
extern "C"
{
int WriteLine( short attr, char *string );
void GoExit( int err );
}

or apply the specification to individual prototypes:

extern "C" int WriteLine( short attr, char *string );
extern "C" void GoExit( int err );
Note the syntax remains the same whether WriteLine and GoExit are
exported C++ functions or imported assembly-language routines. The linkage
specification applies only to called routines, not to external variables. Use the
extern keyword (without the “C”) as you normally would when identifying
objects external to the C++ module.
The FORTRAN/MASM Interface

This section summarizes the specific details important to calling FORTRAN
procedures or receiving arguments from FORTRAN routines that call MASM
routines. It includes a sample MASM and FORTRAN module.
A FORTRAN procedure follows the Pascal calling convention by default. This
convention passes arguments in the order listed, and the calling procedure
removes the arguments from the stack. The naming convention converts all
exported names to uppercase.

This list shows the FORTRAN data types that are equivalent to the MASM 6.1
data types.
FORTRAN Type Equivalent MASM Type
CHARACTER*1 BYTE
INTEGER*1 SBYTE
INTEGER*2 SWORD
REAL*4 REAL4
INTEGER*4 SDWORD
REAL*8, DOUBLE PRECISION REAL8
Naming Restrictions
FORTRAN allows 31 characters for identifier names. A digit or an underscore
cannot be the first character in an identifier name.
By default, FORTRAN passes arguments by reference as far addresses if the
FORTRAN module is compiled in large or huge memory model. It passes them
as near addresses if the FORTRAN module is compiled in medium model.
Versions of FORTRAN prior to Version 4.0 always require large model.

The FORTRAN compiler passes an argument by value when declared with the
VALUE attribute. This declaration can occur either in a FORTRAN
INTERFACE block (which determines how to pass an argument) or in a
function or subroutine declaration (which determines how to receive an
argument).
In FORTRAN you can apply the NEAR (or FAR) attribute to reference
parameters. These keywords override the default. They have no effect when
they specify the same method as the default.

A call to a FORTRAN function or subroutine declared with the PASCAL or C
attribute passes all arguments by value in the parameter list (except for
parameters declared with the REFERENCE attribute). This change in default
passing method applies to function and subroutine definitions as well as to the
functions and subroutines described by INTERFACE blocks.
Array Storage
When you declare FORTRAN arrays, you can specify any integer for the lower
bound (the default is 1). The FORTRAN compiler stores all arrays in column-
major order — that is, the leftmost subscript increments most rapidly. For
example, the first seven elements of an array defined as A[3,4] are stored as
A[1,1], A[2,1], A[3,1], A[1,2], A[2,2], A[3,2], A[1,3]
String Format
FORTRAN stores strings as a series of bytes at a fixed location in memory, with
no delimiter at the end of the string. When passing a variable-length FORTRAN
string to another language, you need to devise a method by which the target
routine can find the end of the string.
Consider the string declared as
CHARACTER*14 MSG
MSG = 'String of text'
The string is stored in 14 bytes of memory like this:
Figure 12.3 FORTRAN String Format

Strings are passed by reference. Although FORTRAN has a method for passing
length, the variable-length FORTRAN strings cannot be used in a mixed-
language interface because other languages cannot access the temporary variable
that FORTRAN uses to communicate string length. However, fixed-length
strings can be passed if the FORTRAN INTERFACE statement declares the
length of the string in advance.
External Data
FORTRAN routines can directly access external data. In FORTRAN you can
declare data to be external by adding the EXTERN attribute to the data
declaration. You can also access a FORTRAN variable from MASM if it is
declared in a COMMON block.
A FORTRAN program can call an external assembly procedure with the use of
the INTERFACE statement. However, the INTERFACE statement is not
strictly necessary unless you intend to change one of the FORTRAN defaults.
Structure Alignment
By default, FORTRAN uses word alignment (unpacked storage) for all data
objects larger than 1 byte. This storage method specifies that occasional bytes
may be added as padding, so that word and doubleword objects start on an even
boundary. In addition, all nested structures and records start on a word
boundary. The MASM default is byte-alignment, so you should specify an
alignment of 2 for MASM structures or use the /Zp1 option when compiling in
FORTRAN.

Use the same memory model for the MASM and FORTRAN modules.
Returning Values
You must use a special convention to return floating-point values, records, user-
defined types, arrays, and values larger than 4 bytes to a FORTRAN module
from an assembly procedure. The FORTRAN module creates space in the stack
segment to hold the actual return value. When the call to the assembly
procedure is made, an extra parameter is passed. This parameter is the last one
pushed. The segment address of the return value is contained in SS.
In the assembly procedure, put the data for the return value at the location
pointed to by the return value offset. Then copy the return-value offset (located
at BP + 6) to AX, and copy SS to DX. This is necessary because the calling
module expects DX:AX to point to the return value.

Structures, Records, and User-Defined Data Types

The FORTRAN structure variable, defined with the STRUCTURE keyword
and declared with the RECORD statement, is equivalent to the Pascal
RECORD and the C struct. You can pass structures as arguments by value or
by reference (the default).
The FORTRAN types COMPLEX*8 and COMPLEX*16 are not directly
implemented in MASM. However, you can write structures that are equivalent.
The type COMPLEX*8 has two fields, both of which are 4-byte floating-point
numbers; the first contains the real component, and the second contains the
imaginary component. The type COMPLEX is equivalent to the type
COMPLEX*8.
The type COMPLEX*16 is similar to COMPLEX*8. The only difference is
that each field of the former contains an 8-byte floating-point number.
A FORTRAN LOGICAL*2 is stored as a 1-byte indicator value (1=true,
0=false) followed by an unused byte. A FORTRAN LOGICAL*4 is stored as a
1-byte indicator value followed by three unused bytes. The type LOGICAL is
equivalent to LOGICAL*4, unless $STORAGE:2 is in effect.
To pass or receive a FORTRAN LOGICAL type, declare a MASM structure
with the appropriate fields.

In FORTRAN, you can call routines with a variable number of arguments by
including the VARYING attribute in your interface to the routine, along with the
C attribute. You must use the C attribute because a variable number of
arguments is possible only with the C calling convention. The VARYING
attribute prevents FORTRAN from enforcing a matching number of parameters.
Pointers and Addresses

FORTRAN programs can determine near and far addresses with the
LOCNEAR and LOCFAR functions. Store the result as INTEGER*2 (with
the LOCNEAR function) or as INTEGER*4 (with the LOCFAR function). If
you pass the result of LOCNEAR or LOCFAR to another language, be sure to
pass by value.

Example
In the following example, the FORTRAN module calls an assembly procedure
that calculates A*2^B, where A and B are the first and second parameters,
respectively. This is done by shifting the bits in A to the left B times.
INTERFACE TO INTEGER*2 FUNCTION POWER2(A, B)
INTEGER*2 A, B
END
PROGRAM MAIN
INTEGER*2 POWER2
INTEGER*2 A, B
A = 3
B = 5
WRITE (*, *) '3 TIMES 2 TO THE B OR 5 IS ',POWER2(A, B)
END
To understand the assembly procedure, consider how the parameters are placed
on the stack, as illustrated in Figure 12.4.
Figure 12.4 FORTRAN Stack Frame
Figure 12.4 assumes that the FORTRAN module is compiled in large model. If
you compile the FORTRAN module in medium model, then each argument is
passed as a 2-byte, not 4-byte, address. The return address is 4 bytes long
because procedures called from FORTRAN must always be FAR.

The assembler code looks like this:

.MODEL LARGE, FORTRAN
Power2 PROTO FORTRAN, pFactor:FAR PTR SWORD, pPower:FAR PTR SWORD
.CODE
Power2 PROC FORTRAN, pFactor:FAR PTR SWORD, pPower:FAR PTR SWORD
les bx, pFactor ; ES:BX points to factor

mov ax, es:[bx] ; AX = value of factor
les bx, pPower ; ES:BX points to power
mov cx, es:[bx] ; CX = value of power
shl ax, cl ; Multiply by 2^power
ret ; Return result in AX
Power2 ENDP
END
The Basic/MASM Interface

This section explains how to call MASM procedures or functions from Basic
and how to receive Basic arguments for the MASM procedure. Pascal is the
default naming and calling convention, so all lowercase letters are converted to
uppercase. Routines defined with the FUNCTION keyword return values, but
routines defined with SUB do not. Basic DEF FN functions and GOSUB
routines cannot be called from another language.
The information provided pertains to Microsoft’s Basic and QuickBasic
compilers. Differences between the two compilers are noted when necessary.

The following list shows the Basic data types that are equivalent to the MASM
6.1 data types.
Basic Type Equivalent MASM Type
STRING*1 WORD
INTEGER (X%) SWORD
SINGLE (X!) REAL4
LONG (X&), SDWORD
CURRENCY
DOUBLE (X#) REAL8
Naming Conventions
Basic recognizes up to 40 characters of a name. In the object code, Basic also
drops any of its reserved characters: %, &, !, #, @, &.

Basic can pass data in several ways and can receive it by value or by near
reference. By default, Basic arguments are passed by near reference as 2-byte
addresses. To pass a near address, pass only the offset; if you need to pass a far
address, pass the segment and offset separately as integer arguments. Pass the
segment address first, unless you have specified C compatibility with the
CDECL keyword.
Basic passes each argument in a call by far reference when CALLS is used to
invoke a routine. You can also use SEG to modify a parameter in a preceding
DECLARE statement so that Basic passes that argument by far reference. To
pass any other variable type by value, apply the BYVAL keyword to the
argument in the DECLARE statement. You cannot pass arrays and user-
defined types by value.
DECLARE SUB Test(BYVAL a%, b%, SEG c%)
CALL Test(x%, y%, z%)

CALLS Test(x%, y%, z%)
This CALL statement passes the first argument (a%) by value, the second
argument (b%) by near reference, and the third argument (c%) by far reference.
The statement
CALLS Test2(x%, y%, z%)
passes each argument by far reference.

Including the CDECL keyword in the Basic DECLARE statement enables the
C calling and naming conventions. This also allows a call to a MASM procedure
with a varying number of arguments.
Array Storage
The DIM statement sets the number of dimensions for a Basic array and also
sets the array’s maximum subscript value. In the array declaration DIM
x(a,b), the upper bounds (the maximum number of values possible) of the
array are a and b. The default lower bound is 0. The default upper bound for an
array subscript is 10.
The default for column storage in Basic is column-major order, as in
FORTRAN. For an array defined as DIM Arr%(3,3), reference the last
element as Arr%(3,3). The first five elements of Arr (3,3) are
Arr(0,0), Arr(1,0), Arr(2,0), Arr(0,1), Arr(1,1)

When you pass an array from Basic to a language that stores arrays in row-
major order, use the command-line option /R when compiling the Basic module.
Most Microsoft languages permit you to reference arrays directly. Basic uses an
array descriptor, however, which is similar in some respects to a Basic string
descriptor. The array descriptor is necessary because Basic handles memory
allocation for arrays dynamically, and thus may shift the location of the array in
memory.
A reference to an array in Basic is really a near reference to an array descriptor.
Array descriptors are always in DGROUP, even though the data may be in far
memory. Array descriptors contain information about type, dimensions, and
memory locations of data. You can safely pass arrays to MASM routines only if
you follow three rules:
u Pass the array’s address by applying the VARPTR function to the first
element of the Basic array and passing the result by value. To pass the far
address of the array, apply both the VARPTR and VARSEG functions and
pass each result by value. The receiving language gets the address of the first
element and considers it to be the address of the entire array. It can then
access the array with its normal array-indexing syntax.
u The MASM routine that receives the array should not call back to one of the
calling program’s routines before it has finished processing the array.
Changing data within the caller’s heap — even data unrelated to the array —
may change the array’s location in the heap. This would invalidate any
further work the called routine performs, since the routine would be
operating on the array’s old location.
u Basic can pass any member of an array by value. When passing individual
array elements, these restrictions do not apply.
You can apply LBOUND and UBOUND to a Basic array to determine lower
and upper bounds, and then pass the results to another routine. This way, the
size of the array does not need to be determined in advance.
String Format
Basic maintains a 4-byte string descriptor for each string, as shown in the
following. The first field of the string descriptor contains a 2-byte integer
indicating the length of the actual string text. The second field contains the offset
address of this text within the caller’s data segment.
Figure 12.5 Basic String Descriptor Format

An assembly-language procedure can store a Basic string descriptor as a simple

structure, like this:
DESC STRUCT
len WORD ? ; Length of string
off WORD ? ; Offset of string
DESC ENDS
string BYTE "This text referenced by a string descriptor"

sdesc DESC (LENGTHOF string, string)
Version 7.0 or later of the Microsoft Basic Compiler provides new functions that
access string descriptors. These functions simplify the process of sharing Basic
string data with routines written in other languages.
Earlier versions of Basic offer the LEN (Length) and SADD (String Address)
functions, which together obtain the information stored in a string descriptor.
LEN returns the length of a string in bytes. SADD returns the offset address of
a string in the data segment. The caller must provide both pieces of information
so the called procedure can locate and read the entire string. The address
returned by SADD is declared as type INTEGER but is actually equivalent to a
C near pointer.
If you need to pass the far address of a string, use the SSEGADD (String
Segment Address) function of Microsoft Basic version 7.0 or later. You can also
determine the segment address of the first element with VARSEG.
External Data
Declaring global data in Basic follows the same two-step process as in other
languages:
1. Declare shareable data in Basic with the COMMON statement.
2. Identify the shared variables in your assembly-language procedures with the
EXTERN keyword. Place the EXTERN statement outside of a code or data
segment when declaring far data.
Structure Alignment
Basic packs user-defined types. For MASM structures to be compatible, select
byte-alignment.

Always use medium model in assembly-language procedures linked with Basic
modules. If you are listing other libraries on the LINK command line, specify
Basic libraries first. (There are differences between the QBX and command-line
compilation. See your Basic documentation.)

Returning Values
Basic follows the usual convention of returning values in AX or DX:AX. If the
value is not floating point, an array, or a structured type, or if it is less than 4
bytes long, then the 2-byte integers should be returned from the MASM
procedure in AX and 4-byte integers should be returned in DX:AX. For all other
types, return the near offset in AX.
User-Defined Data Types

The Basic TYPE statement defines structures composed of individual fields.
These types are equivalent to the C struct, FORTRAN record (declared with
the STRUCTURE keyword), and Pascal Record types.
You can use any of the Basic data types except variable-length strings or
dynamic arrays in a user-defined type. Once defined, Basic types can be passed
only by reference.

You can vary the number of arguments in Basic when you change the calling
convention with CDECL. To call a function with a varying number of
arguments, you also need to suppress the type checking that normally forces a
call to be made with a fixed number of arguments. In Basic, you can remove
this type checking by omitting a parameter list from the DECLARE statement.
Pointers and Addresses

VARSEG returns a variable’s segment address, and VARPTR returns a
variable’s offset address. These intrinsic Basic functions enable your program to
pass near or far addresses.
Example
This example calls the Power2 procedure in the MASM 6.1 module.
DEFINT A-Z
DECLARE FUNCTION Power2 (A AS INTEGER, B AS INTEGER)

PRINT "3 times 2 to the power of 5 is ";
PRINT Power2(3, 5)
END
The first argument, A, is higher in memory than B because Basic pushes

arguments in the same order in which they appear.

Figure 12.6 shows how the arguments are placed on the stack.
Figure 12.6 Basic Stack Frame
The assembly procedure can be written as follows:

.MODEL medium
Power2 PROTO PASCAL, factor:PTR WORD, power:PTR WORD

.CODE
Power2 PROC PASCAL, factor:PTR WORD, power:PTR WORD
mov bx, WORD PTR factor ; BX points to factor

mov ax, [bx] ; Load factor into AX
mov bx, WORD PTR power ; BX points to power
mov cx, [bx] ; Load power into CX
shl ax, cl ; AX = AX * (2 to power of CX)
ret
Power2 ENDP
END
Note that each parameter must be loaded in a two-step process because the
address of each is passed rather than the value. The return address is 4 bytes
long because procedures called from Basic must be FAR.


335
C H A P T E R 1 3
Writing 32-Bit Applications
This chapter is an introduction to 32-bit programming for the 80386. The

guidelines in this chapter also apply to the 80486 processor, which is basically a
faster 80386 with the equivalent of a 80387 floating-point processor. Since you
are already familiar with 16-bit real-mode programming, this chapter covers the
differences between 16-bit programming and 32-bit protected-mode
programming.
The 80386 processor (and its successors such as the 80486) can run in real
mode, virtual-86 mode, and in protected mode. In real and virtual-86 modes, the
80386 can run 8086/8088 programs. In protected mode, it can run 80286
programs. The 386 also extends the features of protected mode to include 32-bit
operations and segments larger than 64K.
The MS-DOS operating system directly supports 8086/8088 programs, which it

runs either in real mode or virtual-86 mode. Native 32-bit 80386 programs can
be run by using a “DOS extender,” by using the WINMEM32.DLL facility of
Microsoft Windows 3.x, or by running a native 32-bit operating system, such as
Microsoft Windows NT. You can use MASM to generate object code (OMF or
COFF) for 32-bit programs. To do this, you will need a software development
kit such as the Windows SDK for the target environment. Such kits include the
linker and other components specific to your chosen operating environment.
32-Bit Memory Addressing

The 80386 has six segment registers. Four of these are familiar to 8086/8088
programmers: CS (Code Segment), SS (Stack Segment), DS (Data Segment),
and ES (Extra Segment). The two additional registers, FS and GS, are used as
data segment registers.
Memory addresses on 80x86 machines consist of two parts — a segment and an

offset. In real-mode programs, the segment is a 16-bit number and the offset is a
16-bit number. Effective addresses are calculated by multiplying the segment by

16 and adding the offset to it. In protected mode, the segment value is not used
directly as a number, but instead is an index to a table of “selectors.” Each
selector describes a block of memory, including attributes such as the size and
location of the block, and the access rights the program has to it (read, write,
execute). The effective address is calculated by adding the offset to the base
address of the memory block described by the selector.
All segment registers are 16 bits wide. The offset in a 32-bit protected-mode
program is itself 32 bits wide, which means that a single segment can address up
to 4 gigabytes of memory. Because of this large range, there is little need to use
segment registers to extend the range of addresses in 32-bit programs. If all six
segment registers are initially set to the same value, then the rest of the program
can ignore them and treat the processor as if it used a 32-bit linear address
space. This is called 0:32, or flat, addressing. (The full segmented 32-bit
addressing mode, in which the segment registers can contain different values, is
called 16:32 addressing.) Flat addressing is used by the Windows NT operating
system.
Figure 13.1 32-Bit Register Set
MASM Directives for 32-Bit Programming

If you use the simplified segment directives, a 32-bit program is surprisingly
similar to a program for MS-DOS. Here are the differences:

Chapter 13 Writing 32-Bit Applications 337
u Supply the .386 directive, which enables the 32-bit programming features of
the 386 and its successors. The .386 directive must precede the .MODEL
directive.
u For flat-model programming, use the directive
.MODEL flat, stdcall
which tells the assembler to assume flat model (0:32) and to use the
Windows NT standard calling convention for subroutine calls.
u Precede your data declarations with the .DATA directive.
u Precede your instruction codes with the .CODE directive.
u At the end of the source file, place an END directive.
Sample Program
The following sample is a 32-bit assembly language subroutine, such as might be
called from a 32-bit C program written for the Windows NT operating system.
The program illustrates the use of a variety of directives to make assembly
language easier to read and maintain. Note that with 32-bit flat model
programming, there is no longer any need to refer to segment registers, since
these are artifacts of segmented addressing.

;* szSearch - An example of 32-bit assembly programming using MASM 6.1

;*
;* Purpose: Search a buffer (rgbSearch) of length cbSearch for the
;* first occurrence of szTok (null terminated string).
;*
;* Method: A variation of the Boyer-Moore method
;* 1. Determine length of szTok (n)
;* 2. Set array of flags (rgfInTok) to TRUE for each character
;* in szTok
;* 3. Set current position of search to rgbSearch (pbCur)
;* 4. Compare current position to szTok by searching backwards
;* from the nth position. When a comparison fails at
;* position (m), check to see if the current character
;* in rgbSearch is in szTok by using rgfInTok. If not,
;* set pbCur to pbCur+(m)+1 and restart compare. If
;* pbCur reached, increment pbCur and restart compare.
;* 5. Reset rgfInTok to all 0 for next instantiation of the
;* routine.
.386
.MODEL flat, stdcall
FALSE EQU 0
TRUE EQU NOT FALSE

Chapter 13 Writing 32-Bit Applications 339
.DATA
; Flags buffer - data initialized to FALSE. We will
; set the appropriate flags to TRUE during initialization
; of szSearch and reset them to FALSE before exit.
rgfInTok BYTE 256 DUP (FALSE);
.CODE
szSearch PROC PUBLIC USES esi edi,

rgbSearch:PBYTE,
cbSearch:DWORD,
szTok:PBYTE
; Initialize flags buffer. This tells us if a character is in

; the search token - Note how we use EAX as an index
; register. This can be done with all extended registers.
mov esi, szTok
xor eax, eax
.REPEAT
lodsb
mov BYTE PTR rgfInTok[eax], TRUE
.UNTIL (!AL)
; Save count of szTok bytes in EDX

mov edx, esi
sub edx, szTok
dec edx
; ESI will always point to beginning of szTok

mov esi, szTok
; EDI will point to current search position

; and will also contain the return value
mov edi, rgbSearch
; Store pointer to end of rgbSearch in EBX

mov ebx, edi
add ebx, cbSearch
sub ebx, edx

; Initialize ECX with length of szTok

mov ecx, edx
.WHILE ( ecx != 0 )
dec ecx ; Move index to current
mov al, [edi+ecx] ; characters to compare
; If the current byte in the buffer doesn't exist in the

; search token, increment buffer pointer to current position
; +1 and start over. This can skip up to 'EDX'
; bytes and reduce search time.
.IF !(rgfInTok[eax])
add edi, ecx
inc edi ; Initialize ECX with
mov ecx, edx ; length of szTok
; Otherwise, if the characters match, continue on as if
; we have a matching token
.ELSEIF (al == [esi+ecx])
.CONTINUE
; Finally, if we have searched all szTok characters,
; and land here, we have a mismatch and we increment
; our pointer into rgbSearch by one and start over.
.ELSEIF (!ecx)
inc edi
mov ecx, edx
.ENDIF
; Verify that we haven't searched beyond the buffer.

.IF (edi > ebx)
mov edi, 0 ; Error value
.BREAK
.ENDIF
.ENDW
; Restore flags in rgfInTok to 0 (for next time).

mov esi, szTok
xor eax, eax
.REPEAT
lodsb
mov BYTE PTR rgfInTok[eax], FALSE
.UNTIL !AL
; Put return value in eax

mov eax, edi
ret
szSearch ENDP
end


341
A P P E N D I X A
Differences Between
MASM 6.1 and 5.1
For the many users who come to version 6.1 of the Microsoft Macro Assembler
directly from the popular MASM 5.1, this appendix describes the differences
between the two versions. Version 6.1 contains significant changes, including:
u An integrated development environment called Programmer’s WorkBench
(PWB) from which you can write, edit, debug, and execute code.
u Expanded functionality for structures, unions, and type definitions.
u New directives for generating loops and decision statements, and for
declaring and calling procedures.
u Simplified methods for applying public attributes to variables and routines in
multiple-module programs.
u Enhancements for writing and using macros.
u Flat-model support for Windows NT and new instructions for the 80486
processor.
The OPTION M510 directive (or the /Zm command-line switch) assures nearly
complete compatibility between MASM 6.1 and MASM 5.1. However, to take
full advantage of the enhancements in MASM 6.1, you will need to rewrite
some code written for MASM 5.1.
The first section of this appendix describes the new or enhanced features in
MASM 6.1. The second section, “Compatibility Between MASM 5.1 and 6.1,”
explains how to:
u Minimize the number of required changes with the OPTION directive.
u Rewrite your existing assembly code, if necessary, to take advantage of the
assembler’s enhancements.
Filename: LMAPGAPA.DOC Project:

New Features of Version 6.1

This section gives an overview of the new features of MASM 6.1 and provides
references to more detailed information elsewhere in the documentation. For full
explanations and coding examples, see the documentation listed in the cross-
references.
The Assembler, Environment, and Utilities

Most of the executable files provided with MASM 6.1 are new or revised. For a
complete list of these files, read the PACKING.TXT file on the distribution
disk. The book Getting Started also provides information about setting up the
environment, assembler, and Help system.
The Assembler
The macro assembler, named ML.EXE, can assemble and link in one step. Its
new 32-bit operation gives ML.EXE the ability to handle much larger source
files than MASM 5.1. The command-line options are new. For example, the /Fl
and /Sc options generate instruction timings in the listing file. Command-line
options are case-sensitive and must be separated by spaces.
For backward compatibility with MASM 5.1 makefiles, MASM 6.1 includes the
MASM.EXE utility. MASM.EXE translates MASM 5.1 command-line options
to the new MASM 6.1 command-line options and calls ML.EXE. See the
Reference book for details.
H2INC
H2INC converts C include files to MASM include files. It translates data
structures and declarations but does not translate executable code. For more
information, see Chapter 20 of Environment and Tools.
NMAKE
NMAKE replaces the MAKE utility. NMAKE provides new functions for
evaluating target files and more flexibility with macros and command-line
options. For more information, see Environment and Tools.
Integrated Environment
PWB is an integrated development environment for writing, developing, and
debugging programs. For information on PWB and the CodeView debugging
application, see Environment and Tools.
Online Help
MASM 6.1 incorporates the Microsoft Advisor Help system. Help provides a
vast database of online help about all aspects of MASM, including the syntax

Appendix A Differences Between MASM 6.1 and 5.1 343
and timings for processor and coprocessor instructions, directives, command-

line options, and support programs such as LINK and PWB.
For information on how to set up the help system, see Getting Started. You can
invoke the help system from within PWB or from the QuickHelp program
(QH).
HELPMAKE
You can use the HELPMAKE utility to create additional help files from ASCII
text files, allowing you to customize the online help system. For more
information, see Environment and Tools.
Other Programs
MASM 6.1 contains the most recent versions of LINK, LIB, BIND, CodeView,
and the mouse driver. The CREF program is not included in MASM 6.1. The
Source Browser provides the information that CREF provided under MASM
5.1. For more information on the Source Browser, see Chapter 5 of
Environment and Tools or Help.
Segment Management
This section lists the changes and additions to memory-model support and
directives that relate to memory model.
Predefined Symbols
The following predefined symbols (also called predefined equates) provide
information about simplified segments:
Predefined Symbol Value
@stack DGROUP for near stacks, STACK for far stacks
@Interface Information about language parameters
@Model Information about the current memory model
@Line The source line in the current file
@Date The current date
@FileCur The current file
@Time The current time
@Environ The current environment variables
For more information about predefined symbols, see “Predefined Symbols” in

Chapter 1.

Enhancements to the ASSUME Directive

MASM automatically generates ASSUME values for the code segment register
(CS). It is no longer necessary to include lines such as
ASSUME CS:MyCodeSegment
in your programs. In addition, the ASSUME directive can include ERROR,

FLAT, or register:type. MASM 6.1 issues a warning when you specify
ASSUME values for CS other than the current segment or group.
For more information, see “Setting the ASSUME Directive for Segment
Registers” in Chapter 2 and “Defining Register Types with ASSUME” in
Chapter 3.
Relocatable Offsets
For compatibility with applications for Windows, the LROFFSET operator can
calculate a relocatable offset, which is resolved by the loader at run time. See
Help for details.
Flat Model
MASM 6.1 supports the flat-memory model of Windows NT, which allows
segments as large as 4 gigabytes. All other memory models limit segment size to
64K for MS-DOS and Windows. For more information about memory models,
see “Defining Basic Attributes with .MODEL” in Chapter 2.
Data Types
MASM 6.1 supports an improved data typing. This section summarizes the
improved forms of data declarations in MASM 6.1.
Defining Typed Variables

You can now use the type names as directives to define variables. Initializers are
unsigned by default. The following example lines are equivalent:
var1 DB 25
var1 BYTE 25
Signed Types
You can use the SBYTE, SWORD, and SDWORD directives to declare signed
data. For more information about these directives, see “Allocating Memory for
Integer Variables” in Chapter 4.

Floating-Point Types
MASM 6.1 provides the REAL4, REAL8, and REAL10 directives for
declaring floating-point variables. For information on these type directives, see
“Declaring Floating-Point Variables and Constants” in Chapter 6 .
Qualified Types
Type definitions can now include distance and language type attributes.
Procedures, procedure prototypes, and external declarations let you specify the
type as a qualified type. A complete description of qualified types is provided in
the section “Data Types” in Chapter 1.
Structures
Changes to structures since MASM 5.1 include:
u Structures can be nested.
u The names of structure fields need not be unique. As a result, you must
qualify references to field names.
u Initialization of structure variables can continue over multiple lines provided
the last character in the line before the comment field is a comma.
u Curly braces and angle brackets are equivalent.
For example, this code works in MASM 6.1:

SCORE STRUCT
team1 BYTE 10 DUP (?)
score1 BYTE ?
team2 BYTE 10 DUP (?)
score2 BYTE ?
SCORE ENDS
first SCORE {"BEARS", 20, ; This comment is allowed.

"CUBS", 10 }
mov al, [bx].score.team1 ; Field name must be qualified

; with structure name.
You can use OPTION OLDSTRUCTS or OPTION M510 to enable MASM

5.1 behavior for structures. See “Compatibility between MASM 5.1 and 6.1,”
later in this appendix. For more information on structures and unions, see
“Structures and Unions” in Chapter 5.
Unions
MASM 6.1 allows the definition of unions with the UNION directive. Unions
differ from structures in that all fields within a union occupy the same data
space. For more information, see “Structures and Unions” in Chapter 5.

Types Defined with TYPEDEF

The TYPEDEF directive defines a type for use later in the program. It is most
useful for defining pointer types. For more information on defining types, see
“Data Types” in Chapter 1, and “Defining Pointer Types with TYPEDEF” in
Chapter 3.
Names of Identifiers
MASM 6.1 accepts identifier names up to 247 characters long. All characters
are significant, whereas under MASM 5.1, names are significant to 31
characters only. For more information on identifiers, see “Identifiers” in Chapter
1.
Multiple-Line Initializers
In MASM 6.1, a comma at the end of a line (except in the comment field)
implies that the line continues. For example, the following code is legal in
MASM 6.1:
longstring BYTE "This string ",
"continues over two lines."
bitmasks BYTE 80h, 40h, 20h, 10h,
08h, 04h, 02h, 01h
For more information, see “Statements” in Chapter 1.
Comments in Extended Lines

MASM 5.1 allows a backslash ( \ ) as the line-continuation character if it is the
last nonspace character in the line. MASM 6.1 permits a comment to follow the
backslash.
Determining Size and Length of Data Labels

The LENGTHOF operator returns the number of data items allocated for a data
label. MASM 6.1 also provides the SIZEOF operator. When applied to a type,
SIZEOF returns the size attribute of the type expression. When applied to a data
label, SIZEOF returns the number of bytes used by the initializer in the label’s
definition. In this case, SIZEOF for a variable equals the number of bytes in the
type multiplied by LENGTHOF for the variable.
MASM 6.1 recognizes the LENGTH and SIZE operators for backward
compatibility. For a description of the behavior of SIZE under OPTION M510,
see “Length and Size of Labels with OPTION M510,” later in this appendix.
For obsolete behavior with the LENGTH operator, see also “LENGTH
Operator Applied to Record Types,” page 356.
For information on LENGTHOF and SIZEOF, see the following sections in
chapter 5: “Declaring and Referencing Arrays,” “Declaring and Initializing

Strings,” “Declaring Structure and Union Variables,” and “Defining Record

Variables.”
HIGHWORD and LOWWORD Operators

These operators return the high and low words for a given 32-bit operand. They
are similar to the HIGH and LOW operators of MASM 5.1 except that
HIGHWORD and LOWWORD can take only constants as operands, not
relocatables (labels).
PTR and CodeView

Under MASM 5.1, applying the PTR operator to a data initializer determines the
size of the data displayed by CodeView. You can still use PTR in this manner in
MASM 6.1, but it does not affect CodeView typing. Defining pointers with the
TYPEDEF directive allows CodeView to generate correct information. See
“Defining Pointer Types with TYPEDEF” in Chapter 3.
Procedures, Loops, and Jumps

With its significant improvements for procedure and jump handling, MASM 6.1
closely resembles high-level – language implementations of procedure calls.
MASM 6.1 generates the code to correctly handle argument passing, check type
compatibility between parameters and arguments, and process a variable number
of arguments. MASM 6.1 can also automatically recast jump instructions to
correct for insufficient jump distance.
Function Prototypes and Calls

The PROTO directive lets you prototype procedures in the same way as high-
level languages. PROTO enables type-checking and type conversion of
arguments when calling the procedure with INVOKE. For more information,
see “Declaring Procedure Prototypes” in Chapter 7.
The INVOKE directive sets up code to call a procedure and correctly pass
arguments according to the prototype. MASM 6.1 also provides the VARARG
keyword to pass a variable number of arguments to a procedure with INVOKE.
For more information about INVOKE and VARARG, see “Calling Procedures
with INVOKE” and “Declaring Parameters with the PROC Directive” in
Chapter 7.
The ADDR keyword is new since MASM 5.1. When used with INVOKE, it
provides the address of a variable, in the same way as the address-of operator
(&) in C. This lets you conveniently pass an argument by reference rather than
value. See “Calling Procedures with INVOKE” in Chapter 7.

High-Level Flow-Control Constructions

MASM 6.1 contains several directives that generate code for loops and decisions
depending on the status of a conditional statement. The conditions are tested at
run time rather than at assembly time.
Directives new since MASM 5.1 include .IF, .ELSE, .ELSEIF, .REPEAT,
.UNTIL, .UNTILCXZ, .WHILE, and .ENDW. MASM 6.1 also provides the
associated .BREAK and .CONTINUE directives for loops and IF statements.
For more information, see “Loops” in Chapter 7 and “Decision Directives” on
page 171.
Automatic Optimization for Unconditional Jumps

MASM 6.1 automatically determines the smallest encoding for direct
unconditional jumps. See “Unconditional Jumps” in Chapter 7.
Automatic Lengthening for Conditional Jumps

If a conditional jump cannot reach its target destination, MASM automatically
recasts the code to use an unconditional jump to the target. See “Jump
Extending,” page 169.
User-Defined Stack Frame Setup and Cleanup

The prologue code generated immediately after a PROC statement sets up the
stack for parameters and local variables. The epilogue code handles stack
cleanup. MASM 6.1 allows user-defined prologues and epilogues, as described
in “Generating Prologue and Epilogue Code” in Chapter 7.
Simplifying Multiple-Module Projects

MASM 6.1 simplifies the sharing of code and data among modules and makes
the use of include files more efficient.
EXTERNDEF in Include Files

MASM 5.1 requires that you declare public and external all data and routines
used in more than one module. With MASM 6.1, a single EXTERNDEF
directive accomplishes the same task. EXTERNDEF lets you put global data
declarations within an include file, making the data visible to all source files that
include the file. For more information, see “Using EXTERNDEF” in Chapter 8.
Search Order for Include Files

MASM 6.1 searches for include files in the directory of the main source file
rather than in the current directory. Similarly, it searches for nested include files
in the directory of the include file. You can specify additional paths to search
with the /I command-line option. For more information on include files, see
“Organizing Modules” in Chapter 8.

Enforcing Case Sensitivity

In MASM 5.1, sensitivity to case is influenced only by command-line options
such as /MX, not the language type given with the .MODEL directive. In
MASM 6.1, the language type takes precedence over the command-line options
in specifying case sensitivity.

Alternate Names for Externals

The syntax for EXTERN allows you to specify an alternate symbol name,
which the linker can use to resolve an external reference to an unused symbol.
This prevents linkage with unneeded library code, as explained in “Using
EXTERN with Library Routines,” Chapter 8.
Expanded State Control

Several directives in MASM 6.1 enable or disable various aspects of the
assembler control. These include 80486 coprocessor instructions and use of
compatibility options.
The OPTION Directive

The new OPTION directive allows you to selectively define the assembler’s
behavior, including its compatibility with MASM 5.1. See “Using the OPTION
Directive” in Chapter 1 and “Compatibility between MASM 5.1 and 6.1,” later
in this appendix.
The .NO87 Directive

The .NO87 directive disables all coprocessor instructions. For more information,
see Help.
The .486 and .486P Directives

MASM 6.1 can assemble instructions specific to the 80486, enabled with the
.486 directive. The .486P directive enables 80486 instructions at the highest
privilege level (recommended for systems-level programs only). For more
information, see Help.
The PUSHCONTEXT and POPCONTEXT Directives

The directive PUSHCONTEXT saves the assembly environment, and
POPCONTEXT restores it. The environment includes the segment register
assumes, the radix, the listing and CREF flags, and the current processor and
coprocessor. Note that .NOCREF (the MASM 6.1 equivalent to .XCREF) still
determines whether information for a given symbol will be added to Browser
information and to the symbol table in the listing file. For more information on
listing files, see Appendix C or Help.

New Processor Instructions

MASM 6.1 supports these instructions for the 80486 processor:
80486 Instruction Description
BSWAP Byte swap
CMPXCHG Compare and exchange
INVD Invalidate data cache
INVLPG Invalidate Translation Lookaside Buffer entry
WBINVD Write back and invalidate data cache
XADD Exchange and add
For full descriptions of these instructions, see the Reference or Help.
Renamed Directives
Although MASM 6.1 still supports the old names in MASM 5.1, the following
directives have been renamed for language consistency:
MASM 6.1 MASM 5.1
.DOSSEG DOSSEG
.LISTIF .LFCOND
.LISTMACRO .XALL
.LISTMACROALL .LALL
.NOCREF .XCREF
.NOLIST .XLIST
.NOLISTIF .SFCOND
.NOLISTMACRO .SALL
ECHO %OUT
EXTERN EXTRN
FOR IRP
FORC IRPC
REPEAT REPT
STRUCT STRUC
SUBTITLE SUBTTL
Specifying 16-Bit and 32-Bit Instructions

MASM 6.1 supports all instructions that work with the extended 32-bit registers
of the 80386/486. For certain instructions, you can override the default operand

size with the W (word) and the D (doubleword) suffixes. For details, see the
Reference or Help.
Macro Enhancements
There are significant enhancements to macro functions in MASM 6.1.
Directives provide for a variable number of arguments, loop constructions,
definitions of text equates, and macro functions.
Variable Arguments
MASM 5.1 ignores extra arguments passed to macros. In MASM 6.1, you can
pass a variable number of arguments to a macro by appending the VARARG
keyword to the last macro parameter in the macro definition. The macro can
then reference additional arguments relative to the last declared parameter. This
procedure is explained in “Returning Values with Macro Functions” in Chapter
9.
Required and Default Macro Arguments

With MASM 6.1, you can use REQ or the := operator to specify required or
default arguments. See “Specifying Required and Default Parameters” in
Chapter 9.
New Directives for Macro Loops

Within a macro definition, WHILE repeats assembly as long as a condition
remains true. Other macro loop directives, IRP, IRPC, and REPT, have been
renamed FOR, FORC, and REPEAT. For more information, see “Defining
Repeat Blocks with Loop Directives” in Chapter 9.
Text Macros
The EQU directive retains its old functionality, but MASM 6.1 also incorporates
a TEXTEQU directive for defining text macros. TEXTEQU allows greater
flexibility than EQU. For example, TEXTEQU can assign to a label the value
calculated by a macro function. For more information, see “Text Macros” in
Chapter 9.
The GOTO Directive for Macros

Within a macro definition, GOTO transfers assembly to a line labeled with a
leading colon(:). For more information on GOTO, see Help.
Macro Functions
At assembly time, macro functions can determine and return a text value using
EXITM. Predefined macro string functions concatenate strings, return the size
of a string, and return the position of a substring within a string. For information

on writing your own macro functions, see “Returning Values with Macro
Functions” in Chapter 9.
Predefined Macro Functions

MASM 6.1 provides the following predefined text macro functions:
Symbol Value Returned
@CatStr A concatenated string
@InStr The position of one string within another
@SizeStr The size of a string
@SubStr A substring
For more information on predefined macros, see “String Directives and

Predefined Functions” in Chapter 9.
MASM 6.1 Programming Practices

MASM 6.1 provides many features that make it easier for you to write assembly
code. If you are familiar with MASM 5.1 programming, you may find it helpful
to adopt the following list of new programming practices for programming with
MASM 6.1. The list summarizes many of the changes covered in the following
section, “Compatibility Between MASM 5.1 and 6.1.”
u Select identifier names that do not begin with the dot operator (.).
u Use the dot operator (.) only to reference structure fields, and the plus
operator (+) when not referencing structures.
u Different structures can have the same field names. However, the assembler
does not allow ambiguous references. You must include the structure type
when referring to field names common to two or more structures.
u Separate macro arguments with commas, not spaces.
u Avoid adding extra ampersands in macros. For a list of the new rules about
using ampersands in macros, see “Substitution Operator” in Chapter 9 and
“OPTION OLDMACROS,” page 372.
u By default, code labels defined with a colon are local. Place two colons after
code labels if you want to reference the label outside the procedure.
Compatibility Between MASM 5.1 and 6.1

MASM 6.1 provides a “compatibility mode,” making it easy for you to transfer
existing MASM 5.1 code to the new version. You invoke the compatibility mode
through the OPTION M510 directive or the /Zm command-line switch. This

section explains the changes you may need to make to get your MASM 5.1 code
to run under MASM 6.1 in compatibility mode.
Rewriting Code for Compatibility

In some cases, MASM 6.1 with OPTION M510 does not support MASM 5.1
behavior. In several cases, this is because bugs in MASM 5.1 were corrected.
To update your code to MASM 6.1, use the instructions in this section. This
usually requires only minor changes.
Many of the topics listed here will not apply to your code. This section discusses
topics in order of likelihood, beginning with the most common. In addition, you
may have conflicts between identifier names and new reserved words. OPTION
NOKEYWORD resolves errors generated from the use of reserved words as
identifiers. See “OPTION NOKEYWORD,” page 376, for more information.
Bug Fixes Since MASM 5.1

This section lists the differences between MASM 5.1 and MASM 6.1 due to
bug corrections since MASM 5.1.
Invalid Use of LOCK, REPNE, and REPNZ

Except in compatibility mode, MASM 6.1 flags illegal uses of the instruction
prefixes LOCK, REPNE, and REPNZ. The error generated for invalid uses of
the LOCK, REPNE, and REPNZ prefixes is error A2068:
instruction prefix not allowed
Table A.1 summarizes the correct use of the instruction prefixes. It lists each
string instruction with the type of repeat prefix it uses, and indicates whether the
instruction works on a source, a destination, or both.
Table A.1 Requirements for String Instructions
Instruction Repeat Prefix Source/Destination Register Pair
MOVS REP Both DS:SI, ES:DI
SCAS REPE/REPNE Destination ES:DI
CMPS REPE/REPNE Both DS:SI, ES:DI
LODS -- Source DS:SI
STOS REP Destination ES:DI
INS REP Destination ES:DI
OUTS REP Source DS:SI

No Closing Quotation Marks in Macro Arguments

In MASM 5.1, you can use both single and double quotation marks (' and ") to
begin strings in macro arguments. The assembler does not generate an error or
warning if the string does not end with quotation marks on a macro call. Instead,
MASM 5.1 considers the remainder of the line to be part of the macro argument
containing the opening quote, as if there were a closing quotation mark at the
end of the line.
By default, MASM 6.1 now generates error A2046:
missing single or double quotation mark in string
so all single and double quotation marks in macro arguments must be matched.
To correct such errors in MASM 6.1, either end the string with a closing
quotation mark as shown in the following example, or use the macro escape
character (!) to treat the quotation mark literally.
; MASM 5.1 code
MyMacro "all this in one argument
; Default MASM 6.1 code

MyMacro "all this in one argument"
Making a Scoped Label Public

MASM 5.1 considers code labels defined with a single colon inside a procedure
to be local to that procedure if the module contains a .MODEL directive with a
language type. Although the label is local, MASM 5.1 does not generate an error
if it is also declared PUBLIC. MASM 6.1 generates error A2203:
cannot declare scoped code label as PUBLIC
If you want to make a label PUBLIC, it must not be local. You can use the
double colon operator to define a non-scoped label, as shown in this example:
PUBLIC publicLabel
publicLabel:: ; Non-scoped label MASM 6.1
Byte Form of BT, BTS, BTC, and BTR Instructions

MASM 5.1 allows a byte argument for the 80386 bit-test instructions, but
encodes it as a word argument. The byte form is not supported by the
processor.
MASM 6.1 does not support this behavior and generates error A2024:
invalid operand size for instruction
Rewrite your code to use a word-sized argument.

Default Values for Record Fields

In MASM 5.1, default values for record fields can range down to –2n , where n
is the number of bits in the field. This results in the loss of the sign bit.
MASM 6.1 allows a range of –2n–1 to 2n–1 for default values. Illegal initializers
generate error A2071:
initializer too large for specified size
Design Change Issues

MASM 6.1 includes design changes that make the language more consistent.
These changes are not affected by the OPTION directive, discussed later in
this appendix. Therefore, the changes require revisions in your code. In most
cases, the necessary revisions are minor and the circumstances requiring
changes are rare.
Operands of Different Size

MASM 5.1 does not require operands to agree in size, as the following code
illustrates:
.DATA?
var1 DB ?
var2 DB ?
.CODE
.
.
.
mov var1, ax ; Copy AX to word at var1
The operands for the MOV instruction do not match in size, yet the instruction
assembles correctly. It places the contents of AL into var1 and AH into var2,
moving a word of data in one step. If the code defined var1 as a word value,
the instruction
mov var1, al
would also assemble correctly, copying AL into the low byte of var1 while
leaving the high byte unaffected. Except at warning level 0, MASM 5.1 issues a
warning to inform you of the size mismatch, but both scenarios are legal.
MASM 6.1 does not accept instructions with operands that do not agree in size.
You must specifically “coerce” the size of the memory operand, like this:

mov BYTE PTR var1, al
Conflicting Structure Declarations

MASM 5.1 allows you to declare two or more structures with the same name.
Each declaration replaces the previous declaration. However, the field names
from previous declarations still remain in the assembler’s list of declared values.
MASM 6.1 does not allow conflicting declarations of a structure. It generates
errors A2160 through A2165 for each conflicting declaration. The errors note a
specific conflict, such as conflicting number of fields, conflicting names of fields,
or conflicting initializers.
Forward References to Text Macros Outside of Expressions

MASM 5.1 allows forward references to text macros in specialized cases.
MASM 6.1 with OPTION M510 also permits forward references, but only
when the text macro is referenced in an expression. To revise your code, place
all macro definitions at the beginning of the file.
HIGH and LOW Applied to Relocatable Operands

In some cases, MASM 5.1 accepts HIGH and LOW applied to relocatable
memory expressions. For example, MASM 5.1 allows this code sequence:
; MASM 5.1 code
EXTRN var1:WORD
var2 DW 0
mov al, LOW var1 ; These two instructions yield the
mov ah, HIGH var1 ; same as mov ax, OFFSET var1
However, the instruction

mov ax, LOW var2
is not legal. MASM 6.1 generates error A2105:

HIGH and LOW require immediate operands
The OFFSET operator is required on these operands in MASM 6.1, as shown in

the following. Rewrite your code if necessary.
; MASM 6.1 code
mov al, LOW OFFSET var1
mov ah, HIGH OFFSET var2
OFFSET Applied to Group Names and Indirect Memory Operands

In MASM 6.1, you cannot apply OFFSET to a group name, indirect argument,
or procedure argument. Doing so generates error A2098:

invalid operand for OFFSET
LENGTH Operator Applied to Record Types

In MASM 5.1, the LENGTH operator, when applied to a record type, returns
the total number of bits in a record definition.

In MASM 6.1, the statement LENGTH recordName returns error A2143:

expected data label
Rewrite your code if necessary. The new SIZEOF operator returns information
about records in MASM 6.1. For more information, see “Defining Record
Variables” in Chapter 5.
Signed Comparison of Hexadecimal Values Using GT, GE, LE, or LT

The rules for two’s-complement comparisons have changed. In MASM 5.1, the
expression
0FFFFh GT -1
is false because the two’s-complement values are equal. However, because

hexadecimal numbers are now treated as unsigned, the expression is true in
MASM 6.1. To update, rewrite the affected code.
RET Used with a Constant in Procedures with Epilogues

By default in MASM 6.1, the RET instruction followed by a constant
suppresses automatic generation of epilogue code. MASM 5.1 ignores the
operand and generates the epilogue. Remove the argument if necessary. See
“Generating Prologue and Epilogue Code” in Chapter 7.
Code Labels at Top of Procedures with Prologues

By default in MASM 5.1, a code label defined on the same line as the first
procedure instruction refers to the first byte of the prologue.
In MASM 6.1, a code label defined at the beginning of a procedure refers to the
first byte of the procedure after the prologue. If you need to label the start of the
prologue code, place the label before the PROC statement. For more
information, see “Generating Prologue and Epilogue Code” in Chapter 7.
Use of % as an Identifier Character

MASM 5.1 allows % as an identifier character. This behavior leads to
ambiguities when % is used as the expansion operator in macros. Since % is not
allowed as a character in MASM 6.1 identifiers, you must change the names of
any identifiers containing the % character. For a list of legal identifier characters,
see “Identifiers” in Chapter 1.
ASSUME CS Set to Wrong Value

With MASM 6.1 you do not need to use the ASSUME statement for the CS
register. Instead, MASM 6.1 generates an automatic ASSUME statement for
the code segment register to the current segment or group, as explained in
“Setting the ASSUME Directive for Segment Registers” in Chapter 2.

Additionally, MASM 6.1 does not allow explicit ASSUME statements for CS
that contradict the automatically set ASSUME statement.
MASM 5.1 allows CS to be assumed to the current segment, even if that
segment is a member of a group. With MASM 6.1, this results in warning
A4004:
cannot ASSUME CS
To avoid this warning with MASM 6.1, delete the ASSUME statement for CS.
Code Requiring Two-Pass Assembly

Unlike version 5.1, MASM 6.1 does most of its work on its first pass, then
performs as many subsequent passes as necessary. In contrast, MASM 5.1
always assembles in two source passes. As a result, you may need to revise or
delete some pass-dependent constructs under MASM 6.1.
Two-Pass Directives
To assure compatibility, MASM 6.1 supports 5.1 directives referring to two
passes. These include .ERR1, .ERR2, IF1, IF2, ELSEIF1, and ELSEIF2. For
second-pass constructs, you must specify OPTION SETIF2, as discussed in
“OPTION SETIF2,” page 377. Without OPTION SETIF2, the IF2 and .ERR2
directives cause error A2061:
[[ELSE]]IF2/.ERR2 not allowed : single-pass assembler
MASM 6.1 handles first-pass constructs differently. It treats the .ERR1

directive as .ERR, and the IF1 directive as IF.
The following examples show you how you can rewrite typical pass-sensitive
code for MASM 6.1:
u Declare var external only if not defined in current module:
; MASM 5.1:
IF2
IFNDEF var
EXTRN var:far
ENDIF
ENDIF
; MASM 6.1:
EXTERNDEF var:far

u Include a file of definitions only once to speed assembly:

; MASM 5.1:
IF1
INCLUDE file1.inc
ENDIF
; MASM 6.1:
INCLUDE FILE1.INC
u Generate a %OUT or .ERR message only once:

; MASM 5.1:
IF2
%OUT This is my message
ENDIF
IF2
.ERRNZ A NE B
ENDIF
; MASM 6.1:
ECHO This is my message
.ERRNZ A NE B <ASSERTION FAILURE: A NE B>
u Generate an error if a symbol is not defined but may be forward referenced:

; MASM 5.1:
IF2
.ERRNDEF var
ENDIF
; MASM 6.1:
.ERRNDEF var
For information on conditional directives, see “Conditional Directives,” Chapter

1.
IFDEF and IFNDEF with Forward-Referenced Identifiers

If you use a symbol name that has not yet been defined in an IFDEF or
IFNDEF expression, MASM 6.1 returns FALSE for the IFDEF expression and
TRUE for the IFNDEF expression. When OPTION M510 is enabled, the
assembler generates warning A6005:
expression condition may be pass-dependent
To resolve the warning, place the symbol definition before the conditional test.

Address Spans as Constants

The value of offsets calculated on the first assembly pass may not be the same
as those calculated on later passes. Therefore, you should avoid comparisons
with an address span, as in the following examples:
IF (OFFSET var1 - OFFSET var2) EQ 10
WHILE dx LT (OFFSET var1 - OFFSET var2)
REPEAT OFFSET var1 - OFFSET var2
However, the DUP operator allows such an expression as its count value. The
assembler evaluates the DUP count on every pass, so even expressions
involving forward references assemble correctly.
You can also use expressions containing span distances with the .ERR
directives, since the assembler evaluates these directives after calculating all
offsets:
.ERRE OFFSET var1 - OFFSET var2 - 10, <span incorrect>
.TYPE with Forward References

MASM 5.1 evaluates .TYPE on both assembly passes. This means it yields zero
on the first pass and nonzero on the second pass, if applied to an expression that
forward-references a symbol.
MASM 6.1 evaluates .TYPE only on the first assembly pass. As a result, if the
operand references a symbol that has not yet been defined, .TYPE yields a
value of zero. This means that .TYPE, if used in a conditional-assembly
construction, may yield different results in MASM 6.1 than in MASM 5.1.
Obsolete Features No Longer Supported

The following two features are no longer supported by MASM 6.1. Because
both are obscure features provided by early versions of the assembler, they
probably do not affect your MASM 5.1 code.
The ESC Instruction

MASM 6.1 no longer supports the ESC instruction, which was used to send
hand-coded commands to the coprocessor. Because MASM 6.1 recognizes and
assembles the full set of coprocessor mnemonics, the ESC instruction is not
necessary. Using the ESC instruction generates error A2205:
ESC instruction is obsolete: ignored
To update MASM 5.1 code, use the coprocessor instructions instead of ESC.

The MSFLOAT Binary Format

MASM 6.1 does not support the .MSFLOAT directive, which provided the
Microsoft Binary Format (MSB) for floating-point numbers in variable
initializers. Using the .MSFLOAT directive generates error A2204:
.MSFLOAT directive is obsolete: ignored
Use IEEE format or, if MSB format is necessary, initialize variables with
hexadecimal values. See “Storing Numbers in Floating-Point Format” in Chapter
6.
Using the OPTION Directive

The OPTION directive lets you control compatibility with MASM 5.1 code.
This section explains the differences in MASM 5.1 and MASM 6.1 behavior
that the OPTION directive can influence.
The OPTION M510 directive (or /Zm command-line option) initiates all aspects
of 5.1 compatibility mode. You can select from among specific characteristics of
MASM 5.1 behavior with the OPTION arguments discussed in following
sections. Each section also explains how to revise your code if you want to
remove OPTION directives from your MASM 5.1 code.
Note If your code includes both .MODEL and OPTION M510, the OPTION
M510 statement must appear first. Wherever this appendix suggests using
OPTION M510 in your code, you can set the /Zm command-line option
instead.
OPTION M510
This section discusses the M510 argument to the OPTION directive, which
selects the MASM 5.1 compatibility mode. In this mode, MASM 6.1
implements MASM 5.1 behavior relating to macros, offsets, scope of code
labels, structures, identifier names, identifier case, and other behaviors.
The OPTION M510 directive automatically sets the following:
OPTION OLDSTRUCTS ; MASM 5.1 structures
OPTION OLDMACROS ; MASM 5.1 macros
OPTION DOTNAME ; Identifiers may begin with a dot (.)
OPTION SETIF2:TRUE ; Two-pass code activates on every pass
If you do not have a .386, 386P .486, or 486P directive in your module, then
OPTION M510 adds:

OPTION EXPR16 ; 16-bit expression precision

; See "OPTION EXPR16," following

If you do not have a .MODEL directive in your module, OPTION M510 adds:
OPTION OFFSET:SEGMENT ; OFFSET operator defaults to
; segment-relative
; See "OPTION OFFSET," following
If you do not have a .MODEL directive with a language specifier in your

module, OPTION M510 also adds:
OPTION NOSCOPED ; Code labels are not local inside
; procedures
; See "OPTION NOSCOPED," following
OPTION PROC:PRIVATE ; Labels defined with PROC are not
; public by default
; See "OPTION PROC," following
If you want to remove OPTION M510 from your code (or /Zm from the
command line), add the OPTION directive arguments to your module according
to the conditions stated earlier.
There may be compatibility issues affecting your code that are supported under
OPTION M510, but are not covered by the other OPTION directive
arguments. Once you have modified your source code so it no longer requires
behavior supported by OPTION M510, you can replace OPTION M510 with
other OPTION directive arguments. These compatibility issues are discussed in
following sections.
Once you have replaced OPTION M510 with other forms of the OPTION
directive and your code works correctly, try removing the OPTION directives,
one at a time. Make appropriate source modifications as necessary, until your
code uses only MASM 6.1 defaults.
Reserved Keywords Dependent on CPU Mode with OPTION M510

With OPTION M510, keywords and instructions not available in the current
CPU mode (such as ENTER under .8086) are not treated as keywords. This
also means the USE32, FLAT, FAR32, and NEAR32 segment types and the
80386/486 registers are not keywords with a processor selection less than .386.
If you remove OPTION M510, any reserved word used as an identifier
generates a syntax error. You can either rename the identifiers or use OPTION
NOKEYWORD. For more information on OPTION NOKEYWORD, see
“OPTION NOKEYWORD,” later in this appendix.
Invalid Use of Instruction Prefixes with OPTION M510

Code without OPTION M510 generates errors for all invalid uses of the
instruction prefixes. OPTION M510 suppresses some of these errors to match
MASM 5.1 behavior. MASM 5.1 does not check for illegal usage of the
instruction prefixes LOCK, REP, REPE, REPZ, REPNE, and REPNZ.

Illegal usage of these prefixes results in error A2068:

instruction prefix not allowed
For more information on these instruction prefixes, see “Overview of String

Instructions” in Chapter 5. See also “Bug Fixes from MASM 5.1,” earlier in this
appendix.
Size of Constant Operands with OPTION M510

In MASM 5.1, a large constant value that can fit only in the processor’s default
word (4 bytes for .386 and .486, 2 bytes otherwise) is assigned a size attribute
of the default word size. The value of the constant affects the number of bytes
changed by the instruction. For example,
; Legal only with OPTION M510
mov [bx], 0100h
is legal in OPTION M510 mode. Since 0100h cannot fit in a byte, the
assembler interprets the value as a word.
Without OPTION M510, the assembler never assigns a size automatically. You
must state it explicitly with the PTR operator, as shown in the following
example:
; Without OPTION M510
mov [bx], WORD PTR 0100h
Code Labels when Defining Data with OPTION M510

MASM 5.1 allows a code label definition in a data definition statement if that
statement does not also define a data label. MASM 6.1 also allows such
definitions if OPTION M510 is enabled; otherwise it is illegal.
; Legal only with OPTION M510
MyCodeLabel: DW 0
SEG Operator with OPTION M510

In MASM 5.1, the SEG operator returns a label’s segment address unless the
frame is explicitly specified, in which case it returns the segment address of the
frame. A statement such as SEG DGROUP:var always returns DGROUP,
whereas SEG var always returns the segment address of var. OPTION M510
forces this same behavior in MASM 6.1.
If you do not use OPTION M510, the behavior of the SEG operator is
determined by the OPTION OFFSET directive, as described in “OPTION
OFFSET,” later in this appendix.
In MASM 6.1, the value returned by the SEG operator applied to a nonexternal
variable depends on compatibility mode:

u Without OPTION M510, SEG returns the address of the frame (the
segment, group, or the value assumed to the segment register) if one has
been explicitly set.
u With OPTION M510, SEG returns the group if one has been specified. In
the absence of a defined group, SEG returns the segment where the variable
is defined.
Expression Evaluation with OPTION M510

By default, MASM 6.1 changes the way expressions are evaluated. In MASM
5.1,
var-2[bx]
is parsed as
(var-2)[bx]
Without OPTION M510, you must rewrite the statement, since the assembler
parses it as
var-(2[bx])
which generates an error.
Length and Size of Labels with OPTION M510

With OPTION M510, you can apply the LENGTH and SIZE operators to any
label. For a code label, SIZE returns a value of 0FFFFh for NEAR and 0FFFEh
for FAR. LENGTH always returns a value of 1. For strings, SIZE and
LENGTH both return 1.
Without OPTION M510, SIZE returns values of 0FF01h, 0FF02h, 0FF04h,
0FF05h, and 0FF06h for SHORT, NEAR16, NEAR32, FAR16, and FAR32
labels, respectively. LENGTH returns 1 except when used with DUP, in which
case it returns the outermost count. For arrays initialized with DUP, SIZE
returns the length multiplied by the size of the type.
The LENGTHOF and SIZEOF operators in MASM 6.1 handle arrays much
more consistently. These operators return the number of data items and the
number of bytes in an initializer. For a description of SIZEOF and
LENGTHOF, see the following sections in Chapter 5: “Declaring and
Referencing Arrays,” “Declaring and Initializing Strings,” “Defining Structure
and Union Variables,” and “Defining Record Variables.”
Comparing Types Using EQ and NE with OPTION M510

With OPTION M510, the assembler converts types to a constant value before
comparisons with EQ and NE. Code types are converted to values of 0FFFFh

(near) or 0FFFEh (far). If OPTION M510 is not enabled, the assembler

converts types to constants only when comparing them with constants. Thus,
MASM 6.1 recognizes only equivalent qualified types as equal expressions.
For existing MASM 5.1 code, these distinctions affect only the use of the TYPE
operator in conjunction with EQ and NE. The following example illustrates how
the assembler compares types with and without compatibility mode:
MYSTRUCT STRUC
f1 DB 0
f2 DB 0
MYSTRUCT ENDS
; With OPTION M510
val = (TYPE MYSTRUCT) EQ WORD ; True: 2 EQ 2

val = 2 EQ WORD ; True: 2 EQ 2
val = WORD EQ WORD ; True: 2 EQ 2
val = SWORD EQ WORD ; True: 2 EQ 2
val = (TYPE MYSTRUCT) EQ WORD ; False: MyStruct NE WORD

val = 2 EQ WORD ; True: 2 EQ 2
val = WORD EQ WORD ; True: WORD EQ WORD
val = SWORD EQ WORD ; False: SWORD NE WORD
Use of Constant and PTR as a Type with OPTION M510

You can use a constant as the left operand to PTR in compatibility mode.
Otherwise, you must use a type expression. With OPTION M510, a constant
must have a value of 1 (BYTE), 2 (WORD), 4 (DWORD), 6 (FWORD), 8
(QWORD) or 10 (TBYTE). The assembler treats the constant as the

parenthesized type. Note that the TYPE operator yields a type expression, but
the SIZE operator yields a constant.
; With OPTION M510
MyData DW 0
mov WORD PTR [bx], 10 ; Legal

mov (TYPE MyData) PTR [bx], 10 ; Legal
mov (SIZE MyData) PTR [bx], 10 ; Legal
mov 2 ptr [bx], 10 ; Legal
MyData WORD 0
mov WORD PTR [bx], 10 ; Legal

mov (TYPE MyData) PTR [bx], 10 ; Legal
; mov (SIZE MyData) PTR [bx], 10 ; Illegal
; mov 2 PTR [bx], 10 ; Illegal
Structure Type Cast on Expressions with OPTION M510

In compatibility mode, use the PTR operator to type-cast a constant to a
structure type. This is most often done in data initializers to affect the CodeView
information of the data label. Without OPTION M510, the assembler generates
an error.
MYSTRC STRUC
f1 DB 0
MYSTRC ENDS
MyPtr DW MYSTRC PTR 0 ; Illegal without OPTION M510
In MASM 6.1, the initializer type does not influence CodeView’s type
information.
Hidden Coercion of OFFSET Expression Size with OPTION M510

When programming for the 80386 or 80486, the size of an OFFSET expression
can be 2 bytes for a symbol in a USE16 segment, or 4 bytes for a symbol in a
USE32 or FLAT segment. With OPTION M510, you can use a 32-bit
OFFSET expression in a 16-bit context. Without OPTION M510, you must use
the LOWWORD operator to convert the offset size.

.386
; With OPTION M510
seg32 SEGMENT USE32

MyLabel WORD 0
seg32 ENDS
seg16 SEGMENT USE16 'code' ; With OPTION M510:

mov ax, OFFSET MyLabel ; Legal
mov ax, LOWWORD OFFSET MyLabel ; Legal
mov eax, OFFSET MyLabel ; Legal
seg16 ENDS
seg32 SEGMENT USE32

MyLabel WORD 0
seg32 ENDS
seg16 SEGMENT USE16 'code' ; Without OPTION M510:

; mov ax, OFFSET MyLabel ; Illegal
mov ax, LOWWORD offset MyLabel ; Legal
mov eax, OFFSET MyLabel ; Legal
seg16 ENDS
Specifying Radixes with OPTION M510

If the current radix in your code is greater than 10 decimal, MASM 6.1 allows
the radix specifiers B (binary) and D (decimal) only in compatibility mode. You
must change B to Y for binary, and D to T for decimal, since both B and D are
legitimate hexadecimal values, making numbers such as 12D ambiguous. If you
want to keep B and D as radix specifiers when the current radix is greater than
10, you must specify OPTION M510. For more information about radixes, see
“Integer Constants and Constant Expressions” in Chapter 1.
Naming Conventions with OPTION M510

By default, MASM 5.1 does not write the names of public variables in
uppercase to the object file, even when a language type of PASCAL,
FORTRAN, or BASIC is specified.
Unless you use OPTION M510, these language types in MASM 6.1 write
identifier names in uppercase, even with the /Cp or /Cx command-line options.
When you link with /NOI, case must match in the object files to resolve
externals.

Length Significance of Symbol Names with OPTION M510

With MASM 5.1, only the first 31 characters of a symbol name are considered
significant, and only the first 31 characters of a public or external symbol name
are placed in the object file.
Without OPTION M510, the entire name is considered significant. The
maximum number of characters placed in the object file is controlled with the
/Hnumber command-line option, with a default of 247 (the maximum length of
an identifier in MASM 6.1).
String Defaults in Structure Variables with OPTION M510

In compatibility mode, a constant initializer can override a structure field
initialized with a string value. Without OPTION M510, only another string or a
list can override a string initializer. To update your code, surround the constant
override value with angle brackets or curly braces to indicate a list with one
element.
MTSTRUCT STRUCT
MyString BYTE "This is a string"
MTSTRUCT ENDS
; With OPTION M510
MyInst MTSTRUCT <0>
; Without OPTION M510, either of these statements is correct
MyInst MTSTRUCT <<0>>

MyInst MTSTRUCT {<0>}
Effects of the ? Initializer in Data Definitions with OPTION M510

As described in “Declaring and Initializing Strings” in Chapter 5, the assembler
treats the ? initializer as either zero or as an unspecified value. In compatibility
mode, however, the assembler always treats the ? initializer as zero unless it is
used with the DUP operator. In this case, the assembler allocates space, but
does not initialize it with any value.
Current Address Operator with OPTION M510

In compatibility mode, the current address operator ($) applied to a structure
returns the offset of the first byte of the structure. When OPTION M510 is not
enabled, $ returns the offset of the current field in the structure.
Segment Association for FAR Externals with OPTION M510

In MASM 5.1, you must place an EXTRN directive for a variable in the same
segment that holds the variable. For far data, this often entails opening and
closing a segment just to place the EXTRN statement.

MASM 6.1 offers much greater flexibility in where EXTERN and

EXTERNDEF statements can appear, as described in “Positioning External
Declarations” in Chapter 8. However, in compatibility mode, MASM 6.1
emulates the behavior of MASM 5.1.
Defining Aliases Using EQU with OPTION M510

In MASM 5.1, you can equate one symbol with another. These equates are
called “aliases.”
Unless you specify OPTION M510, MASM 6.1 does not allow aliases defined
with EQU. An immediate expression or text must appear as the right operand of
an EQU directive. Change aliases to use the TEXTEQU directive, described in
“Text Macros” in Chapter 9. This change may cause an expression to evaluate
differently.
The following examples illustrate the differences between MASM 5.1 code,
MASM 6.1 code with OPTION M510, and MASM 6.1 code without OPTION
M510:
; MASM 5.1 code
var1 EQU 3
var2 EQU var1 ; var2 taken as an alias
; var2 references var1 anywhere var2 is
; used as a symbol
; MASM 6.1 with OPTION M510

var1 EQU 3
var2 EQU var1 ; var2 taken as a var2 EQU <var1>
; var2 substituted for var1 whenever
; text macros substituted
; MASM 6.1 without OPTION M510

var1 EQU 3
var2 EQU var1 ; Treated as var2 EQU 3

Difference in Text Macro Expansions with OPTION M510

MASM 6.1 recursively expands text macros used as values, whereas MASM 5.1
simply replaces the text macro with its value. The following example illustrates
the difference:
; With OPTION M510
tm1 EQU <contains tm2>

tm2 EQU <value>
tm3 CATSTR tm1 ; == <contains tm2>
tm3 CATSTR tm1 ; == <contains value>
Conditional Directives and Missing Operands with OPTION M510

MASM 5.1 considers a missing argument to be a zero. MASM 6.1 requires an
argument unless OPTION M510 is enabled.
OPTION OLDSTRUCTS
This section describes changes in MASM 6.1 that apply to structures. With
OPTION OLDSTRUCTS or OPTION M510:
u You can use plus operator (+) in structure field references.
u Labels and structure field names cannot have the same name with OPTION
OLDSTRUCTS.
Plus Operator Not Allowed with Structures

By default, each reference to structure member names must use the dot
operator (.) to separate the structure variable name from the field name. You
cannot use the dot operator as the plus operator (+) or vice versa.
To convert your code so that it does not need OPTION OLDSTRUCTS:
u Qualify all structure field references.
u Change all uses of the dot operator ( . ) that occur outside of structure
references to use the plus operator ( + ).

If you remove OPTION OLDSTRUCTS from your code, the assembler

generates errors for all lines requiring change. Using the dot operator in any
context other than for a structure field results in error A2166:
structure field expected
Unqualified structure references result in error A2006:

undefined symbol : identifier
The following example illustrates how to change MASM 5.1 code from the old
structure references to the new type in MASM 6.1:
; OPTION OLDSTRUCTS (simulates MASM 5.1)
structname STRUC
a BYTE ?
b WORD ?
structname ENDS
structinstance structname <>
mov ax, [bx].b ; This code assembles

mov al, structinstance.a ; correctly only with
mov ax, [bx].4 ; OPTION OLDSTRUCTS
; or OPTION M510
; OPTION NOOLDSTRUCTS (the MASM 6.1 default)

structname STRUCT
a BYTE ?
b WORD ?
structname ENDS
structinstance structname <>
mov ax, [bx].structname.b ; Add qualifying type

mov al, structinstance.a ; No change needed
mov ax, [bx]+4 ; Change dot to plus
; Alternative methods in MASM 6.1

; Either this:
ASSUME bx:PTR structname
mov ax, [bx]
; or this:
mov ax, (structname PTR[bx]).b

Duplicate Structure Field Names

With the default, OPTION NOOLDSTRUCTS, label and structure field names
may have the same name. With OPTION OLDSTRUCTS (the MASM 5.1
default), labels and structure fields cannot have the same name. For more
information, see “Structures and Unions” in Chapter 5.
OPTION OLDMACROS
This section describes how MASM 5.1 and 6.1 differ in their handling of
macros. Without OPTION OLDMACROS or OPTION M510, MASM 6.1
changes the behavior of macros in several ways. If you want the MASM 5.1
macro behavior, add OPTION OLDMACROS or OPTION M510 to your
MASM 5.1 code.
Separating Macro Arguments with Commas

MASM 5.1 allows white spaces or commas to separate arguments to macros.
MASM 6.1 with OPTION NOOLDMACROS (the default) requires commas
between arguments. For example, in the macro call
MyMacro var1 var2 var3, var4
OPTION OLDMACROS causes the assembler to treat all four items as

separate arguments. With OPTION NOOLDMACROS, the assembler treats
var1 var2 var3
as one argument, since the items are not separated with commas. To convert
your macro code, replace spaces between macro arguments with a single
comma.
New Behavior with Ampersands in Macros

The default OPTION NOOLDMACROS causes the assembler to interpret
ampersands (&) within a macro differently than does MASM 5.1. MASM 5.1
requires one ampersand for each level of macro nesting. OPTION
OLDMACROS emulates this behavior.
Without OPTION OLDMACROS, MASM 6.1 removes ampersands only once
no matter how deeply nested the macro. To update your MASM 5.1 macros,
follow this simple rule: replace every sequence of ampersands with a single
ampersand. The only exception is when macro parameters immediately precede
and follow the ampersand, and both require substitution. In this case, use two
ampersands. For a description of the new rules, see “Substitution Operator” in
Chapter 9.

This example shows how to update a MASM 5.1 macro:

; OPTION OLDMACROS (emulates MASM 5.1 behavior)
createNames macro arg

irp tail, <Next, Last>
irp num, <1, 2>
; Define more names of the form: abcNext1?
arg&&tail&&&num&&&? label BYTE
ENDM
ENDM
ENDM
; OPTION NOOLDMACROS (the MASM 6.1 default)
createNames macro arg

for tail, <Next, Last> ; FOR is the MASM 6.1
for num, <1, 2> ; synonym for irp
; Define more names of the form: abcNext1?
arg&&tail&&num&? label BYTE
ENDM
ENDM
ENDM
OPTION DOTNAME
MASM 5.1 allows names of identifiers to begin with a period. The MASM 6.1
default is OPTION NODOTNAME. Adding OPTION DOTNAME to your
code enables the MASM 5.1 behavior.
If you don’t want to use this directive in your source code, rename the
identifiers whose names begin with a period.
OPTION EXPR16
MASM 5.1 treats expressions as 16-bit words if you do not specify .386 or
.386P directives. MASM 6.1 by default treats expressions as 32-bit words,
regardless of the CPU type. You can force MASM 6.1 to use the smaller
expression size with the OPTION EXPR16 statement.
Unless your MASM 5.1 code specifies .386 or .386P, OPTION M510 also sets
16-bit expression size. You can selectively disable this by following OPTION
M510 with the OPTION EXPR32 directive, which sets the size back to 32 bits.
You cannot have both OPTION EXPR32 and OPTION EXPR16 in your
program.
It may not be easy to determine the effect of changing from 16-bit internal
expression size to 32-bit size. In most cases, the 32-bit word size does not affect
the MASM 5.1 code. However, problems may arise because of differences in

intermediate values during evaluation of expressions. You can compare the files
for differences by generating listing files with the /Fl and /Sa command-line
options with and without OPTION EXPR16.
OPTION OFFSET
The information in this section is relevant only if your MASM 5.1 code does not
use the .MODEL directive. With no .MODEL, MASM 5.1 computes offsets
from the start of the segment, whereas MASM 6.1 computes offsets from the
start of the group. (With .MODEL, MASM 5.1 also computes offsets from the
start of the group.)
To force MASM 6.1 to emulate 5.1 behavior, specify either
OFFSET:SEGMENT or OPTION M510. Both directives cause the assembler
to compute offsets relative to the segment if you do not include .MODEL.
To selectively enable MASM 6.1 behavior, place the directive OPTION
OFFSET:GROUP after OPTION M510. In this case, you should ensure each
OFFSET statement has a segment override where appropriate. The following
example shows how OPTION OFFSET:SEGMENT affects code written for
MASM 5.1:
OPTION OFFSET:SEGMENT
MyGroup GROUP MySeg
MySeg SEGMENT 'data'

MyLabel LABEL BYTE
DW OFFSET MyLabel ; Relative to MySeg
DW OFFSET MyGroup:MyLabel ; Relative to MyGroup
DW OFFSET MySeg:MyLabel ; Relative to MySeg
MySeg ENDS
In the preceding example, the first OFFSET statement computes the offset of
MyLabel relative to MySeg. Without OFFSET:SEGMENT, MASM 6.1 returns
the offset relative to MyGroup. To maintain the correct behavior with
OFFSET:GROUP, specify a segment override, as shown in the following. The
other two OFFSET statements already include overrides, and so do not require
modification.
OPTION OFFSET:GROUP
MyGroup GROUP MySeg
MySeg SEGMENT 'data'

MyLabel LABEL BYTE
DW OFFSET MyGroup:MyLabel ; Relative to MyGroup
MySeg ENDS

When not in compatibility mode, the OPTION OFFSET directive determines

whether the SEG operator returns a value relative to the group or segment. With
OPTION M510, SEG is always segment-relative by default, regardless of the
current value of OPTION OFFSET.
OPTION NOSCOPED
The information in this section applies only if the .MODEL directive in your
MASM 5.1 code does not specify a language type. Without a language type,
MASM 5.1 assumes code labels in procedures have no “scope” — that is, the
labels are not local to the procedure. When not in compatibility mode, MASM
6.1 always gives scope to code labels, even without a language type.
To force MASM 5.1 behavior, specify either OPTION M510 or OPTION
NOSCOPED in your code. To selectively enable MASM 6.1 behavior, place
the directive OPTION SCOPED after OPTION M510.
To determine which labels require change, assemble the module without the
OPTION NOSCOPED directive. For each reference to a label that is not local,
the assembler generates error A2006:
undefined symbol : identifier
OPTION PROC
The information in this section applies only if the .MODEL directive in your
MASM 5.1 code does not specify a language type. Without a language type,
MASM 5.1 makes procedures private to the module. By default, MASM 6.1
makes procedures public. You can explicitly change the default visibility to
private with either OPTION M510, OPTION PROC:PRIVATE, or OPTION
PROC:EXPORT.
To selectively enable MASM 6.1 behavior, place the directive OPTION
PROC:PUBLIC after OPTION M510. You can override the default by adding
the PUBLIC or PRIVATE keyword to selected procedures. The following
example shows how to change MASM 5.1 code to keep a procedure private:
; MASM 5.1 (OPTION PROC:PRIVATE)
MyProc PROC NEAR
; MASM 6.1 (OPTION PROC:PUBLIC)

MyProc PROC NEAR PRIVATE
This is necessary only to avoid naming conflicts between public names in

multiple modules or libraries. The symbol table in a listing file shows the
visibility (public, private, or export) of each procedure.

OPTION NOKEYWORD
MASM 6.1 has several new keywords that MASM 5.1 does not recognize as
reserved. To resolve any conflicts, you can:
u Rename any offending symbols in your code.
u Selectively disable keywords with the OPTION NOKEYWORD directive.
The second option lets you retain the offending symbol names in your code by
forcing MASM 6.1 to not recognize them as keywords. For example,
OPTION NOKEYWORD:<INVOKE STRUCT>
removes the keywords INVOKE and STRUCT from the assembler’s list of
reserved words. However, you cannot then use the keywords in their intended
function, since the assembler no longer recognizes them.
The following list shows MASM 6.1 reserved words new since MASM 5.1:
.BREAK .UNTILCXZ FRSTORD
.CONTINUE .WHILE FRSTORW
.DOSSEG ADDR FSAVED
.ELSE ALIAS FSAVEW
.ELSEIF BSWAP FSTENVD
.ENDIF CARRY? FSTENVW
.ENDW CMPXCHG GOTO
.EXIT ECHO HIGHWORD
.IF EXTERN INVD
.LISTALL EXTERNDEF INVLPG
.LISTIF FAR16 INVOKE
.LISTMACRO FAR32 IRETDF
.LISTMACROALL FLAT IRETF
.NO87 FLDENVD LENGTHOF
.NOCREF FLDENVW LOOPD
.NOLIST FNSAVED LOOPED
.NOLISTIF FNSAVEW LOOPEW
.NOLISTMACRO FNSTENVD LOOPNED
.REPEAT FNSTENVW LOOPNEW
.STARTUP FOR LOOPNZD
.UNTIL FORC LOOPNZW

LOOPW PUSHCONTEXT SWORD

LOOPZW PUSHD SYSCALL
LOWWORD PUSHW TEXTEQU
LROFFSET REAL10 TR3
NEAR16 REAL4 TR4
NEAR32 REAL8 TR5
OPATTR REPEAT TYPEDEF
OPTION SBYTE UNION
OVERFLOW? SDWORD VARARG
PARITY? SIGN? WBINVD
POPAW SIZEOF WHILE
POPCONTEXT STDCALL XADD
PROTO STRUCT ZERO?
PUSHAW SUBTITLE
OPTION SETIF2
By default, MASM 6.1 does not recognize pass-dependent constructs. Both the
OPTION M510 and OPTION SETIF2 statements force MASM 6.1 to handle
MASM 5.1 constructs that activate on the second assembly pass, such as
.ERR2, IF2, and ELSEIF2.
Invoke the option like this:
OPTION SETIF2: {TRUE | FALSE}
When set to TRUE, OPTION SETIF2 forces all second-pass constructs to
activate on every assembly pass. When set to FALSE, second-pass constructs
do not activate on any pass. OPTION M510 implies OPTION SETIF2:TRUE.
Changes to Instruction Encodings

MASM 6.1 contains changes to the encodings for several instructions. In some
cases, the changes help optimize code size.
Coprocessor Instructions
For the 8087 coprocessor, MASM 5.1 adds an extra NOP before the no-wait
versions of coprocessor instructions. MASM 6.1 does not. In the rare case that
the missing NOP affects timing, insert NOP.

Chapter 1 Chapter Head 381
For the 80287 coprocessor or better, MASM 5.1 inserts FWAIT before certain
instructions. MASM 6.1 does not prefix any 80287, 80387, or 80486
coprocessor instruction with FWAIT, except for wait forms of instructions that
have a no-wait form.
RET Instruction
MASM 5.1 generates a 3-byte encoding for RET, RETN, or RETF instructions
with an operand value of zero, unless the operand is an external absolute. In this
case, MASM 5.1 ignores the parameter and generates a 1-byte encoding.
MASM 6.1 does the opposite. It ignores a zero operand for the return
instructions and generates a 1-byte encoding, unless the operand is an external
absolute. In this case, MASM 6.1 generates a 3-byte encoding.
Thus, you can suppress epilogue code in a procedure but still specify the default
size for RET by coding the return as
ret 0
Arithmetic Instructions
Versions 5.1 and 6.1 differ in the way they encode the arithmetic instructions
ADC, ADD, AND, CMP, OR, SUB, SBB, and XOR, under the following
conditions:
u The first operand is either AX or EAX.
u The second operand is a constant value between 0 and 127.
For the AX register, there is no size or speed difference between the two
encodings. For the EAX register, the encoding in MASM 6.1 is 2 bytes smaller.
The OPTION NOSIGNEXTEND directive forces the MASM 5.1 behavior for
AND, OR,
and XOR.

379
A P P E N D I X B
BNF Grammar
This appendix provides a complete description of symbols, operators, and

directives for MASM 6.1. It uses the Backus-Naur Form (BNF) for grammar
notation. You can use BNF grammar to determine the exact syntax for any
language component and find all available options for any MASM command.
BNF definitions consist of “nonterminals” and “terminals.” Nonterminals are
placeholders within a BNF definition, defined elsewhere in the BNF grammar.
Terminals are endpoints in a BNF definition, consisting of MASM 6.1
keywords. In this Appendix, all nonterminals appear in italics type and all
terminals appear in bold type.
BNF Conventions
The conventions use different font attributes for different items in the BNF. The
symbols and formats are as follows:
Attribute Description
nonterminal Italic type indicates nonterminals.
RESERVED Terminals in boldface type are literal reserved words and
symbols that must be entered as shown. Characters in this
context are always case insensitive.
[[ ]] Objects enclosed in double brackets ([[ ]]) are optional. The
brackets do not actually appear in the source code.
| A vertical bar indicates a choice between the items on each side
of the bar.
.8086 Underlined items indicate the default option if one is given.
default typeface Characters in the set described or listed can be used as
terminals in MASM statements.
Filename: LMAPGAPB.DOC Project:

How to Use the BNF Grammar

To illustrate the use of the BNF, Figure B.1 diagrams the definition of the
TYPEDEF directive, starting with the nonterminal typedefDir.
The entries under each horizontal brace in Figure B.1 are terminals (such as
NEAR16, NEAR32, FAR16, and FAR32) or nonterminals (such as qualifier,
qualifiedType, distance, and protoSpec) that can be further defined. Each
italicized nonterminal in the typedefDir definition is also an entry in the BNF.
Three vertical dots indicate a branching definition for a nonterminal that, for the
sake of simplicity, this figure does not illustrate.
The BNF grammar allows recursive definitions. For example, the grammar uses
qualifiedType as a possible definition for qualifiedType, which is also a
component of the definition for qualifier.
Figure B.1 BNF Definition of the TYPEDEF Directive
Nonterminal Definition
;; endOfLine
| comment
=Dir id = immExpr ;;
addOp +|-
aExpr term
| aExpr && term

Appendix B BNF Grammar 381
altId id
arbitraryText charList
asmInstruction mnemonic [[ exprList ]]
assumeDir ASSUME assumeList ;;
| ASSUME NOTHING ;;
assumeList assumeRegister
| assumeList , assumeRegister
assumeReg register : assumeVal
assumeRegister assumeSegReg
| assumeReg
assumeSegReg segmentRegister : assumeSegVal
assumeSegVal frameExpr
| NOTHING | ERROR
assumeVal qualifiedType
| NOTHING | ERROR
bcdConst [[ sign ]] decNumber
binaryOp == | != | >= | <= | > | < | &
bitDef bitFieldId : bitFieldSize [[ = constExpr ]]
bitDefList bitDef
| bitDefList , [[ ;; ]] bitDef
bitFieldId id
bitFieldSize constExpr
blockStatements directiveList
| .CONTINUE [[ .IF cExpr ]]
| .BREAK [[ .IF cExpr ]]
bool TRUE | FALSE
byteRegister AL | AH | BL | BH | CL | CH | DL | DH
cExpr aExpr
| cExpr || aExpr
character Any character with ordinal in the range 0–255
except linefeed (10)
charList character
| charList character
className string
commDecl [[ nearfar ]] [[ langType ]] id : commType
[[ : constExpr ]]
commDir COMM commList ;;
comment ; text ;;

commentDir COMMENT delimiter
text
text delimiter text ;;
commList commDecl
| commList , commDecl
commType type
| constExpr
constant digits [[ radixOverride ]]
constExpr expr
contextDir PUSHCONTEXT contextItemList ;;
| POPCONTEXT contextItemList ;;
contextItem ASSUMES | RADIX | LISTING | CPU | ALL
contextItemList contextItem
| contextItemList , contextItem
controlBlock whileBlock
| repeatBlock
controlDir controlIf
| controlBlock
controlElseif .ELSEIF cExpr ;;
directiveList
[[ controlElseif ]]
controlIf .IF cExpr ;;
directiveList
[[ controlElseif ]]
[[ .ELSE ;;
directiveList ]]
.ENDIF ;;
coprocessor .8087 | .287 | .387 | .NO87
crefDir crefOption ;;
crefOption .CREF
| .XCREF [[ idList ]]
| .NOCREF [[ idList ]]
cxzExpr expr
| ! expr
| expr == expr
| expr != expr
dataDecl DB | DW | DD | DF | DQ | DT | dataType | typeId
dataDir [[ id ]] dataItem ;;

dataItem dataDecl scalarInstList
| structTag structInstList
| typeId structInstList
| unionTag structInstList
| recordTag recordInstList
dataType BYTE | SBYTE | WORD | SWORD | DWORD
| SDWORD | FWORD | QWORD | TBYTE
| REAL4 | REAL8 | REAL10
decdigit 0|1|2|3|4|5|6|7|8|9
decNumber decdigit
| decNumber decdigit
delimiter Any character except whiteSpaceCharacter
digits decdigit
| digits decdigit
| digits hexdigit
directive generalDir
| segmentDef
directiveList directive
| directiveList directive
distance nearfar
| NEAR16 | NEAR32 | FAR16 | FAR32
e01 e01 orOp e02
| e02
e02 e02 AND e03
| e03
e03 NOT e04
| e04
e04 e04 relOp e05
| e05
e05 e05 addOp e06
| e06
e06 e06 mulOp e07
| e06 shiftOp e07
| e07
e07 e07 addOp e08
| e08
e08 HIGH e09
| LOW e09
| HIGHWORD e09
| LOWWORD e09
| e09

e09 OFFSET e10
| SEG e10
| LROFFSET e10
| TYPE e10
| THIS e10
| e09 PTR e10
| e09 : e10
| e10
e10 e10 . e11
| e10 [[ expr ]]
| e11
e11 ( expr )
| [[ expr ]]
| WIDTH id
| MASK id
| SIZE sizeArg
| SIZEOF sizeArg
| LENGTH id
| LENGTHOF id
| recordConst
| string
| constant
| type
| id
|$
| segmentRegister
| register
| ST
| ST ( expr )
echoDir ECHO arbitraryText ;;
%OUT arbitraryText ;;
elseifBlock elseifStatement ;;
directiveList
[[ elseifBlock ]]
elseifStatement ELSEIF constExpr
| ELSEIFE constExpr
| ELSEIFB textItem
| ELSEIFNB textItem
| ELSEIFDEF id
| ELSEIFNDEF id
| ELSEIFDIF textItem , textItem
| ELSEIFDIFI textItem , textItem
| ELSEIFIDN textItem , textItem
| ELSEIFIDNI textItem , textItem
| ELSEIF1
| ELSEIF2

endDir END [[ immExpr ]] ;;
endpDir procId ENDP ;;
endsDir id ENDS ;;
equDir textMacroId EQU equType ;;
equType immExpr
| textLiteral
errorDir errorOpt ;;
errorOpt .ERR [[ textItem ]]
| .ERRE constExpr [[ optText ]]
| .ERRNZ constExpr [[ optText ]]
| .ERRB textItem [[ optText ]]
| .ERRNB textItem [[ optText ]]
| .ERRDEF id [[ optText ]]
| .ERRNDEF id [[ optText ]]
| .ERRDIF textItem , textItem [[ optText ]]
| .ERRDIFI textItem , textItem [[ optText ]]
| .ERRIDN textItem , textItem [[ optText ]]
| .ERRIDNI textItem , textItem [[ optText ]]
| .ERR1 [[ textItem ]]
| .ERR2 [[ textItem ]]
exitDir .EXIT [[ expr ]] ;;
exitmDir: EXITM
| EXITM textItem
exponent E [[ sign ]] decNumber
expr SHORT e05
| .TYPE e01
| OPATTR e01
| e01
exprList expr
| exprList , expr
externDef [[ langType ]] id [[ ( altId ) ]] : externType
externDir externKey externList ;;
externKey EXTRN | EXTERN | EXTERNDEF
externList externDef
| externList , [[ ;; ]] externDef
externType ABS
| qualifiedType
fieldAlign constExpr
fieldInit [[ initValue ]]
| structInstance

fieldInitList fieldInit
| fieldInitList , [[ ;; ]] fieldInit
fileChar delimiter
fileCharList fileChar
| fileCharList fileChar
fileSpec fileCharList
| textLiteral
flagName ZERO? | CARRY? | OVERFLOW?
| SIGN? | PARITY?
floatNumber [[ sign ]] decNumber . [[ decNumber ]] [[ exponent ]]
| digits R
| digits r
forcDir FORC | IRPC
forDir FOR | IRP
forParm id [[ : forParmType ]]
forParmType REQ
| = textLiteral
frameExpr SEG id
| DGROUP : id
| segmentRegister : id
| id
generalDir modelDir | segOrderDir | nameDir
| includeLibDir | commentDir
| groupDir | assumeDir
| structDir | recordDir | typedefDir
| externDir | publicDir | commDir | protoTypeDir
| equDir | =Dir | textDir
| contextDir | optionDir | processorDir
| radixDir
| titleDir | pageDir | listDir
| crefDir | echoDir
| ifDir | errorDir | includeDir
| macroDir | macroCall | macroRepeat | purgeDir
| macroWhile | macroFor | macroForc
| aliasDir
gpRegister AX | EAX | BX | EBX | CX | ECX | DX | EDX
| BP | EBP | SP | ESP | DI | EDI | SI | ESI
groupDir groupId GROUP segIdList
groupId id
hexdigit a|b|c|d|e|f
|A|B|C|D|E|F

id alpha
| id alpha
| id decdigit
idList id
| idList , id
ifDir ifStatement ;;
directiveList
[[ elseifBlock ]]
[[ ELSE ;;
directiveList ]]
ENDIF ;;
ifStatement IF constExpr
| IFE constExpr
| IFB textItem
| IFNB textItem
| IFDEF id
| IFNDEF id
| IFDIF textItem , textItem
| IFDIFI textItem , textItem
| IFIDN textItem , textItem
| IFIDNI textItem , textItem
| IF1
| IF2
immExpr expr
includeDir INCLUDE fileSpec ;;
includeLibDir INCLUDELIB fileSpec ;;
initValue immExpr
| string
|?
| constExpr DUP ( scalarInstList )
| floatNumber
| bcdConst
inSegDir [[ labelDef ]] inSegmentDir
inSegDirList inSegDir
| inSegDirList inSegDir

inSegmentDir instruction
| dataDir
| controlDir
| startupDir
| exitDir
| offsetDir
| labelDir
| procDir [[ localDirList ]] [[ inSegDirList ]] endpDir
| invokeDir
| generalDir
instrPrefix REP | REPE | REPZ | REPNE | REPNZ | LOCK
instruction [[ instrPrefix ]] asmInstruction
invokeArg register :: register
| expr
| ADDR expr
invokeDir INVOKE expr [[ , [[ ;; ]] invokeList ]] ;;
invokeList invokeArg
| invokeList , [[ ;; ]] invokeArg
keyword Any reserved word
keywordList keyword
| keyword keywordList
labelDef id :
| id ::
| @@:
labelDir id LABEL qualifiedType ;;
langType C | PASCAL | FORTRAN | BASIC
| SYSCALL | STDCALL
listDir listOption ;;
listOption .LIST
| .NOLIST | .XLIST
| .LISTALL
| .LISTIF | .LFCOND
| .NOLISTIF | .SFCOND
| .TFCOND
| .LISTMACROALL | .LALL
| .NOLISTMACRO | .SALL
| .LISTMACRO | .XALL
localDef LOCAL idList ;;
localDir LOCAL parmList ;;
localDirList localDir
| localDirList localDir

localList localDef
| localList localDef
macroArg % constExpr
| % textMacroId
| % macroFuncId ( macroArgList )
| string
| arbitraryText
| < arbitraryText >
macroArgList macroArg
| macroArgList , macroArg
macroBody [[ localList ]]
macroStmtList
macroCall id macroArgList ;;
| id ( macroArgList )
macroDir id MACRO [[ macroParmList ]] ;;
macroBody
ENDM ;;
macroFor forDir forParm , < macroArgList > ;;
macroBody
ENDM ;;
macroForc forcDir id , textLiteral ;;
macroBody
ENDM ;;
macroFuncId id
macroId macroProcId
| macroFuncId
macroIdList macroId
| macroIdList , macroId
macroLabel id
macroParm id [[ : parmType ]]
macroParmList macroParm
| macroParmList , [[ ;; ]] macroParm
macroProcId id
macroRepeat repeatDir constExpr ;;
macroBody
ENDM ;;
macroStmt directive
| exitmDir
| : macroLabel
| GOTO macroLabel

macroStmtList macroStmt ;;
| macroStmtList macroStmt ;;
macroWhile WHILE constExpr ;;
macroBody
ENDM ;;
mapType ALL | NONE | NOTPUBLIC
memOption TINY | SMALL | MEDIUM | COMPACT
| LARGE | HUGE | FLAT
mnemonic Instruction name
modelDir .MODEL memOption [[ , modelOptlist ]] ;;
modelOpt langType
| stackOption
modelOptlist modelOpt
| modelOptlist , modelOpt
module [[ directiveList ]] endDir
mulOp * | / | MOD
nameDir NAME id ;;
nearfar NEAR | FAR
nestedStruct structHdr [[ id ]] ;;
structBody
ENDS ;;
offsetDir offsetDirType ;;
offsetDirType EVEN
| ORG immExpr
| ALIGN [[ constExpr ]]
offsetType GROUP | SEGMENT | FLAT
oldRecordFieldList [[ constExpr ]]
| oldRecordFieldList , [[ constExpr ]]
optionDir OPTION optionList ;;

optionItem CASEMAP : mapType
| DOTNAME | NODOTNAME
| EMULATOR | NOEMULATOR
| EPILOGUE : macroId
| EXPR16 | EXPR32
| LANGUAGE : langType
| LJMP | NOLJMP
| M510 | NOM510
| NOKEYWORD : < keywordList >
| NOSIGNEXTEND
| OFFSET : offsetType
| OLDMACROS | NOOLDMACROS
| OLDSTRUCTS | NOOLDSTRUCTS
| PROC : oVisibility
| PROLOGUE : macroId
| READONLY | NOREADONLY
| SCOPED | NOSCOPED
| SEGMENT : segSize
| SETIF2 : bool
optionList optionItem
| optionList , [[ ;; ]] optionItem
optText , textItem
orOp OR | XOR
oVisibility PUBLIC | PRIVATE | EXPORT
pageDir PAGE [[ pageExpr ]] ;;
pageExpr +
| [[ pageLength ]] [[ , pageWidth ]]
pageLength constExpr
pageWidth constExpr
parm parmId [[ : qualifiedType ]]
| parmId [[ constExpr ]] [[ : qualifiedType ]]
parmId id
parmList parm
| parmList , [[ ;; ]] parm
parmType REQ
| = textLiteral
| VARARG
pOptions [[ distance ]] [[ langType ]] [[ oVisibility ]]
primary expr binaryOp expr
| flagName
| expr

procDir procId PROC [[ pOptions ]] [[ < macroArgList > ]]
[[ usesRegs ]] [[ procParmList ]]
processor .8086
| .186
| .286 | .286C | .286P
| .386 | .386C | .386P
| .486 | .486P
processorDir processor ;;
| coprocessor ;;
procId id
procParmList [[ , [[ ;; ]] parmList ]]
[[ , [[ ;; ]] parmId :VARARG]]
protoArg [[ id ]] : qualifiedType
protoArgList [[ , [[ ;; ]] protoList ]]
[[ , [[ ;; ]] [[ id ]] :VARARG ]]
protoList protoArg
| protoList , [[ ;; ]] protoArg
protoSpec [[ distance ]] [[ langType ]] [[ protoArgList ]]
| typeId
protoTypeDir id PROTO protoSpec
pubDef [[ langType ]] id
publicDir PUBLIC pubList ;;
pubList pubDef
| pubList , [[ ;; ]] pubDef
purgeDir PURGE macroIdList
qualifiedType type
| [[ distance ]] PTR [[ qualifiedType ]]
qualifier qualifiedType
| PROTO protoSpec
quote “
|‘
radixDir .RADIX constExpr ;;
radixOverride h|o|q|t|y
|H|O|Q|T|Y
recordConst recordTag { oldRecordFieldList }
| recordTag < oldRecordFieldList >
recordDir recordTag RECORD bitDefList ;;
recordFieldList [[ constExpr ]]
| recordFieldList , [[ ;; ]] [[ constExpr ]]

recordInstance { [[ ;; ]] recordFieldList [[ ;; ]] }
| < oldRecordFieldList >
| constExpr DUP ( recordInstance )
recordInstList recordInstance
| recordInstList , [[ ;; ]] recordInstance
recordTag id
register specialRegister
| gpRegister
| byteRegister
regList register
| regList register
relOp EQ | NE | LT | LE | GT | GE
repeatBlock .REPEAT ;;
blockStatements ;;
untilDir ;;
repeatDir REPEAT | REPT
scalarInstList initValue
| scalarInstList , [[ ;; ]] initValue
segAlign BYTE | WORD | DWORD | PARA | PAGE
segAttrib PUBLIC
| STACK
| COMMON
| MEMORY
| AT constExpr
| PRIVATE
segDir .CODE [[ segId ]]
| .DATA
| .DATA?
| .CONST
| .FARDATA [[ segId ]]
| .FARDATA? [[ segId ]]
| .STACK [[ constExpr ]]
segId id
segIdList segId
| segIdList , segId
segmentDef segmentDir [[ inSegDirList ]] endsDir
| simpleSegDir [[ inSegDirList ]] [[ endsDir ]]
segmentDir segId SEGMENT [[ segOptionList ]] ;;
segmentRegister CS | DS | ES | FS | GS | SS

segOption segAlign
| segRO
| segAttrib
| segSize
| className
segOptionList segOption
| segOptionList segOption
segOrderDir .ALPHA | .SEQ | .DOSSEG | DOSSEG
segRO READONLY
segSize USE16 | USE32 | FLAT
shiftOp SHR | SHL
sign -|+
simpleExpr ( cExpr )
| primary
simpleSegDir segDir ;;
sizeArg id
| type
| e10
specialChars : | . | [[ | ]] | ( | ) | < | > | { | }
|+|-|/|*|&|%|!
|’|\|=|;|,|“
| whiteSpaceCharacter
| endOfLine
specialRegister CR0 | CR2 | CR3
| DR0 | DR1 | DR2 | DR3 | DR6 | DR7
| TR3 | TR4 | TR5 | TR6 | TR7
stackOption NEARSTACK | FARSTACK
startupDir .STARTUP ;;
stext stringChar
| stext stringChar
string quote [[ stext ]] quote
stringChar quote quote
| Any character except quote
structBody structItem ;;
| structBody structItem ;;
structDir structTag structHdr [[ fieldAlign ]]
[[, NONUNIQUE ]] ;;
structBody
structTag ENDS ;;
structHdr STRUC | STRUCT | UNION

structInstance < [[ fieldInitList ]] >
| { [[ ;; ]] [[ fieldInitList ]] [[ ;; ]] }
| constExpr DUP ( structInstList )
structInstList structInstance
| structInstList , [[ ;; ]] structInstance
structItem dataDir
| generalDir
| offsetDir
| nestedStruct
structTag id
term simpleExpr
| ! simpleExpr
text textLiteral
| text character
| ! character text
| character
| ! character
textDir id textMacroDir ;;
textItem textLiteral
| textMacroId
| % constExpr
textLen constExpr
textList textItem
| textList , [[ ;; ]] textItem
textLiteral < text >;;
textMacroDir CATSTR [[ textList ]]
| TEXTEQU [[ textList ]]
| SIZESTR textItem
| SUBSTR textItem , textStart [[ , textLen ]]
| INSTR [[ textStart , ]] textItem , textItem
textMacroId id
textStart constExpr
titleDir titleType arbitraryText ;;
titleType TITLE | SUBTITLE | SUBTTL
type structTag
| unionTag
| recordTag
| distance
| dataType
| typeId
typedefDir typeId TYPEDEF qualifier

typeId id
unionTag id
untilDir .UNTIL cExpr ;;
.UNTILCXZ [[ cxzExpr ]] ;;
usesRegs USES regList
whileBlock .WHILE cExpr ;;
blockStatements ;;
.ENDW
whiteSpaceCharacter ASCII 8, 9, 11–13, 26, 32

397
A P P E N D I X
Generating and Reading

Assembly Listings
A listing file shows precisely how the assembler translates your source file into
machine code. The listing documents the assembler’s assumptions, memory
allocations, and optimizations.
MASM creates an assembly listing of your source file whenever you do one of
the following:
u Select the appropriate option in PWB.
u Use one of the related source code directives.
u Specify the /Fl option on the MASM command line.
The assembly listing contains both the statements in the source file and the
binary code (if any) generated for each statement. The listing also shows the
names and values of all labels, variables, and symbols in your file.
The assembler creates tables for macros, structures, unions, records, segments,
groups, and other symbols, and places the tables at the end of the assembly
listing. Only the types of symbols encountered in the program are included. For
example, if your program has no macros, the symbol table does not have a
macros section.
Generating Listing Files

To generate a listing file from within PWB, follow these steps:
1. From the Options menu, choose MASM Options.
2. In the MASM Options dialog box, choose Set Debug or Release Options.
The dialog box for Set Debug or Release Options lists the choices summarized
in Table C.1. This table also shows the equivalent source code directives and
command-line options.
Filename: LMAPGAPC.DOC Project:

Table C.1 Options for Generating or Modifying Listing Files
To generate this information:

To generate this In source From command
information: In PWB1, select: code, enter: line, enter:
Default listing — includes all Generate Listing File .LIST (default) /Fl
assembled lines
Turn off all source Generate Listing File (turn .NOLIST —
listings (overrides all listing off) (synonym = .SFCOND)
directives)
List all source lines, including Include All Source Lines .LISTALL /Fl /Sa
false conditionals and
generated code
Show instruction timings List Instruction — /Fl /Sc
Timings
Show assembler-generated List Generated — /Fl /Sg
code Instructions
Include false conditionals2 List False .LISTIF /Fl /Sx
Conditionals (synonym = .LFCOND)
Suppress listing of any List False Conditionals .NOLISTIF —
subsequent conditional blocks (turn off) (synonym = .SFCOND)
whose condition is false
Toggle between .LISTIF and — .TFCOND —
.NOLISTIF
Suppress symbol table Generate Symbol Table — /Fl /Sn
generation (turn off the default)
List all processed macro — .LISTMACROALL —
statements (synonym = .LALL)
List only instructions, data, — .LISTMACRO —
and segment directives in (default)
macros (synonym = .XALL)
Turn off all listing during — .NOLISTMACRO —
macro expansion (synonym = .SALL)
Specify title for each page — TITLE name /St name
(use only once per file)
Specify subtitle for page — SUBTITLE name /Ss name
Designate page length and line — PAGE /Sp length
width, increment section [[length,width]][[+]] /Sl width
number, or generate page
breaks
Generate first-pass listing — — /Ep
1 Select MASM Options from the Options menu, then choose Set Dialog Options from the MASM Options dialog box.

Appendix C Generating and Reading Assembly Listings 399
2 See “Conditional Directives” in Chapter 1
Precedence of Command-Line Options and Listing Directives

Since command-line options and source code directives can specify opposite
behavior for the same listing file option, the assembler interprets the commands
according to the following precedence levels. Selecting PWB options is
equivalent to specifying /Fl /Sx on the command line:
u /Sa overrides any source code directives that suppress listing.
u Source code directives override all command-line options except /Sa.
u .NOLIST overrides other listing directives such as .NOLISTIF and
.LISTMACROALL.
u The /Sx, /Ss, /Sp, and /Sl options set initial values for their respective
features. Directives in the source file can override these command-line
options.
Reading the Listing File

The first half of the listing shows macros from the include file DOS.MAC,
structure declarations, and data. After the .DATA directive, the columns on the
left show offsets and initialized byte values within the data segment.
Instructions begin after the .CODE directive. The three columns on the left
show offsets, instruction timings, and binary code generated by the assembler.
The columns on the right list the source statements exactly as they appear in the
source file or as expanded by a macro. Various symbols and abbreviations in the
middle column provide information about the code, as explained in the following
section. The subsequent section, “Symbols and Abbreviations,” explains the
meanings of listing symbols.
Generated Code
The assembler lists the code generated from the statements of a source file.
With the /Sc command-line switch, which generates instruction timings, each
line has this syntax:
offset [[timing]] [[code]]
The offset is the offset from the beginning of the current code segment. The
timing shows the number of cycles the processor needs to execute the
instruction. The value of timing reflects the CPU type; for example, specifying
the .386 directive produces instruction timings for the 80386 processor. If the
statement generates code or data, code shows the numeric value in hexadecimal

notation if the value is known at assembly time. If the value is calculated at run
time, the assembler indicates what action is necessary to compute the value.
When assembling under the default .8086 directive, timing includes an effective
address value if the instruction accesses memory. The 80186/486 processors do
not use effective address values. For more information on effective address
timing, see the “Processor” section in the Reference book.
Error Messages
If any errors occur during assembly, each error message and error number
appears directly below the statement where the error occurred. An example of
an error line and message is:
mov ax, [dx][di]
listtst.asm(77): error A2031: must be index or base register
Symbols and Abbreviations

The assembler uses the symbols and abbreviations shown in Table C.2 to
indicate addresses that need to be resolved by the linker or values that were
generated in a special way. The example in this section illustrates many of these
symbols.
The example listing was produced using “List Generated Instructions” and “List
Instruction Timings” in PWB. These options correspond to the ML command-
line switches /Fl /Sg /Sc.
Table C.2 Symbols and Abbreviations in Listings
Character Meaning
C Line from include file
= EQU or equal-sign (=) directive
nn[xx] DUP expression: nn copies of the value xx
---- Segment/group address (linker must resolve)
R Relocatable address (linker must resolve)
* Assembler-generated code
E External address (linker must resolve)
n Macro-expansion nesting level (+ if more than 9)
| Operator size override
& Address size override
nn: Segment override in statement
nn/ REP or LOCK prefix instruction

Table C.3 explains the five symbols that may follow timing values in your
listing. The Reference book will help you determine correct timings for those
values marked with a symbol.

Table C.3 Symbols in Timing Column

Symbol Meaning
m Add cycles depending on next executed instruction.
n Add cycles depending on number of iterations or size of data.
p Different timing value in protected mode.
+ Add cycles depending on operands or combination of the preceding.
, Separates two values for “jump taken” and “jump not taken.”
Microsoft (R) Macro Assembler Version 6.10 09/20/00 12:00:00

listtst.asm Page 1 - 1
.MODEL small, c
.386
.DOSSEG
.STACK 256
INCLUDE dos.mac
C StrDef MACRO name1, text
C name1 BYTE &text
C BYTE 13d, 10d, '$'
C l&name1 EQU LENGTHOF name1
C ENDM
C
C Display MACRO string
C mov ah, 09h
C mov dx, OFFSET string
C int 21h
C ENDM
= 0020 num EQU 20h
COLOR RECORD b:1, r:3=1, i:1=1, f:3=7
= 35 value TEXTEQU %3 + num
= 32 tnum TEXTEQU %num
= 04 strpos TEXTEQU @InStr( , <person>,
<son> )
PutStr PROTO pMsg:PTR BYTE
0004 DATE STRUCT

0000 01 month BYTE 1
0001 01 day BYTE 1
0002 0000 year WORD ?
DATE ENDS

0002 U1 UNION
0000 0028 fsize WORD 40
bsize BYTE 60
U1 ENDS
0000 .DATA
0000 00000000 ddData DWORD ?

0004 1F text COLOR <>
0005 01 14 07C9 today DATE <01, 20, 1993>
0009 00 flag BYTE 0
000A 001E [ buffer WORD 30 DUP (0)
0000
]
StrDef ending, "Finished."

0046 46 69 6E 69 73 68 1 ending BYTE "Finished."
65 64 2E
004F 0D 0A 24 1 BYTE 13d, 10d, '$'
= 0009 1 lending EQU LENGTHOF ending
0052 54 68 69 73 20 69 Msg BYTE "This is a
string","0"
73 20 61 20
73 74 72 69
6E 67 30
float TYPEDEF REAL4

FPBYTE TYPEDEF FAR PTR BYTE
0063 ---- 0052 R FPMSG FPBYTE Msg
NPWORD TYPEDEF NEAR PTR WORD
PVOID TYPEDEF PTR
PPBYTE TYPEDEF PTR PBYTE
0000 .CODE
.STARTUP
0000 *@Startup:
0000 2 B8 ---- R * mov ax, DGROUP
0003 2p 8E D8 * mov ds, ax
0005 2 8C D3 * mov bx, ss
0007 2 2B D8 * sub bx, ax
0009 3 C1 E3 04 * shl bx, 004h
000C 2p 8E D0 * mov ss, ax
000E 2 03 E3 * add sp, bx
EXTERNDEF work:NEAR
0010 7m E8 0000 E call work

INVOKE PutStr, ADDR msg

0013 2 68 0052 R * push OFFSET Msg
0016 7m E8 0029 * call PutStr
0019 2 83 C4 02 * add sp, 00002h
001C 2 B8 ---- R mov ax, @data

001F 2p 8E C0 mov es, ax
0021 2 B0 63 mov al, 'c'
0023 4 26: 8B 0E mov cx, es:num
0020
0028 2 BF 0052 mov di, 82
002B 7n F2/ AE repne scasb
002D 4 66| A1 0000 R mov eax, ddData
0031 6 67& FE 03 inc BYTE PTR [ebx]
EXTERNDEF morework:NEAR
0034 7m E8 0000 E call morework
Display ending
0037 2 B4 09 1 mov ah, 09h
0039 2 BA 0046 R 1 mov dx, OFFSET ending
003C 37 CD 21 1 int 21h
.EXIT
003E 2 B4 4C * mov ah, 04Ch
0040 37 CD 21 * int 021h
0042 PutStr PROC pMsg:PTR BYTE
0042 2 55 bp * push
0043 4 8B EC bp, sp * mov
0045 2 B4 02 mov ah, 02H
0047 4 8B 7E 04 mov di, pMsg
004A 4 8A 15 mov dl, [di]
mov ax, [dx][di]
listtst.asm(77): error A2031: must be index or base register
.WHILE (dl)
004C 7m EB 10 * jmp @C0001
0059 *@C0002:
0059 37 CD 21 int 21h
005B 2 47 inc di
005C 4 8A 15 mov dl, [di]
.ENDW
005E *@C0001:
005E 2 0A D2 * or dl, dl
0060 7m,3 75 F7 * jne @C0002
ret

0062 4 5D * pop bp
0063 10m C3 * ret 00000h
0064 PutStr ENDP
END
Reading Tables in a Listing File

The tables at the end of a listing file list the macros, structures, unions, records,
segments, groups, and symbols that appear in a source file. These tables are not
printed in the previous sample listing, but are summarized as follows.
Macro Table
Lists all macros in the main file or the include files. Differentiates between
macro functions and macro procedures.
Structures and Unions Table

Provides the size in bytes of the structure or union and the offset of each field.
The type of each field is also given.
Record Table
“Width” gives the number of bits of the entire record. “Shift” provides the offset
in bits from the low-order bit of the record to the low-order bit of the field.
“Width” for fields gives the number of bits in the field. “Mask” gives the
maximum value of the field, expressed in hexadecimal notation. “Initial” gives
the initial value supplied for the field.
Type Table
The “Size” column in this table gives the size of the TYPEDEF type in bytes,
and the “Attr” column gives the base type for the TYPEDEF definition.
Segment and Group Table

“Size” specifies whether the segment is 16 bit or 32 bit. “Length” gives the size
of the segment in bytes. “Align” gives the segment alignment (WORD, PARA,
and so on). “Combine” gives the combine type (PUBLIC, STACK, and so on).
“Class” gives the segment’s class (CODE, DATA, STACK, or CONST).
Procedures, Parameters, and Locals

Gives the types and offsets from BP of all parameters and locals defined in each
procedure, as well as the size and memory location of each procedure.
Symbol Table
All symbols (except names for macros, structures, unions, records, and
segments) are listed in a symbol table at the end of the listing. The “Name”

column lists the names in alphabetical order. The “Type” column lists each
symbol’s type.
The length of a multiple-element variable, such as an array or string, is the
length of a single element, not the length of the entire variable.
If the symbol represents an absolute value defined with an EQU or equal sign
(=) directive, the “Value” column shows the symbol’s value. The value may be
another symbol, a string, or a constant numeric value (in hexadecimal),
depending on the type. If the symbol represents a variable or label, the “Value”
column shows the symbol’s hexadecimal offset from the beginning of the
segment in which it is defined.
The “Attr” column shows the attributes of the symbol. The attributes include
the name of the segment (if any) in which the symbol is defined, the scope of
the symbol, and the code length. A symbol’s scope is given only if the symbol is
defined using the EXTERN and PUBLIC directives. The scope can be external,
global, or communal. The “Attr” column is blank if the symbol has no attribute.


407
A P P E N D I X D
MASM Reserved Words
This appendix lists the reserved words recognized by MASM. They are divided
primarily by their use in the language. The primary categories are:
u Operands and symbols
u Registers
u Operators and directives
u Processor instructions
u Coprocessor instructions
Reserved words in MASM 6.1 are reserved under all CPU modes. Words
enabled in .8086 mode, the default, can be used in all higher CPU modes. To
use words from subcategories such as “Special Operands for the 80386” (later
in this appendix) requires .386 mode or higher.
You can disable the recognition of any reserved word specified in this appendix
by setting the NOKEYWORD option for the OPTION directive. Once
disabled, the word can be used in any way as a user-defined symbol (provided
the word is a valid identifier). If you want to remove the STR instruction, the
MASK operator, and the NAME directive, for instance, from the set of words
MASM recognizes as reserved, add this statement to your program:
OPTION NOKEYWORD:<STR MASK NAME>
Words in this appendix identified with an asterisk (*) are new since MASM 5.1.
Operands and Symbols

The words on the two lists in this section are the operands to certain directives.
They have special meaning to the assembler. The words on the first list are not
reserved words. They can be used in every way as normal identifiers, without
affecting their use as operands to directives. The assembler interprets their use
from context.
Filename: LMAPGAPD.DOC Project:

Even though the words on the first list are not reserved, they should not be
defined to be text macros or text macro functions. If they are, they will not be
recognized in their special contexts. The assembler does not give a warning if
such a redefinition occurs.
ABS LARGE NOTHING
ALL LISTING* NOTPUBLIC*
ASSUMES LJMP* OLDMACROS*
AT LOADDS* OLDSTRUCTS*
CASEMAP* M510* OS_DOS*
COMMON MEDIUM PARA
COMPACT MEMORY PRIVATE*
CPU* NEARSTACK* PROLOGUE*
DOTNAME* NODOTNAME* RADIX*
EMULATOR* NOEMULATOR* READONLY*
EPILOGUE* NOKEYWORD* REQ*
ERROR* NOLJMP* SCOPED*
EXPORT* NOM510* SETIF2*
EXPR16* NONE SMALL
EXPR32* NONUNIQUE* STACK
FARSTACK* NOOLDMACROS* TINY
FLAT NOOLDSTRUCTS* USE16
FORCEFRAME NOREADONLY* USE32
HUGE NOSCOPED* USES
LANGUAGE* NOSIGNEXTEND*
These operands are reserved words. Reserved words are not case sensitive.
$ DWORD PASCAL
? FAR QWORD
@B FAR16* REAL4*
@F FORTRAN REAL8*
ADDR* FWORD REAL10*
BASIC NEAR SBYTE*
BYTE NEAR16* SDWORD*
C OVERFLOW?* SIGN?*
CARRY?* PARITY?* STDCALL*

Appendix D MASM Reserved Words 409
SWORD* TBYTE WORD

SYSCALL* VARARG* ZERO?*
Special Operands for the 80386/486

FLAT* NEAR32* FAR32*
Predefined Symbols
Unlike most MASM reserved words, predefined symbols are case sensitive.
@CatStr* @Environ* @Model*
@code @fardata @SizeStr*
@CodeSize @fardata? @stack*
@Cpu @FileCur* @SubStr*
@CurSeg @FileName @Time*
@data @InStr* @Version
@DataSize @Interface* @WordSize
@Date* @Line*
Registers
AH CS DX SI
AL CX EAX SP
AX DH EBP SS
BH DI EBX ST
BL DL ECX TR3*
BP DR0 EDI TR4*
BX DR1 EDX TR5*
CH DR2 ES TR6
CL DR3 ESI TR7
CR0 DR6 ESP
CR2 DR7 FS
CR3 DS GS

Operators and Directives

.186 .ERRDIFI .TYPE
.286 .ERRE .UNTIL*
.286C .ERRIDN .UNTILCXZ*
.286P .ERRIDNI .WHILE*
.287 .ERRNB .XALL
.386 .ERRNDEF .XCREF
.386C .ERRNZ .XLIST
.386P .EXIT* ALIAS*
.387 .FARDATA ALIGN
.486* .FARDATA? ASSUME
.486P* .IF* CATSTR
.8086 .LALL COMM
.8087 .LFCOND COMMENT
.ALPHA .LIST DB
.BREAK* .LISTALL* DD
.CODE .LISTIF* DF
.CONST .LISTMACRO* DOSSEG
.CONTINUE* .LISTMACROALL* DQ
.CREF .MODEL DT
.DATA .NO87* DUP
.DATA? .NOCREF* DW
.DOSSEG* .NOLIST* ECHO*
.ELSE* .NOLISTIF* ELSE
.ELSEIF* .NOLISTMACRO* ELSEIF
.ENDIF* .RADIX ELSEIF1
.ENDW* .REPEAT* ELSEIF2
.ERR .SALL ELSEIFB
.ERR1 .SEQ ELSEIFDEF
.ERR2 .SFCOND ELSEIFDIF
.ERRB .STACK ELSEIFDIFI
.ERRDEF .STARTUP* ELSEIFE
.ERRDIF .TFCOND ELSEIFIDN

ELSEIFIDNI IFIDN POPCONTEXT*

ELSEIFNB IFIDNI PROC
ELSEIFNDEF IFNB PROTO*
END IFNDEF PTR
ENDIF INCLUDE PUBLIC
ENDM INCLUDELIB PURGE
ENDP INSTR PUSHCONTEXT*
ENDS INVOKE* RECORD
EQ IRP REPEAT*
EQU IRPC REPT
EVEN LABEL SEG
EXITM LE SEGMENT
EXTERN* LENGTH SHORT
EXTERNDEF* LENGTHOF* SIZE
EXTRN LOCAL SIZEOF*
FOR* LOW SIZESTR
FORC* LOWWORD* STRUC
GE LROFFSET* STRUCT*
GOTO* LT SUBSTR
GROUP MACRO SUBTITLE*
GT MASK SUBTTL
HIGH MOD TEXTEQU*
HIGHWORD* .MSFLOAT THIS
IF NAME TITLE
IF1 NE TYPE
IF2 OFFSET TYPEDEF*
IFB OPATTR* UNION*
IFDEF OPTION* WHILE*
IFDIF ORG WIDTH
IFDIFI %OUT
IFE PAGE

412
Processor Instructions
Processor instructions are not case sensitive.

8086/8088 Processor Instructions

AAA JAE LDS
AAD JB LEA
AAM JBE LES
AAS JC LODS
ADC JCXZ LODSB
ADD JE LODSW
AND JG LOOP
CALL JGE LOOPE
CBW JL LOOPEW*
CLC JLE LOOPNE
CLD JMP LOOPNEW*
CLI JNA LOOPNZ
CMC JNAE LOOPNZW*
CMP JNB LOOPW*
CMPS JNBE LOOPZ
CMPSB JNC LOOPZW*
CMPSW JNE MOV
CWD JNG MOVS
DAA JNGE MOVSB
DAS JNL MOVSW
DEC JNLE MUL
DIV JNO NEG
ESC JNP NOP
HLT JNS NOT
IDIV JNZ OR
IMUL JO OUT
IN JP POP
INC JPE POPF
INT JPO PUSH
INTO JS PUSHF
IRET JZ RCL
JA LAHF RCR

RET SCAS STOSB

RETF SCASB STOSW
RETN SCASW SUB
ROL SHL TEST
ROR SHR WAIT
SAHF STC XCHG
SAL STD XLAT
SAR STI XLATB
SBB STOS XOR
80186 Processor Instructions

BOUND INSW OUTSW
ENTER LEAVE POPA
INS OUTS PUSHA
INSB OUTSB PUSHW*

ARPL SIDT VERR
LAR SLDT VERW
LSL SMSW
SGDT STR
80286 and 80386 Privileged-Mode Instructions

CLTS LIDT LMSW
LGDT LLDT LTR

BSF CDQ IRETF*
BSR CMPSD JECXZ
BT CWDE LFS
BTC INSD LGS
BTR IRETD LODSD
BTS IRETDF* LOOPD*

LOOPED* SETAE SETNGE

LOOPNED* SETB SETNL
LOOPNZD* SETBE SETNLE
LOOPZD* SETC SETNO
LSS SETE SETNP
MOVSD SETG SETNS
MOVSX SETGE SETNZ
MOVZX SETL SETO
OUTSD SETLE SETP
POPAD SETNA SETPE
POPFD SETNAE SETPO
PUSHAD SETNB SETS
PUSHD* SETNBE SETZ
PUSHFD SETNC SHLD
SCASD SETNE SHRD
SETA SETNG STOSD

BSWAP* INVD* WBINVD*
CMPXCHG* INVLPG* XADD*
Instruction Prefixes
LOCK REPE REPNZ
REP REPNE REPZ
Coprocessor Instructions
Coprocessor instructions are not case sensitive.
8087 Coprocessor Instructions

F2XM1 FADDP FCHS
FABS FBLD FCLEX
FADD FBSTP FCOM

FCOMP FLDCW FRNDINT

FCOMPP FLDENV FRSTOR
FDECSTP FLDENVW* FRSTORW*
FDISI FLDL2E FSAVE
FDIV FLDL2T FSAVEW*
FDIVP FLDLG2 FSCALE
FDIVR FLDLN2 FSQRT
FDIVRP FLDPI FST
FENI FLDZ FSTCW
FFREE FMUL FSTENV
FIADD FMULP FSTENVW*
FICOM FNCLEX FSTP
FICOMP FNDISI FSTSW
FIDIV FNENI FSUB
FIDIVR FNINIT FSUBP
FILD FNOP FSUBR
FIMUL FNSAVE FSUBRP
FINCSTP FNSAVEW* FTST
FINIT FNSTCW FWAIT
FIST FNSTENV FXAM
FISTP FNSTENVW* FXCH
FISUB FNSTSW FXTRACT
FISUBR FPATAN FYL2X
FLD FPREM FYL2XP1
FLD1 FPTAN
80287 Privileged-Mode Instruction

FSETPM
80387 Instructions
FCOS FPREM1 FSINCOS
FLDENVD* FRSTORD* FSTENVD*
FNSAVED* FSAVED* FUCOM
FNSTENVD* FSIN FUCOMP

FUCOMPP

417
A P P E N D I X E
Default Segment Names
If you use simplified segment directives by themselves, you do not need to

know the names assigned for each segment. However, it is possible to mix full
segment definitions with simplified segment directives, in which case you need
to know the segment names.
Table E.1 shows the default segment names created by each directive.
If you use .MODEL, a _TEXT segment is always defined, even if all .CODE
directives specify a name. The default segment name used as part of far-code
segment names is the filename of the module. The default name associated with
the .CODE directive can be overridden, as can the default names for
.FARDATA and .FARDATA?.
The segment and group table at the end of listings always shows the actual
segment names. However, the GROUP and ASSUME statements generated by
the .MODEL directive are not shown in listing files. For a program that uses all
possible segments, group statements equivalent to the following would be
generated:
DGROUP GROUP _DATA, CONST, _BSS, STACK
For the tiny model, these ASSUME statements would be generated:

ASSUME cs:DGROUP, ds:DGROUP, ss:DGROUP
For small and compact models with NEARSTACK, these ASSUME statements
would be generated:
ASSUME cs: _TEXT, ds:DGROUP, ss:DGROUP
Filename: LMAPGAPE.DOC Project:

For medium, large, and huge models with NEARSTACK, these ASSUME
statements would be generated:
ASSUME cs:name_TEXT, ds:DGROUP, ss:DGROUP
Table E.1 Default Segments and Types for Standard Memory Models
Model Directive Name Align Combine Class Group
Tiny .CODE _TEXT WORD PUBLIC 'CODE' DGROUP
.FARDATA FAR_DATA PARA PRIVATE 'FAR_DATA'
.FARDATA? FAR_BSS PARA PRIVATE 'FAR_BSS'
.DATA _DATA WORD PUBLIC 'DATA' DGROUP
.CONST CONST WORD PUBLIC 'CONST' DGROUP
.DATA? _BSS WORD PUBLIC 'BSS' DGROUP
Small .CODE _TEXT WORD PUBLIC 'CODE'
.STACK STACK PARA STACK 'STACK' DGROUP *
Medium .CODE name_TEXT WORD PUBLIC 'CODE'
.STACK STACK PARA STACK 'STACK' DGROUP*
Compact .CODE _TEXT WORD PUBLIC 'CODE'

Appendix E Default Segment Names 419
Table E.1 (continued)

Model Directive Name Align Combine Class Group
Large or huge .CODE name_TEXT WORD PUBLIC 'CODE'
Flat .CODE _TEXT DWORD PUBLIC 'CODE'
.FARDATA _DATA DWORD PUBLIC 'DATA'
.FARDATA? _BSS DWORD PUBLIC 'BSS'
.DATA _DATA DWORD PUBLIC 'DATA'
.CONST CONST DWORD PUBLIC 'CONST'
.DATA? _BSS DWORD PUBLIC 'BSS'
.STACK STACK DWORD PUBLIC 'STACK'
* unless the stack type is FARSTACK


421
Glossary
8087, 80287, or 80387 coprocessor Intel chips that arg In PWB, a function modifier that introduces
perform high-speed floating-point and binary an argument or an editing function. The
coded decimal number processing. Also called argument may be of any type and is passed to
math coprocessors. Floating-point instructions are the next function as input. For example, the
supported directly by the 80486 processor. PWB command Arg textarg Copy passes
the text argument textarg to the function
Copy.
A
address The memory location of a data item or argument A value passed to a procedure or
procedure. The expression can represent just the function. See “parameter.”
offset (in which case the default segment is
assumed), or it can be in segment:offset format. array An ordered set of continuous elements of
the same type.
address constant In an assembly-language
instruction, an immediate operand derived by ASCII (American Standard Code for Information
applying the SEG or OFFSET operator to an Interchange) A widely used coding scheme
identifier. where 1-byte numeric values represent letters,
numbers, symbols, and special characters. There
address range A range of memory bounded by are 256 possible codes. The first 128 codes are
two addresses. standardized; the remaining 128 are special
characters defined by the computer
addressing modes The various ways a memory manufacturer.
address or device I/O address can be generated.
See “far address,” “near address.” assembler A program that converts a text file
containing mnemonically coded microprocessor
aggregate types Data types containing more instructions into the corresponding binary
than one element, such as arrays, structures, and machine code. MASM is an assembler. See
unions. “compiler.”
animate A debugging feature in which each line assembly language A programming language in
in a running program is highlighted as it executes. which each line of source code corresponds to a
The Animate command from the CodeView specific microprocessor instruction. Assembly
debugger Run menu turns on animation. language gives the programmer full access to the
API (application programming interface) A set of computer’s hardware and produces the most
system-level routines that can be used in an compact, fastest executing code. See “high-level
application program for tasks such as basic language.”
input/output and file management. In a graphics-
assembly mode The mode in which the
oriented operating environment like Microsoft CodeView debugger displays the assembly-
Windows, high-level support for video graphics language equivalent of the high-level code being
output is part of the Windows graphical API. executed. CodeView obtains the assembly-
Filename: LMAPGGLO.DOC Project: Glossary

Template: GLOSS.DOT Author: Ruth L Silverio Last Saved By: Ruth L Silverio
422 base address
language code by disassembling the executable breakpoint A user-defined condition that pauses
file. See “source mode.” program execution while debugging. CodeView
can set breakpoints at a specific line of code, for
a specific value of a variable, or for a
B combination of these two conditions.
base address The starting address of a stack
frame. Base addresses are usually stored in the buffer A reserved section of memory that holds
BP register. data temporarily, most often during input/output
operations.
base name The portion of the filename that
precedes the extension. For example, SAMPLE byte The smallest unit of measure for computer
is the base name of the file SAMPLE.ASM. memory and data storage. One byte consists of 8
bits and can store one 8-bit character (a letter,
BCD (binary coded decimal) A way of number, punctuation mark, or other symbol). It
representing decimal digits where 4 bits of 1 byte can represent unsigned values from 0 to 255 or
are a decimal digit, coded as the equivalent signed values between –128 and +127.
binary number.
binary Referring to the base-2 counting system, C

whose digits are 0 and 1. C calling convention The convention that follows
the C standard for calling a procedure—that is,
binary expression A Boolean expression pushing arguments onto the stack from right to
consisting of two operands joined by a binary left (in reverse order from the way they appear
operator and resolving to a binary number. in the argument list). The C calling convention
permits a variable number of arguments to be
binary file A file that contains numbers in binary
passed.
form (as opposed to ASCII characters
representing the same numbers). For example, a chaining (to an interrupt) Installing an interrupt
program file is a binary file. handler that shares control of an interrupt with
other handlers. Control passes from one handler
binary operator A Boolean operator that takes
to the next until a handler breaks the chain by
two arguments. The AND and OR operators in
terminating through an IRET instruction. See
assembly language are examples of binary
“interrupt handler,” “hooking (an interrupt).”
operators.
character string See “string.”
BIOS (Basic Input/Output System) The software
in a computer’s ROM which forms a hardware- clipboard In PWB, a section of memory that
independent interface between the CPU and its holds text deleted with the Copy, Ldelete, or
peripherals (for example, keyboard, disk drives, Sdelete functions. Any text attached to the
video display, I/O ports). clipboard deletes text already there. The Paste
function inserts text from the clipboard at the
bit Short for binary digit. The basic unit of current cursor position.
binary counting. Logically equivalent to decimal
digits, except that bits can have a value of 0 or 1, .COM The filename extension for executable
whereas decimal digits can range from 0 through files that have a single segment containing both
9. code and data. Tiny model produces .COM files.

device driver 423
combine type The segment-declaration specifier description file A text file used as input for the
(AT, COMMON, MEMORY, PUBLIC, or NMAKE utility.
STACK) which tells the linker to combine all
segments of the same type. Segments without a
combine type are private and are placed in
separate physical segments.
compact A memory model with multiple data

segments but only one code segment.
compiler A program that translates source code

into machine language. Usually applied only to
high-level languages such as Basic, FORTRAN,
or C. See “assembler.”
constant A value that does not change during

program execution. A variable, on the other
hand, is a value that can—and usually does—
change. See “symbolic constant.”
constant expression Any expression that

evaluates to a constant. It may include integer
constants, character constants, floating-point
constants, or other constant expressions.
D
debugger A utility program that allows the
programmer to execute a program one line at a
time and view the contents of registers and
memory in order to help locate the source of
bugs or other problems. Examples are CodeView
and Symdeb.
declaration A construct that associates the name

and the attributes of a variable, function, or type.
See “variable declaration.”
default A setting or value that is assumed unless

specified otherwise.
definition A construct that initializes and

allocates storage for a variable, or that specifies
either code labels or the name, formal
parameters, body, and return type of a
procedure. See “type definition.”

424 device driver
device driver A program that transforms I/O subroutines to be packaged, distributed, and
requests into the operations necessary to make a maintained independently of their callers.
specific piece of hardware fulfill that request. Windows extends the dynamic-link mechanism
to serve as the primary method by which all
Dialog Command window The window at the system and nonsystem services are obtained. See
bottom of the CodeView screen where dialog “linking.”
commands can be entered, and previously
entered dialog commands can be reviewed. dynamic-link library (DLL) A library file that
contains the executable code for a group of
direct memory operand In an assembly-language dynamically linked routines.
instruction, a memory operand that refers to the
contents of an explicitly specified memory dynamic-link routine A routine in a dynamic-link
location. library that can be linked at load time or run
time.
directive An instruction that controls the
assembler’s state.
E
displacement In an assembly-language element A single member variable of an array of
instruction, a constant value added to an like variables.
effective address. This value often specifies the
starting address of a variable, such as an array or environment block The section of memory
multidimensional table. containing the MS-DOS environment variables.
DLL See “dynamic-link library.” errorlevel code See “exit code.”
double-click To rapidly press and release a .EXE The filename extension for a program that
mouse button twice while pointing the mouse can be loaded and executed by the computer.
cursor at an object on the screen. The small, compact, medium, large, huge, and
flat models generate .EXE files. See “.COM,”
double precision A real (floating-point) value “tiny.”
that occupies 8 bytes of memory (MASM type
REAL8). Double-precision values are accurate exit code A code returned by a program to the
to 15 or 16 digits. operating system. This usually indicates whether
the program ran successfully.
doubleword A 4-byte word (MASM type
DWORD). expanded memory Increased memory available
after adding an EMS (Expanded Memory
drag To move the mouse while pointing at an Specification) board to an 8086 or 80286
object and holding down one of the mouse machine. Expanded memory can be simulated in
buttons. software. The EMS board can increase memory
from 1 megabyte to 8 megabytes by swapping
dump To display or print the contents of segments of high-end memory into lower
memory in a specified memory range. memory. Applications must be written to the
EMS standard in order to make use of expanded
dynamic linking The resolution of external memory. See “extended memory.”
references at load time or run time (rather than
link time). Dynamic linking allows the called expression Any valid combination of

forward declaration 425
mathematical or logical variables, constants, extended memory Physical memory above 1

strings, and operators that yields a single value. megabyte that can be addressed by 80286–
80486 machines in protected mode. Adding a
memory card adds extended memory. On
80386-based machines, extended memory can be
made to simulate expanded memory by using a
memory-management program.
extension The part of a filename (of up to three

characters) that follows the period (.). An
extension is not required but is usually added to
differentiate similar files. For example, the
source-code file MYPROG.ASM is assembled
into the object file MYPROG.OBJ, which is
linked to produce the executable file
MYPROG.EXE.
external variable A variable declared in one

module and referenced in another module.
F
far address A memory location specified with a
segment value plus an offset from the start of
that segment. Far addresses require 4 bytes—
two for the segment and two for the offset. See
“near address.”
field One of the components of a structure,

union, or record variable.
fixup The linking process that supplies addresses

for procedure calls and variable references.
flags register A register containing information

about the status of the CPU and the results of
the last arithmetic operation performed by the
CPU.
flat A nonsegmented linear address space.

Selectors in flat model can address the entire 4
gigabytes of addressable memory space. See
“segment,” “selector.”
formal parameters The variables that receive

values passed to a function when the function is
called.

426 forward declaration
forward declaration A function declaration that H

establishes the attributes of a symbol so that it
can be referenced before it is defined, or called handle An arbitrary value that an operating
from a different source file. system supplies to a program (or vice versa) so
that the program can access system resources,
frame The segment, group, or segment register files, peripherals, and so forth, in a controlled
that specifies the segment portion of an address. fashion.
handler See “interrupt handler.”

G
General-Protection (GP) fault An error that hexadecimal The base-16 numbering system
occurs in protected mode when a program whose digits are 0 through F (the letters A
accesses invalid memory locations or accesses through F represent the decimal numbers 10
valid locations in an invalid way (such as writing through 15). This is often used in computer
into ROM areas). programming because it is easily converted to
and from the binary (base-2) numbering system
gigabyte 1,024 megabytes, or 1,073,741,824 the computer itself uses.
bytes.
high-level language A programming language
global See “visibility.” that expresses operations as mathematical or
logical relationships, which the language’s
global constant A constant available throughout compiler then converts into machine code. This
a module. Symbolic constants defined in the contrasts with assembly language, in which the
module-level code are global constants. program is written directly as a sequence of
explicit microprocessor instructions. Basic, C,
global data segment A data segment that is COBOL, and FORTRAN are examples of high-
shared among all instances of a dynamic-link level languages. See “assembly language,”
routine; in other words, a single segment that is “compiler.”
accessible to all processes that call a particular
dynamic-link routine. hooking (an interrupt) Replacing an address in
the interrupt vector table with the address of
global variable A variable that is available another interrupt handler. See “interrupt
(visible) across multiple modules. handler,” “interrupt vector table,” “unhooking
(an interrupt).”
granularity The degree to which library
procedures can be linked as individual blocks of huge A memory model (similar to large model)
code. In Microsoft libraries, granularity is at the with more than one code segment and more than
object-file level. If a single object file containing one data segment. However, individual data
three procedures is added to a library, all three items can be larger than 64K, spanning more
procedures will be linked with the main program than one segment. See “large.”
even if only one of them is actually called.
group A collection of individually defined I

segments that have the same segment base identifier A name that identifies a register or
address. memory location.

linked list 427
IEEE format A standard created by the Institute interrupt handler A routine that receives
of Electrical and Electronics Engineers for processor control when a specific interrupt
representing floating-point numbers, performing occurs.
math with them, and handling
underflow/overflow conditions. The 8087 family interrupt service routine See “interrupt handler.”
of coprocessors and the emulator package
implement this format. interrupt vector An address that points to an
interrupt handler.
immediate expression An expression that
evaluates to a number that can be either a interrupt vector table A table maintained by the
component of an address or the entire address. operating system. It contains addresses (vectors)
of current interrupt handlers. When an interrupt
immediate operand In an assembly-language occurs, the CPU branches to the address in the
instruction, a constant operand that is specified at table that corresponds to the interrupt’s number.
assembly time and stored in the program file as See “interrupt handler.”
part of the instruction opcode.
import library A pseudo library that contains K

addresses rather than executable code. The linker keyword A word with a special, predefined
reads the addresses from an import library to meaning for the assembler. Keywords cannot be
resolve references to external dynamic-link used as identifiers.
library routines.
kilobyte (K) 1,024 bytes.
include file A text file with the .INC extension
whose contents are inserted into the source-code L
file and immediately assembled.
label A symbol (identifier) representing the
indirect memory operand In an assembly- address of a code label or data objects.
language instruction, a memory operand whose
language type The specifier that establishes the
value is treated as an address that points to the
naming and calling conventions for a procedure.
location of the desired data. See “pointer.”
These are BASIC, C, FORTRAN, PASCAL,
instruction The unit of binary information that a STDCALL, and SYSCALL.
CPU decodes and executes. In assembly
large A memory model with more than one
language, instruction refers to the mnemonic
code segment and more than one data segment,
(such as LDS or SHL) that the assembler
but with no individual data item larger than 64K
converts into machine code.
(a single segment). See “huge.”
instruction prefix See “prefix.”
library A file that contains modules of compiled
interrupt A signal to the processor to halt its code. MS-DOS programs use normal run-time
current operation and immediately transfer libraries, from which the linker extracts modules
control to an interrupt handler. Interrupts are and combines them with other object modules to
triggered either by hardware, as when the create executable program files. Windows-based
keyboard detects a keypress, or by software, as programs can use dynamic-link libraries (see),
when a program executes the INT instruction. which the operating system loads and links to
See “interrupt handler.” calling programs. See also “import library.”

428 linked list
linked list A data structure in which each entry macro A block of text or instructions that has
includes a pointer to the location of the adjoining been assigned an identifier. When the assembler
entries. sees this identifier in the source code, it
substitutes the related text or instructions and
linking In normal static linking, the process in assembles them.
which the linker resolves all external references
by searching run-time and user libraries, and then main module The module containing the point
computes absolute offset addresses for these where program execution begins (the program’s
references. Static linking results in a single entry point). See “module.”
executable file. In dynamic linking (see), the
operating system, rather than the linker, provides math coprocessor See “8087, 80287, or 80387
the addresses after loading the modules into coprocessor.”
separate parts of memory.
medium A memory model with multiple code
local constant A constant whose scope is limited segments but only one data segment.
to a procedure or a module.
megabyte 1,024 kilobytes or 1,048,576 bytes.
local variable A variable whose scope is confined
to a particular unit of code, such as module-level member One of the elements of a structure or
code, or a procedure. See “module-level code.” union; also called a field.
logical device A symbolic name for a device that memory address A number through which a
can be mapped to a physical (actual) device. program can reference a location in memory.
logical line A complete program statement in memory map A representation of where in

source code, including the initial line of code and memory the computer expects to find certain
any extension lines. types of information.
logical segment A memory area in which a memory model A convention for specifying the
program stores code, data, or stack information. number and types of code and data segments in
See “physical segment.” a module. See “tiny,” “small,” “medium,”
“compact,” “large,” “huge,” and “flat.”
low-level input and output routines Run-time
library routines that perform unbuffered, memory operand An operand that specifies a
unformatted input/output operations. memory location.
LSB (least-significant bit) The bit lowest in meta A prefix that modifies the subsequent
memory in a binary number. PWB function.
mnemonic A word, abbreviation, or acronym

M that replaces something too complex to
machine code The binary numbers that a remember or type easily. For example, ADC is
microprocessor interprets as program the mnemonic for the 8086’s add-with-carry
instructions. See “instruction.” instruction. The assembler converts it into
machine (binary) code, so it is not necessary to
remember or calculate the binary form.

overlay 429
module A discrete group of statements. Every the value 0.

program has at least one module (the main
module). In most cases, a module is the same as
a source file. O
.OBJ Default filename extension for an object
module-definition file A text file containing file.
information that the linker uses to create a
Windows-based program. object file A file (normally with the extension
.OBJ) produced by assembling source code. It
module-level code Program statements within contains relocatable machine code. The linker
any module that are outside procedure combines object files with run-time and library
definitions. code to create an executable file.
offset The number of bytes from the beginning
MSB (most-significant bit) The bit farthest to the
of a segment to a particular byte within that
left in a binary number. It represents 2(n-1) ,
segment.
where n is the number of bits in the number.
opcode The binary number that represents a
multitasking operating system An operating specific microprocessor instruction.
system in which two or more programs,
processes, or threads can execute operand A constant or variable value that is
simultaneously. manipulated in an expression or instruction.
operator One or more symbols that specify how
N the operand or operands of an expression are
naming convention The way the compiler or manipulated.
assembler alters the name of a routine before
option A variable that modifies the way a
placing it into an object file.
program performs. Options can appear on the
NAN Acronym for “not a number.” Math command line, or they can be part of an
coprocessors generate NANs when the result of initialization file (such as TOOLS.INI). An
an operation cannot be represented in IEEE option is sometimes called a switch.
format. For example, if two numbers being output screen The CodeView screen that
multiplied have a product larger than the displays program output. Choosing the Output
maximum value permitted, the coprocessor command from the View menu or pressing F4
returns a NAN instead of the product. switches to this screen.
near address A memory location specified by overflow An error that occurs when the value
the offset from the start of the value in a assigned to a numeric variable is larger than the
segment register. A near address requires only 2 allowable limit for that variable’s type.
bytes. See “far address.”
overlay A program component loaded into
nonreentrant See “reentrant procedure.” memory from disk only when needed. This
technique reduces the amount of free RAM
null character The ASCII character encoded as needed to run the program.
the value 0.
null pointer A pointer to nothing, expressed as

430 parameter
P passing by reference Transferring the address of

an argument to a procedure. This allows the
parameter The name given in a procedure procedure to modify the argument’s value.
definition to a variable that is passed to the
procedure. See “argument.” passing by value Transferring the value (rather
than the address) of an argument to a procedure.
This prevents the procedure from changing the
argument’s original value.
physical segment The true memory address of a
segment, referenced through a segment register.
pointer A variable containing the address of
another variable. See “indirect memory
operand.”
precedence The relative position of an operator
in the hierarchy that determines the order in
which expression elements are evaluated.
preemptive Having the power to take precedence
over another event.
prefix A keyword (LOCK, REP, REPE,
REPNE, REPNZ, or REPZ) that modifies the
behavior of an instruction. MASM 6.1 ensures
the prefix is compatible with the instruction.
private Data items and routines local to the
module in which they are defined. They cannot
be accessed outside that module. See “public.”
privilege level A hardware-supported feature of
the 80286–80486 processors that allows the
programmer to specify the exclusivity of a
program or process. Programs running at low-
numbered privilege levels can access data or
resources at higher-numbered privilege levels,
but the reverse is not true. This feature reduces
the possibility that malfunctioning code will
corrupt data or crash the operating system.
privileged mode The term applied to privilege
level 0. This privilege level should be used only
by a protected-mode operating system. Special
privileged instructions are enabled by .286P,
.386P, and .486P. Privileged mode should not be
confused with protected mode.

ROM (read-only memory) 431
procedure call An expression that invokes a from. RAM data is volatile; it is usually lost
procedure and passes actual arguments (if any) when the computer is turned off. Programs are
to the procedure. loaded into and executed from RAM. See
“ROM.”
procedure definition A definition that specifies a
procedure’s name, its formal parameters, the real mode The normal operating mode of the
declarations and statements that define what it 8086 family of processors. Addresses correspond
does, and (optionally) its return type and storage to physical (not mapped) memory locations, and
class. there is no mechanism to keep one application
from accessing or modifying the code or data of
procedure prototype A procedure declaration another. See “protected mode.”
that includes a list of the names and types of
formal parameters following the procedure name. record A MASM variable that consists of a
sequence of bit values.
process Generally, any executing program or
code unit. This term implies that the program or reentrant procedure A procedure that can be
unit is one of a group of processes executing safely interrupted during execution and restarted
independently. from its beginning in response to a call from a
preemptive process. After servicing the
Program Segment Prefix (PSP) A 256-byte data preemptive call, the procedure continues
structure at the base of the memory block execution at the point at which it was interrupted.
allocated to a transient program. It contains data
and addresses supplied by MS-DOS that a register operand In an assembly-language
program can read during execution. instruction, an operand that is stored in the
register specified by the instruction.
protected mode The 80286–80486 operating
mode that permits multiple processes to run and register window The optional CodeView window
not interfere with each other. This feature should in which the CPU registers and the flag register
not be confused with privileged mode. bits are displayed.
public Data items and procedures that can be registers Memory locations in the processor that
accessed outside the module in which they are temporarily store data, addresses, and processor
defined. See “private.” flags.
regular expression A text expression that
Q specifies a pattern of text to be matched (as
qualifiedtype A user-defined type consisting of opposed to matching specific characters).
an existing MASM type (intrinsic, structure,
union, or record), or a previously defined relocatable Not having an absolute address. The
TYPEDEF type, together with its language or assembler does not know where the label, data,
distance attributes. or code will be located in memory, so it
generates a fixup record. The linker provides the
address.
R
radix The base of a number system. The default return value The value returned by a function.
radix for MASM and CodeView is 10.
ROM (read-only memory) Computer memory that
RAM (random-access memory) Computer can only be read from and cannot be modified.
memory that can be both written to and read ROM data is permanent; it is not lost when the

432 routine
machine is turned off. A computer’s ROM often after they scroll off the top. This mode is
contains BIOS routines and parts of the required with computers that are not IBM
operating system. See “RAM.” compatible.
routine A generic term for a procedure or selector A value that indirectly references a
function. segment address. A protected-mode operating
system, such as Windows, assigns selector
run-time dynamic linking The act of establishing values to programs, which use them as segment
a link when a process is running. See “dynamic addresses. If a program attempts to use an
linking.” unassigned selector, it triggers a General-
run-time error A math or logic error that can be Protection fault (see).
detected only when the program runs. Examples shared memory A memory segment that can be
of run-time errors are dividing by a variable accessed simultaneously by more than one
whose value is zero or calling a DLL function process.
that doesn’t exist.
shell escape A method of gaining access to the
S operating system without leaving CodeView or
losing the current debugging context. It is
scope The range of statements over which a possible to execute MS-DOS commands, then
variable or constant can be referenced by name. return to the debugger.
See “global constant,” “global variable,” “local
constant,” “local variable.” sign extended The process of widening an
integer (for example, going from a byte to a
screen swapping A screen-exchange method that word, or a word to a doubleword) while retaining
uses buffers to store the debugging and output its correct value and sign.
screens. When you request the other screen, the
two buffers are exchanged. This method is signed integer An integer value that uses the
slower than flipping (the other screen-exchange most-significant bit to represent the value’s sign.
method), but it works with most adapters and If the bit is one, the number is negative; if zero,
most types of programs. the number is positive. See “two’s complement,”
“unsigned integer,” “MSB.”
scroll bars The bars that appear at the right side
and bottom of a window and some list boxes. single precision A real (floating-point) value that
Dragging the mouse on the scroll bars allows occupies 4 bytes of memory. Single-precision
scrolling through the contents of a window or values are accurate to six or seven decimal
text box. places.
segment A section of memory, limited to 64K single-tasking environment An environment in
with 16-bit segments or 4 gigabytes with 32-bit which only one program runs at a time. MS-
segments, containing code or data. Also refers to DOS is a single-tasking environment.
the starting address of that memory area.
small A memory model with only one code
sequential mode The mode in CodeView in segment and only one data segment.
which no windows are available. Input and
output scroll down the screen, and the old output source file A text file containing symbols that
scrolls off the top of the screen when the screen define the program.
is full. You cannot examine previous commands source mode The mode in which CodeView

status bar 433
displays the assembly-language source code that represents the machine code currently being
executed.
stack An area of memory in which data items
are consecutively stored and removed on a last-
in, first-out basis. A stack can be used to pass
parameters to procedures.
stack frame The portion of a stack containing a
particular procedure’s local variables and
parameters.
stack probe A short routine called on entry to a
function to verify that there is enough room in
the program stack to allocate local variables
required by the function.
stack switching Changing the stack pointers to
point to another stack area.
stack trace A symbolic representation of the
functions that are being executed to reach the
current instruction address. As a function is
executed, the function address and any function
arguments are pushed on the stack. Therefore,
tracing the stack shows the active functions and
their arguments.
standard error The device to which a program
can send error messages. The display is normally
standard error.
standard input The device from which a
program reads its input. The keyboard is
normally standard input.
standard output The device to which a program
can send its output. The display is normally
standard output.
statement A combination of labels, data
declarations, directives, or instructions that the
assembler can convert into machine code.
status bar See “linking.”

434 static linking
static linking The line at the bottom of the PWB text box In PWB, a box where you type
or CodeView screen. The status bar displays text information needed to carry out a command. A
position, keyboard status, current context of text box appears within a dialog box. The text
execution, and other program information. box may be blank or contain a default entry.
STDCALL A calling convention that uses caller tiny Memory model with a single segment for
stack cleanup if the VARARG keyword is both code and data. This limits the total program
specified. Otherwise the called routine must size to 64K. Tiny programs have the filename
clean up the stack. extension .COM.
string A contiguous sequence of characters toggle A function key or menu selection that
identified with a symbolic name. turns a feature off if it is on, or on if it is off.
Used as a verb, “toggle” means to reverse the
string literal A string of characters and escape status of a feature.
sequences delimited by single quotation marks
(' ') or double quotation marks (" "). TOOLS.INI A file containing initialization
information for many of the Microsoft utilities,
structure A set of variables that may be of including PWB.
different types, grouped under a single name.
two’s complement A form of base-2 notation in
structure member One of the elements of a which negative numbers are formed by inverting
structure. Also called a field. the bit values of the equivalent positive number
switch See “option.” and adding 1 to the result.
symbol A name that identifies a memory type A description of a set of values and a valid
location (usually for data). set of operations on items of that type. For
example, a variable of type BYTE can have any
symbolic constant A constant represented by a of a set of integer values within the range
symbol rather than the constant itself. Symbolic specified for the type on a particular machine.
constants are defined with EQU statements.
They make a program easier to read and modify. type checking An operation in which the
assembler verifies that the operands of an
SYSCALL A language type for a procedure. Its operator are valid or that the actual arguments in
conventions are identical to C’s, except no a function call are of the same types as the
underscore is prefixed to the name. function definition’s parameters.
type definition The storage format and attributes
T for a data unit.
tag The name assigned to a structure, union, or
enumeration type.
U
task See “process.” unary expression An expression consisting of a
single operand preceded or followed by a unary
text Ordinary, readable characters, including the operator.
uppercase and lowercase letters of the alphabet,
the numerals 0 through 9, and punctuation unary operator An operator that acts on a single
marks. operand, such as NOT.

word 435
underflow An error condition that occurs when a visibility The characteristic of a variable or
calculation produces a result too small for the function that describes the parts of the program
computer to represent. in which it can be accessed. An item has global
visibility if it can be referenced in every source
unhooking (an interrupt) The act of removing file constituting the program. Otherwise, it has
your interrupt handler and restoring the original local visibility.
vector. See “hooking (an interrupt).”
union A set of values (in fields) of different W
types that occupy the same storage space. watch window The window in CodeView that
unresolved external See “unresolved reference.” displays watch statements and their values. A
variable or expression is watchable only while
unresolved reference A reference to a global or execution is occurring in the section of the
external variable or function that cannot be program (context) in which the item is defined.
found, either in the modules being linked or in
the libraries linked with those modules. An window A discrete area of the screen in PWB or
unresolved reference causes a fatal link error. CodeView used to display part of a file or to
enter statements.
unsigned integer An integer in which the most-
significant bit serves as part of the number, window commands Commands that work only in
rather than as an indication of sign. For example, CodeView’s window mode. Window commands
an unsigned byte integer can have a value from 0 consist of function keys, mouse selections, CTRL
to 255. A signed byte integer, which reserves its and ALT key combinations, and selections from
eighth bit for the sign, can range from -127 to pop-up menus.
+128. See “signed integer,” “MSB.” window mode The mode in which CodeView
user-defined type A data type defined by the displays separate windows, which can change
user. It is usually a structure, union, record, or independently. CodeView has mouse support
pointer. and a wide variety of window commands in
window mode.
V word A data unit containing 16 bits (2 bytes). It
variable declaration A statement that initializes can store values from 0 to 65,535 (or -32,768 to
and allocates storage for a variable of a given +32,767).
type.
virtual disk A portion of the computer’s random
access memory reserved for use as a simulated
disk drive. Also called an electronic disk or RAM
disk. Unless saved to a physical disk, the
contents of a virtual disk are lost when the
computer is turned off.
virtual memory Memory space allocated on a
disk, rather than in RAM. Virtual memory allows
large data structures that would not fit in
conventional memory, at the expense of slow
access.

436

435
Index
! (literal-character operator) 235 || (logical OR operator) 178
!= (not equal operator) 178 32-bit programming 335
” (double quotation marks) 109, 353 80186 processor 3
$ (current address operator) 368 80188 processor 3
% (expansion operator) 235, 248, 357 80286 processor 3
& (substitution operator) 238, 372 80287 math coprocessor 3, 135
&& (logical AND operator) 178 80386 processor 3, 335
’ (single quotation mark) 109, 353 80387 math coprocessor 3, 135
( ) (parentheses) 106 80486 processor 3, 135
+ (plus operator) 63, 66, 352, 370 8086-based processors 2– 3
. (dot operator) 126 8087 math coprocessor 3, 135
. (structure-member operator) 64, 67, 352, 370 8088 processor 3
.186 directive 38
.286 directive 38
.286P directive 38 A
.287 directive 38 AAD instruction 160
.386 directive AAM instruction 160
FLAT, with 26, 36 ABS operand 220
processor mode, specifying 38, 336 Accessing data with pointers See Pointer variables
segment mode, setting 46, 68 ADC instruction 92– 94
.386P directive 38 ADD instruction 92– 94
.387 directive 38 ADDR operator 197
.486 directive Addresses
FLAT, with 36 displacement of 65
processor mode, specifying 38 dynamic 79
segment mode, setting 46, 68 effective 65
.486P directive 38 errors in 54
.8087 directive 38 far 57, 74, 80
: (colon) 22, 352, 354 near 57, 80
: (segment-override operator) 50, 59– 60, 64 physical 7
:: (double colon) 197, 215, 352– 354 registers, loading into 80
; (semicolon) 21 relocatable 57
;; (double semicolon) 227 segmented 7– 8, 53
< (less than operator) 178 Addressing
< > (angle brackets) See Angle brackets direct registers, used in 62– 63
= = (equal operator) 178 indirect registers, used in 65, 68
> (greater than operator) 178 scaling operands 70
? (question mark initializer) specifying 60
array elements 109 Aliases 87, 369
described 368 ALIGN directive 3
variables 87 Align types 45
@ (at sign) 10 See also individual entries
@@: (anonymous label) 170 .ALPHA directive 47
[ ] (brackets) 107 AND instruction 27, 99, 100
[ ] (index operator) 63 Angle brackets (< >)
\ (backslash character), MASM code 22 default parameters 230
\ (line-continuation character) 121 epilogues 202
{} (curly braces) 121, 131 FOR loops 242
Filename: LMAPGINX.DOC Project:

Template: INDEX.DOT Author: Samuel G. Dawson Last Saved By: Ruth L Silverio
436 Index
FORC loops 244

Index 437
Angle brackets (< >) (continued)

macro text delimiters 234
prologues 202
records 131
structures and unions 121
Anonymous label (@@) 170
API (Application Programming Interface) 257
Architecture, segmented 2, 5
Architecture, unsegmented 5
Arguments
errors 196
macro 252
mixed-language programs, passing in 314
qualifiedtypes, with 16
stack, on 182
Arrays
accessing elements in 105
declaring 105
defined 105
defining 15
DUP, declaring with 106, 124
instructions for processing 110
length of 108
multiple-line declarations for 105
number of bytes in 108
referencing 108, 316
size of elements 108
with DUP operator See DUP operator
with SIZEOF operator See SIZEOF operator
with TYPE operator See TYPE operator
ASCIIZ 267
Assembly
actions during 23
conditional See Conditional assembly
INCLUDE files 212
language
book list xviii
mixed-language programs 312
listing files See Listing files
two-pass 358
Assembly pointers See Conditional assembly
Assembly-time variables 233
ASSUME directive
.MODEL, generated with 37
code segments, changing 357
enhancements 344
general-purpose registers 77
segment registers, setting 49– 55, 58– 59, 357
AT address combine type 46
/AT command-line option, ML 36
At sign (@) 10

438 Index
B OPTION directive 25
Backslash character (\) 22

Backus-Naur Form See BNF grammar
Base Pointer (BP) register 73
Basic calling conventions 308– 310
Basic/MASM programs 328– 332
Binary Coded Decimals
calculating with 156– 160
defining 156
instructions for 156– 160
packed 158
unpacked 159– 160
Bits
mask 99– 102
rotating 100
shifting 100
BNF grammar 16, 379– 380
BOUND instruction 108, 204
BP (Base Pointer) register 73
Brackets ([ ]) 107
.BREAK directive 173, 176
BSF instruction 100
BSR instruction 100
BYTE
align type 45
directive 86
C
C calling convention 309
C++/MASM programs 322– 323
C/MASM programs 315– 321
CALL instruction 180
Calling conventions 309
Basic 308– 310
directives, specifying 37
FORTRAN 308– 310
(list) 308
mixed-language programming 308– 309
Pascal 310
STDCALL 311
SYSCALL 308– 311
CARRY? flag as operand 178
Case sensitivity
enforcing 348
macro functions, predefined 245
MASM statements 22
radix specifiers 11
reserved words 9, 407
specifying
command-line options, in 25
language type 348

Index 439
Case sensitivity (continued) Constants

symbols, predefined 10 defined 11
CASEMAP expressions 12
ALL argument, OPTION directive 25 immediate 61
NONE argument, OPTION directive 25 integer 11– 12
NOTPUBLIC argument, OPTION directive 25 size 363
CATSTR directive 245– 247 size of 12
CATSTR, compared with TEXTEQU directive See symbolic 12
TEXTEQU directive .CONST directive 33, 39– 40
@CatStr predefined string function 245– 247 .CONTINUE directive 173, 176
CBW instruction 90 Coprocessors
CDQ instruction 90 architecture 140– 144
CLC instruction 104 control registers 156
Cleaning the stack 185 data format in registers 140
CLI instruction 5, 209 defined 135
Client program 257, 266 described 3, 139
CMC instruction 104 instructions
CMP instruction 166 arithmetic 148– 150
CMPS instruction 110– 114, 353 data transfer 146
CMPSB instruction 114 described 146
.CODE directive 33, 40– 42 (list) 414
Code segment See Segments, code overview 141
Code, near or far 57 program control 151– 155
@CodeSize predefined symbol 40 memory access 145
CodeView for Windows 264 operand formats
Combine types classical stack 141
(list) 46 memory 142
See also individual entries overview 141
.COM files register 143
relocatable segment expression, lacking 62 register-pop 144
starting address 56 specifying 37, 140
tiny model, using 36, 46– 47 status word register 156
COMM directive 16, 211, 217– 218 steps for using 145
Command-line driver, ML xvi /Cp command-line option, ML 10, 245
Command-line options See ML command-line options @Cpu predefined symbol 254
COMMENT directive 22 Curly braces ({})
Comments records 131
extended lines, in 346 structures and unions 121
macros, in 227 Current address operator ($) 368
source code 21– 22 @CurSeg predefined symbol 39, 219
COMMON combine type 46 CWD instruction 90
Communal variables 217 CWDE instruction 90
Compact model See Memory models, compact /Cx command-line option, ML 158
Compatibility, MASM 5.1 See MASM 5.1 compatibility
Conditional assembly
assembly behavior, changing 23 D
conditions, testing for 28 DAA instruction 162
directives 28 DAS instruction 162
pointers 83, 187 .DATA directive 33, 39– 40
Conditional-error directives (table) 29 .DATA? directive 33, 39– 40
Conditional jumps 164– 170 @data predefined symbol 39
Conditions, testing for conditional assembly See Conditional Data segment See Segments, data
assembly @DataSize predefined symbol 39, 83

440 Index

Index 441
Data types
arrays See Arrays
attributes for 15
Binary Coded Decimals 159
defined 14
defining 87
directives 14
floating-point 136
initializers, as 14
integers, allocating memory for 85– 86
new features, MASM 6.1 344
qualifiedtypes 15, 214
real 14, 136
signed 14, 86
strings See Strings
structures 117
unions 117
user-defined 15
Data, near or far 57, 58
Data-sharing methods 211
Data-sharing methods, multiple-module programs See
Multiple-module programs
Date, system 11
DB directive 86
DD directive 86
DEC instruction 92– 94
DF directive 86
DGROUP group name
.MODEL, defined by 34, 39, 51
DS registers, initializing to 56
MS-DOS programs, for 41– 42
near data, accessing 57– 58
segment 35– 37, 51– 52, 57
Direct memory operands
loading offset of 82
overview 60– 64
Directives
.286P 38
.287 38
.386 See .386 directive
.386P 38
.387 38
.486P 38
.8087 38
.ALPHA 47
ALIGN 3
.BREAK 173, 176
BYTE 86
CATSTR 245– 247
.CODE 33, 40– 42
COMM 16, 211, 217– 218
COMMENT 22
Conditional assembly 28
Conditional error 29, 358

442 Index
Directives (continued)
.CONST 33, 39– 40
.CONTINUE 173, 176
.DATA 33, 39– 40
.DATA? 33, 39– 40
Data declarations, for 87
Data types, for 14
Data-sharing See EXTERN directive
DB 86
DD 86, 136
Decision 171
DF 86
.DOSSEG 47
DQ 86, 136
DT 86, 136
DW 86
DWORD 86
ECHO 236
.ELSE 171
ELSE 28
.ELSEIF 171
ELSEIF 28
ELSEIF1 29, 358
ELSEIF2 29, 358
END 33, 56
.ENDIF 171
ENDIF 28
ENDM 227– 239
ENDP 180– 181, 206
ENDS 44
.ENDW 173
EQU 12, 369
.ERR 30
.ERR1 30, 358
.ERR2 30, 358
.ERRB 30, 231
.ERRDEF 30
.ERRDIF 30
.ERRE 30
.ERRIDN 29
.ERRNB 29, 231
.ERRNDEF 29
.ERRNZ 29
EVEN 3
.EXIT 33, 41– 43
EXITM 248
EXTERN See EXTERN directive
EXTERNDEF See EXTERNDEF directive
FARDATA 33, 39– 40
.FARDATA 39– 40
.FARDATA? 33, 39– 40
Floating-point 136
FOR 242– 243, 249
FORC 244

Index 443
Directives (continued)
FWORD 86
GROUP 51– 52
.IF 171
IF 28– 29
IF1 29, 358
IF2 29, 358
IFB 29, 231
IFDEF 29, 359
IFDIF 29
IFE 29
IFIDN 29
IFNB 29, 231
IFNDEF 29, 359
INCLUDE 212
INCLUDELIB 222
INSTR 245– 246
INVOKE See INVOKE directive
LABEL 16
LOCAL 188– 191, 232
loop-generating 173
.MODEL See .MODEL directive
.MSFLOAT 361
Naming conventions 37
.NO87 38, 349
obsolete 361
OPTION See OPTION directive
ORG 56
POPCONTEXT 255, 349
PROC 180– 184, 193, 206, 312
PUBLIC 185, 211, 220
PUSHCONTEXT 255, 349
QWORD 86
.RADIX 11
REAL4 136– 137
REAL8 136– 137
REAL10 136– 137
RECORD 130– 131
Renamed since MASM 5.1 350
.REPEAT 173– 177
REPEAT 240
SBYTE 86
SDWORD 86
SEGMENT 44– 47
Segment order, controlling 47
.SEQ 47
SIZESTR 245– 246
STACK See STACK directive
.STARTUP See .STARTUP directive
STARTUP See .STARTUP directive
STRUCT 118– 129
SUBSTR 245– 246
SWORD 86
TBYTE 86, 159

444 Index
Directives (continued) ELSE directive 28

TEXTEQU See TEXTEQU directive
UNION 118– 119, 122, 125– 129
.UNTIL 173
.UNTILCXZ 173
.WHILE 173– 177
WHILE 241
WORD 86
Directives: 36– 38, 46
Displacement 66
Distance attributes 15
DIV instruction 97– 98
Division 97, 102
DLLs
client program 257, 266
data segment 265– 269
defined 257, 266
example 267– 268
extension name 266
heap 261– 262, 265– 267
IMPLIB utility 258
initialization 261– 262, 268– 269
loading 258– 260
programming requirements 260– 261, 267
prologue and epilogue 264– 267
stacks in 46, 264– 267
summary 266
termination 262– 264, 270
Document conventions vii
DOS See MS-DOS
.DOSSEG directive 47
Dot (.) operator See Structure-member operator
DOTNAME argument, OPTION directive 25
Double colon (::) 197, 215
Double quotation marks (”) 109
Double semicolon (;;) 227
Doublewords 86
DQ directive 86
DT directive 86
DUP operator
arrays, with 106, 124
record variables, with 131
structures and unions, with 121
DW directive 86
DWORD
align type 45
directive 86
Dynamic -link libraries See DLLs
E
ECHO directive 236
.ELSE directive 171

Index 445
.ELSEIF directive 171

ELSEIF directive 28
ELSEIF1 directive 358
ELSEIF2 directive 29, 358
EMULATOR argument, OPTION directive 27, 157
Emulator libraries 155– 156
END directive 33, 56
.ENDIF directive 171
ENDIF directive 28
ENDM directive 227– 239
ENDP directive 180– 181, 206
ENDS directive 44
.ENDW directive 173
ENTER instruction 183
Environment
target 4
variables
INCLUDE 213
LIB 222
returning values of 10
/EP command-line option, ML 342
EPILOGUE argument, OPTION directive 26, 201– 203
Epilogue code
defined 198
macros 201– 202, 264– 265
PROC statement, specifying arguments in 185
procedures, with 26
RET instruction 357
standard 199
user-defined 201
EQ operator 365
EQU directive 12, 369
Equal directive (=) 12
Equates, predefined See Predefined symbols
.ERR directive 29
.ERR1 directive 30, 358
.ERR2 directive 30, 358
.ERRB directive 29, 231
.ERRDEF directive 29
.ERRDIF directive 29
.ERRE directive 29
.ERRIDN directive 29
.ERRNB directive 29, 231
.ERRNDEF directive 29
.ERRNZ directive 29
Error detection 196
ERROR operand 49–50
Errors, argument passing 196
ESC instruction 360
EVEN directive 3
Executable (.EXE) files, controlling size of 223
Exit codes, Windows operating system 263
.EXIT directive 33, 41– 43
EXITM directive 248

446 Index
Expansion operator (%) 235– 236, 248, 357 relocatable segment expression, lacking 62
Explicit loading 258
Exponent bias 139
EXPORT operand 185
EXPORTS statement 261, 270
EXPR16 argument, OPTION directive 13, 26, 361, 373
EXPR32 argument, OPTION directive 13, 26, 373
Expression operators 178
Expressions
assembly-time evaluation 23
constant 12
loop conditions, evaluating 179
OPTION M510 behavior 364, 373
order of evaluation 14
size 366, 373
word size 13, 26
Extension, filename 266
EXTERN directive
data-sharing 211
executable file size, limiting 223
module-specific 220
overview 16
positioning 218
procedure prototypes, declaring 193
External declarations 216– 218
External variables 217, 369
EXTERNDEF directive
data-sharing 211
overview 16
positioning 218
procedure prototypes, declaring 193
symbols, declaring 214– 215
F
Far addresses, invoking 57, 74, 80– 81, 197
Far code 57
Far data 58– 60
.FARDATA directive 33, 39– 40
.FARDATA? directive 39– 40
FAR operator 169, 185
Far pointer 74, 80– 81
FARSTACK operand
example 35
grouping 34
in Windows-based programs 266
MS-DOS program, initializing 43
special cases, setting for 37
Farwords 86
FCOM instruction 153
Fields, statements in 21– 22
Files
.COM

Index 447
Files (continued)
.COM (continued)
starting address 56
tiny model, using 36, 46– 47
executable 24
include 212– 213, 348
line numbers 11
naming 11
Flags
CARRY? 178
operands, as 178
OVERFLOW? 178
PARITY? 178
SIGN? 178
stack, saving on 73
ZERO? 178
Flags register See Registers, flags
Flat model See Memory models, flat
FLAT operand 46, 49– 50
FLD1 instruction 147
FLDZ instruction 147
Floating-point
calculations 3
constants
decimal form 137
encoded hexadecimal format 137
syntax for defining 136
emulation 157– 158
IEEE format 139
instructions
arithmetic 148– 149
controlling 26
data transfer 147
not emulated (list) 158
program control 152– 153, 156
operations 146
values
double precision 139
single precision 139
variables
IEEE format 138
Microsoft binary format 138
.MSFLOAT format 138
ranges 136
FOR directive 242– 243, 249
FORC directive 244
FORCEFRAME operand 200– 201
FORTRAN calling convention 308– 310
FORTRAN/MASM programs 323– 326
/Fpi command-line option, ML 26, 157
Frame 62
FS register 17
FTST instruction 153

448 Index
Full segment definitions

described 32
segment registers, initializing 54– 56
using 44– 51
Full segment defintions See .STARTUP directive
FWORD directive 86
FXCH instruction 144
G
Global variables 211
GROUP directive 51– 52
Groups
defined 51
DGROUP 51
SEG operator, returned by 62
GS register 17
H
H2INC 318
Heap space 261– 262, 265– 267
HEAPSIZE statement 261, 271
Help, online See Microsoft Advisor
HIGH operator 356
HIGHWORD operator 346
Huge model See Memory models, huge
I
/I command-line option, ML 213
Identifiers
ABS, using 220
naming restrictions 9, 346, 353, 357, 368
OPTION DOTNAME 373
OPTION NOKEYWORD 376
IDIV instruction 97– 98
IEEE format 139
.IF directive 171
IF directive 28– 29
IF1 directive 29, 358
IF2 directive 29, 358
IFB directive 29, 231
IFDEF directive 29, 359
IFDIF directive 29
IFE directive 29
IFIDN directive 29
IFNB directive 29, 231
IFNDEF directive 29, 359
Immediate operands 60– 62
IMPLIB utility 258
Implicit loading 258
Import libraries 258

Index 449
IMPORTS statement 266 Instructions (continued)

IMUL instruction 95– 96 FLDZ 147
IN instruction 5 floating-point See Floating-point instructions
INC instruction 92– 94 FTST 153
INCLUDE directive 212 FXCH 144
INCLUDE environment variable 213 IDIV 97– 98
Include files IMUL 95– 96
assembling 213 IN 5
nested 213 INC 92– 94
overview 212, 348 INT 204– 205
INCLUDELIB directive 222 INTO 207
Index operator ([ ]) 63 JCXZ 170– 173
Indirect memory operands 60, 64– 70 JECXZ 170– 173
Indirect procedure calls See INVOKE directive JMP 49, 162
Initializers JO 165
allocating 87 jump 165–167, 170, 173
directives for 15 LAHF 73
multiple-line 346 LDS 81
Instance 261, 266 LEA 82, 104
INSTR directive 245– 246 LEAVE 183
@InStr predefined string function 245– 246 LES 81
Instruction Pointer (IP) register 20, 57, 161 (list) 412
Instructions LOCK 353, 363
ADC 92– 94 LODS 110– 115, 353
ADD 92– 94 logical 99– 102
AND 26, 99– 100 LOOP 172
arithmetic 378 LOOPE 172
bit-test 354 LOOPNE 172
BOUND 108, 204 LOOPNZ 172
BSF 100 LOOPZ 172
BSR 100 MOV 49, 82, 89
CALL 180 MOVS 110– 113, 353
CBW 90 MOVSX 92
CDQ 90 MOVZX 92
CLC 104 MUL 95– 96
CLI 5, 209 NOP 377
CMC 104 NOT 99– 100
CMP 166 obsolete 360
CMPS 110– 114, 353 operands for 60
CMPSB 114 OR 26, 99– 100, 168
conditional-jump 165– 167 OUT 5
coprocessor 377 POP 49, 71
CWD 90 POPA 74
CWDE 90 POPAD 74
DAA 162 POPF 73
DAS 162 POPFD 73
DEC 92– 94 privileged 2, 38
default segments, requiring 49 PUSH 49, 71
DIV 97– 98 PUSHA 74
encodings, changes to 377– 378 PUSHAD 74
ENTER 183 PUSHF 73
ESC 360 PUSHFD 73
FCOM 153 RCL 101– 104
FLD1 147 RCR 101– 104

450 Index
Instructions (continued)
REP 110– 112, 363
REPE 110– 112, 363
REPNE 110– 112, 353, 363
REPNZ 110– 112, 353, 363
REPZ 110– 112, 363
RET 378
RETF 181, 378
RETN 181, 378
ROL 101– 104
ROR 101– 104
SAL 101– 104
SAR 101– 104
SBB 92– 94
SCAS 110– 115, 353
SHL 101– 104
SHR 101– 104
STC 104
STI 5, 209
STOS 110– 113, 353
SUB 92– 94
TEST 167– 168
timing xvii, 399– 400
XCHG 90
XLAT 116
XLATB 116
XOR 26, 99– 100
Integers
adding 92– 94
allocating memory for 85– 86
Binary Coded Decimal (BCD) 159
bit operations on 99
constants, defining 11– 12
dividing 97– 98
exchanging 90
hexadecimal 12
initializing 87
memory format 86
moving 89
multiplying 95– 96
operations with 88
popping off stack 71
pushing onto stack 71
radix specifiers for 11
sign-extending 90
signed 86
size of 86
stack 71
subtracting 92– 94
translating 116
types, defining 14, 86
value range 86
@Interface predefined symbol 37
Interrupt vector 205

Index 451
Interrupt-enable flag 205

Interrupts
L
CLI instruction 209 LABEL directive 16
handlers 206– 207
INT instruction 204– 205
MS-DOS 204, 285
operation 206
overview 204
redefining 207
STI instruction 209
vector table 205
INTO instruction 207
INVOKE directive
actions 194
ADDR, invoking 197
arguments, widening 196
error detection 196
far addresses, invoking 197
generated code, checking 198
indirect procedure calls 198
mixed-language programs 312– 313
procedures, calling 193– 197, 216
type conversions 194– 195
J
JCXZ instruction 170– 173
JECXZ instruction 170– 173
JMP instruction 49, 162
JO instruction 165
Jumps
anonymous 170
automatic 169
conditional
bit status 167
comparisons 166
extending 26, 169
flag status 165– 166
instructions (list) 165– 167
overview 164
zero value 168
directives for 171
extension, automatic 26, 169
instructions 165– 167
optimization, automatic 162
overview 161
unconditional
indirect operands 163
jump tables 163
overview 162

452 Index
Labels
anonymous 170
code
length 346
OPTION M510 behavior 363
OPTION NOSCOPED 375
procedures, in 357
referencing 352
size 346
visibility 354
LAHF instruction 73
LANGUAGE
BASIC argument, OPTION directive 26
C argument, OPTION directive 26
FORTRAN argument, OPTION directive 26
PASCAL argument, OPTION directive 26
STDCALL argument, OPTION directive 26
SYSCALL argument, OPTION directive 26
LANGUAGE argument, OPTION directive 193
Language attributes
.MODEL directive, with 34, 37
OPATTR operator 253
OPTION directive, with 26
Large model See Memory models, large
LDS instruction 81
LEA instruction 82, 104
LEAVE instruction 183
Length of strings See LENGTHOF operator
LENGTH operator 356– 357, 364
LENGTHOF operator
number of items, returning 110, 124, 132, 346
structures, defining 108
unions, with 125
LES instruction 81
Libraries
C run-time 271
emulator 155– 156
overview 221
source files, specifying in 222
LIBRARY statement 270
Line-continuation character (\) 121
LINK, command-line options See individual entries
Linkage specification 322– 323
Linking
actions during 24, 45
segment order in 48
Listing files
code generated 399
command-line options 397– 399
error messages 400
examples 401
generating 397
PWB options 397– 399
reading 399, 405

Index 453
Listing files (continued)

symbols used in (list) 400
tables in 405– 406
Literal-character operator (!) 235
LJMP argument, OPTION directive 27
LOADDS operand 200– 201
Loading local address variables See Local variables
Loading, actions during 24
Local addresses, loading See Local variables
LOCAL directive 188– 191, 232
Local variables
creating 188
loading addresses of 82
procedures, in 188
LOCK instruction 353, 363
LODS instruction 110– 115, 353
Logical AND 178
Logical instruction 99– 100
Logical line 22
Lookup tables 241
LOOP instruction 172
LOOPE instruction 172
LOOPNE instruction 172
LOOPNZ instruction 172
Loops
conditions
expression evaluation 179
precedence 179
PTR operator in 178
relational operators for (list) 178
signed operands 178
writing 178
controlling 176
directives
.REPEAT 173– 177
.WHILE 173– 177
instructions (list) 172
macros
FOR 242– 243, 249
FORC 244
REPEAT 240
WHILE 241
LOOPZ instruction 172
LOW operator 356
LOWWORD operator 346, 366
LROFFSET operator 344
M
M510 argument, OPTION directive
compatibility with MASM 5.1 26, 353– 370
expression word size, setting 13
structures, with 119

454 Index
Macros
arguments
commas 352, 372
quotation marks 353
testing 29, 252
variable 242, 249
calling 227
checking argument types with 253
comments (;;) 227
expansion 23
functions
defined 248
epilogues 201
EXITM 248
prologues 201
returning values 248
local symbols in 232
loops
FOR 242– 243, 249
FORC 244
REPEAT 240
WHILE 242– 243
MASM 5.1 behavior 25, 356, 372
nested 251
new features 351
operators
behavior in macro functions 251
expansion (%) 235– 236, 248, 357
(list) 234
literal-character (!) 235
substitution (&) 238, 352, 372
OPTION OLDMACROS 372
parameters
default values 230
procedure parameters, compared to 234
required 229
substitution 238
passing arguments to 228, 235
predefined string functions 11
procedures
defined 226
functions, compared to 228
recursive 255
redefining 251
text
defined 226
forward referencing 356
numeric equates, compared to 234
syntax 226
VARARG keyword 242, 249, 351
writing 227
Mask
defined 99

Index 455
Mask (continued)
logic instructions, with 102
record operators, with 133
MASK operator 133
MASM 5.1 compatibility
address fixups 26
macro behavior 25, 356, 372
OPTION directive, specifying 25
overview xvi
structures 25
updating code 353– 360
MASM utility xvi, 342
Math coprocessor See Coprocessors
Medium model See Memory models, medium
Memory
access 64
allocation 24
virtual 5
MEMORY combine type 46
Memory models
attributes 35
compact 36
described 34
determining 10
far code segments 40
far data segments 40
flat 36, 58, 336
huge 36
large 36
medium 36
model-independent code 83
near code segments 40
small 36
specifying in PROC statement 185
tiny 36, 46– 47
Memory-resident programs See TSRs
Microsoft Advisor xiii, 342
Minus operator (– ) 64
Mixed-language programming
argument passing 314
assembly procedures 312
Basic/MASM programs 328– 332
C prototypes, converting with H2INC 318
C++/MASM programs 322– 323
C/MASM programs 315– 321
calling conventions
Basic 308– 310
FORTRAN 308– 310
(list) 308
Pascal 310
STDCALL 311
SYSCALL 308– 311
column-major order 315

456 Index
Mixed-language programming (continued)

compatible data types
Basic (list) 328
C (list) 315
FORTRAN (list) 323
external data 314
FORTRAN/MASM programs 323– 326
initialization code 313, 321
INVOKE, using 312– 313
naming conventions 308– 309
overview 307
register preservation 314
row-major order 315
ML command-line options
/AT 36
/Cp 10, 245
described xvi
/EP 342
/Fpi 26, 157
/I 213
listing options (list) 397
overview xvi
/X 213
/Zm 62, 119
/Zp 119
Mode, real, protected See Real mode; Protected mode
.MODEL directive
attributes 34– 35
DGROUP 51
language types, specifying 26, 308
memory model, defining 35– 36
mode default 46
overview 34
positioning 46
simplified segment directives 33
@Model predefined symbol 35, 83
Module-definition file
described 270
statements
EXPORTS 261, 270
HEAPSIZE 261, 271
IMPORTS 266
LIBRARY 270
STUB 266
Module-specific EXTERN directive See EXTERN directive
MOV instruction 49, 82, 89
MOVS instruction 110– 113, 353
MOVSX instruction 92
MOVZX instruction 92
MS-DOS interrupts 204, 285
MS-DOS operating system 2– 6
MUL instruction 95– 96
Multiple-module programs
alternatives to include files 219

Index 457
Multiple-module programs (continued)

COMM, using 217
data-sharing methods 211
declaring symbols public and external 214
EXTERN with library routines 223
external declarations, positioning 218
EXTERNDEF, using 214
include files 212– 213
libraries 221– 222
modules 212
PROTO, using 216
PUBLIC and EXTERN, using 220
sharing symbols with include files 212
Multiplex interrupt 291, 304
Multiplication
instructions 95
shift operations 102
N
Naming conventions
directives 37
(list) 308
mixed-language programming 308– 309
Naming restrictions 9
Naming restrictions, identifers See Identifiers
NE operator 365
Near address 57, 80
NEAR operator 169, 185
NEARSTACK operand
ASSUME statement 54
default stack type 37, 42
described 35
New features, MASM 6.1 xiv– xv, 342– 351
NMAKE 270
.NO87 directive 38, 349
NODOTNAME argument, OPTION directive 25
NOEMULATOR argument, OPTION directive 27
NOKEYWORD argument, OPTION directive 9, 27, 353, 376
NOLJMP argument, OPTION directive 27, 170
NOM510 argument, OPTION directive 25
NONUNIQUE operand 118, 126
NOOLDMACROS argument, OPTION directive 26
NOOLDSTRUCTS argument, OPTION directive 26
NOP instruction 377
NOREADONLY argument, OPTION directive 27
NOSCOPED argument, OPTION directive 26, 362, 375
NOSIGNEXTEND argument, OPTION directive 27, 378
NOT instruction 99– 100
NOTHING operand 49– 50
Number of items with LENGTHOF operator See LENGTHOF
operator
Numeric equates, compared to text macros 234

458 Index
O LENGTHOF 346
Operators (continued)
OFFSET LOW 356
FLAT argument, OPTION directive 27 LOWWORD 346, 366
GROUP argument, OPTION directive 27 LROFFSET 344
SEGMENT argument, OPTION directive 27, 62 macro 251
OFFSET operator 61, 82, 356, 374 MASK 133
Offsets minus (– ) 64
accessing data with 74 NE 365
addresses 7 NEAR 169, 185
described 5– 7 OFFSET 61, 82, See OFFSET operator
determining 23– 24, 360, 374 OPATTR 252– 254
fixups for 26 plus (+) 63, 66
OLDMACROS argument, OPTION directive 25, 239, 361, precedence 14
372 PTR See PTR operator
OLDSTRUCTS argument, OPTION directive PTR, example See PTR operator
MASM 5.1 compatibility 25, 361, 370– 372 relational 357, 365
structures, with 119, 126 relational (list) 178
Online help See Microsoft Advisor SEG 50, 62, 363
OPATTR operator 252– 253 segment-override (:) 59, 64
Operands SHORT 169
ABS 220 SIZE 364– 365
direct memory 60– 64 size See PTR operator
EXPORT 185 SIZEOF 86, 346
FAR 15 structure-member (.) 64– 67, 126, 352, 370
FARSTACK See FARSTACK operand substitution (&) 238
FLAT 46, 49– 50 .TYPE 252, 360
FORCEFRAME 244 TYPE 86
immediate 60– 62 WIDTH 133
indirect memory 60, 64– 70 OPTION directive
NEAR 15 CASEMAP 25
PRIVATE READONLY 44– 45 described 23
registers 61 DOTNAME 25, 361, 373
size 66, 355 emulation mode 157
USE16 44– 46 EMULATOR 26, 157
USE32 44– 46 EPILOGUE 26, 201– 203
Operating systems EXPR16 OPTION directive 13, 26, 361, 373
(list) 4 EXPR32 OPTION directive 13, 26, 373
.MODEL, specifying with 34 LANGUAGE 26, 193
multitasking 6 language types, specifying 308
types See MS-DOS, Windows operating systems list of arguments for 25
Operators LJMP 26
ADDR 197 M510 See M510 argument, OPTION directive
current address ($) 368 NODOTNAME 25
dot (.) 126, 352, 370 NOEMULATOR 26
EQ 365 NOKEYWORD See NOKEYWORD argument, OPTION
expansion (%) 235– 236, 248, 357 directive
expressions, in 12– 13 NOLJMP 27, 170
FAR 169, 185 NOM510 25
HIGH 356 NOOLDMACROS 26
HIGHWORD 346 NOOLDSTRUCTS 26
index ([ ]) 63 NOREADONLY 27
instructions, compared to 13 NOSCOPED 26, 362, 375
LENGTH 356– 357, 364 NOSIGNEXTEND 27, 378

Index 459
OPTION directive (continued) @InStr 245– 246

OFFSET 26, 62, 362, 374– 375
OLDMACROS 25, 237
OLDSTRUCTS See OLDSTRUCTS argument, OPTION
directive
PROC 185, 375
procedure use 26
PROLOGUE 26, 201– 203
READONLY 26
SCOPED 25
SETIF2 25, 29– 30
using 25, 361
OR instruction 27, 99– 100, 168
ORG directive 56
OUT instruction 5
OVERFLOW? flag as operand 178
P
PAGE align type 45
PARA align type 45
Parentheses [( )] 106
PARITY? flag as operand 178
Pascal convention 310
Physical line 22
Plus operator (+) 66, 352, 370
Pointer variables 74– 78
Pointers
accessing data with 74
arguments, as 80
copying 79
far 74, 80– 81
initializing 78
location 74
operations 78
TYPEDEF, defined with 15, 75– 78
types, to 15
Pointers and conditional Assembly See Conditional assembly
Pointers defined by TYPEDEF See TYPEDEF directive
POP instruction 49, 71
POPA instruction 74
POPAD instruction 74
POPCONTEXT directive 255, 349
POPF instruction 73
POPFD instruction 73
Positioning
EXTERN directive See EXTERN directive
EXTERNDEF directive See EXTERNDEF directive
Precedence operators 14
Predefined equates See Predefined symbols
Predefined functions for macros 11
Predefined string functions
@CatStr 245– 247

460 Index
Predefined string functions (continued)

@SizeStr 245– 246
@SubStr 245– 246
Predefined symbols 39, 83
@Codesize 40
@Cpu 254
@CurSeg 39, 219
@Data 39
@DataSize 39, 83
@Interface 37
(list) 10, 409
@Model 35, 83
@stack 37
@Wordsize 39
case sensitivity 9– 10
new to MASM 6.1 (list) 343
PRIVATE operand 185
Privilege levels 5
Problems, reporting xx
PROC
EXPORT argument, OPTION directive 25
PRIVATE argument, OPTION directive 25, 362
PUBLIC argument, OPTION directive 25, 185
PROC directive 180– 184, 193, 206, 312
PROC statements with visibility See also Visibility
PROC with RET instruction See RET instruction
Procedure prototypes
declaring See EXTERNDEF directive
defined with See PROTO directive
defined with PROTO directive See PROTO directive
writing See PROTO directive
Procedures
arguments
far pointers 197
near addresses 197
passing 182
pointers 80
type conversions 195, 196
CALL instruction 180
calling See INVOKE directive
calls
indirect 198
optimizing 181
defining 180
epilogues 26
EXTERNDEF directive 214–215
See also EXTERNDEF directive,
include files 214
INVOKE directive 193– 197, 216
libraries 221
local variables 188– 192
See also Local variables
Macro See Macros, procedures
new features 347

Index 461
Procedures (continued)
OPTION PROC 375
overview 180
parameters
declaring 184– 186
variable numbers of 186– 188, 194
PROC attributes, specifying 185
prologues 26
PROTO directive 193, 214, 216
See also PROTO directive
prototypes, writing 193
RET instruction 180
RETF instruction 181
RETN instruction 181
syntax description 184
VARARG keyword 186– 188, 194
visibility 25, 375
Processors
See also Real mode; Protected mode
8086-based 2– 3
.MODEL directive 37
modes, determining 10
target 2
timing xvii, 399– 400
Product assistance xx
Program Segment Prefix (PSP) 56
Programming, MASM 6.1 practices 352
Programs
exiting 41
mixed-language 307
starting 41
PROLOGUE argument, OPTION directive 25, 201– 203
Prologue code
arguments, specifying 185
code labels in 357
defined 198
macros for 201– 203, 264– 265
standard 199
user-defined 26, 201
Protected mode
described 2– 7, 335
flat model 335
read-only segments 45
PROTO directive
include files 211, 214– 216
procedure prototypes, defined with 193
procedure prototypes, writing 312
Prototypes
procedure
directives for 193
overview 193
qualifiedtypes, defined with 15
PTR operator
example 92

462 Index
PTR operator (continued) RECORD syntax 130– 131

pointer to type, as 15
signed number, specifying 178
size 66, 88
TYPEDEF, used with 75
PUBLIC combine type 45
PUBLIC directive 185, 211, 220
PUSH instruction 49, 71
PUSHA instruction 74
PUSHAD instruction 74
PUSHCONTEXT directive 255, 349
PUSHF instruction 73
PUSHFD instruction 73
Q
Quadwords 86
Qualifiedtypes
BNF grammar 16
defined 15
pointers, defining 75– 76
prototypes, as 15
rules for use 15– 16
Question mark initializer ( ? )
array elements 109
described 368
variables 87
Quotation marks (' or ") 109
QWORD directive 86
R
.RADIX directive 11
Radix specifiers
(list) 11
RCL instruction 101– 104
RCR instruction 101– 104
Read-only code 27
READONLY argument, OPTION directive 26
READONLY operand 44– 45
Real mode 2, 4, 7
Real numbers See Floating-point
REAL4 directive 136– 137
RECORD directive 130– 131
Records
defined 129
field ranges 354
LENGTH operator 357
operators 133– 134

Index 463
Records with SIZEOF operator See SIZEOF operator

Records with TYPE operator See TYPE operator
Recursive macros 255
Register operands 61
Registers
16-bit 16– 17, 67
32-bit 335
base 65– 70
coprocessor 140
copying pairs of 82
division (table) 98
Eflags 20
extended 17
flags 20
FS 17
general purpose 19
GS 17
index 65– 69
indirect addressing 65
indirect operands 67– 68
initializing 44
Instruction Pointer (IP) 20, 57, 161
(list) 409
loading addresses into 80
mixed 16-bit, 32-bit 70
pointers as 77
scaling 67– 69
segment See Segment registers
Stack Pointer (SP) 19
Stack Segment (SS) 73
stacks, saving on 74
types, defined with ASSUME 77
Relational operators (list) 178
Relocatable
addresses 57
expressions 62, 65
REP instruction 110– 112, 363
REPE instruction 110– 112, 363
Repeat blocks 239
.REPEAT directive 173
REPEAT directive 240
REPNE instruction 110– 112, 353, 363
REPNZ instruction 110– 112, 353, 363
Reporting problems xx
REPZ instruction 110– 112
Reserved words
described 8, 26
(list) 407
OPTION NOKEYWORD 376
RET instruction
epilogue code, generating 200, 378
instruction encodings, changes to 357
PROC, with 180

464 Index
RETF instruction 181, 378

RETN instruction 181, 378
ROL instruction 101– 104
ROM-BIOS interrupts See Interrupts
ROR instruction 101– 104
Rotate instructions 101
Routines, interrupt 206
S
SAL instruction 101– 104
SAR instruction 101– 104
SBB instruction 92– 94
SBYTE directive 86
Scaling factor 107
Scaling index registers 67– 69
SCAS instruction 110– 112, 115, 353
Scope within visibility See also Visibility
SCOPED argument, OPTION directive 26
SDWORD directive 86
SEG operator 49, 62, 363
SEGMENT
FLAT argument, OPTION directive 27
USE16 argument, OPTION directive 27
USE32 argument, OPTION directive 27
Segment arithmetic 7
SEGMENT directive 44– 47
Segment mode, setting See .386 directive; .486 directive
Segment registers
32-bit 335
assigning 59, 62
ASSUME directive 49– 55, 58– 59, 357
changing 57
default 60, 64
described 18
FS 18
GS 18
initializing 43, 54– 57
MS-DOS, under 24, 43
near code 57
restoring 59
segment-override operator (:) 50, 59– 60, 64
Segment registers
initializing See STARTUP directive
setting See STACK directive
Segment selectors 5
Segment-override operator (:) 50, 59– 60, 64
Segmented architecture 2, 5
Segments
32-bit 36, 335
accessing data 74
aligning 44– 45
class types 44, 47– 48

Index 465
Segments (continued)
code
creating 40
far 40
memory model support for 36
near 40
combining 40, 44– 46
current 10
data
creating 39
default 49, 54– 55, 59
far 40
memory model support for 36
near 39
defined 31
described 5– 7
determining order of 47– 48
determining position of 23– 24
determining size of 44
fixups for 26
full segment definitions, defining 32
groups, defining 51
initializing 55
location of 6
naming 40
ordering with the linker 48
protection 6
READONLY 45
simplified segment directives 37– 42
size, determining 10
types 44
USE16 44
USE32 44
values 55
word size, setting 46
Selector 335
Semicolon (;), comments 21
.SEQ directive 47
SETIF2 argument, OPTION directive 25, 29– 30
Shift instructions 100
SHL instruction 101– 104
SHORT operator 169
SHR instruction 101– 104
Sign-extending integers 90
SIGN? flag as operand 178
Signed data 14, 91
Signed numbers, specifying See PTR operator
Significand 139
Simplified segment directives
code segments 41
code, starting and ending 42
data segments 40
described 32
language convention 36

466 Index
Simplified segment directives (continued)

memory model 35
.MODEL, defining with 34
operating system 35
processor 38
segment registers, initializing 54– 56
stack 39
stack distance 37
using 33
Single quotation mark (') 109
Size attribute, segments
FLAT 46
USE16 46
USE32 46
Size mismatch 355
Size of strings See SIZEOF operator
SIZE operator 364, 365
@SizeStr predefined string function 245– 246
SIZEOF operator
arrays, with 108
described 346
records, with 132
strings, with 110
structures, with 124
types 86
unions, with 125
SIZESTR directive 245– 246
Small model See Memory models, small
Source code, statements in 21
SP (Stack Pointer) register 19, 71– 73
SS (Stack Segment) register 73
STACK combine type 45
.STACK directive
described 33
segment registers, setting 56
Stack distance 37
Stack frame 73, 200, 264– 265
Stack Pointer (SP) register 19
@stack predefined symbol 37
Stack Segment (SS) register 73
Stacks
cleaning 185
creating 38
described 71
distance 37
far 10
FARSTACK 35, 37
in DLLs 264– 267
local variables on 188– 191
near 10
NEARSTACK 33, 35– 37
operations with 72– 74
operators 71
passing arguments 182

Index 467
Stacks (continued)
pointer 71– 73
POP instructions 71
PUSH instructions 71
saving flags 73
saving registers 74
segment register 18
separate 46
trace 264
.STARTUP directive
described 33
initializing segments 54– 56
program, starting 41– 42
segment address 37
Statements
case sensitivity 22
syntax 21
Status flags, saving 73
STC instruction 104
STDCALL calling convention 311, 336
STI instruction 5, 209
STOS instruction 110– 113, 353
Strings
declaring 109
defined 105
defining 15
initializing 109
instructions
processing, for 110
requirements (table) 112, 353
length of 110
multiple-line declarations for 109
overview 111
predefined functions for macros 11
See also Predefined string functions
size of 110
type of 110
STRUCT directive 118– 129
Structure-member operator (.) 64– 67, 126, 352, 370
Structures
alignment of fields 118– 119
array initializers 122
arrays 124
compatibility with MASM 5.1 25, 118
current address operator ($) 368
default field values 122
defined 117
fields
accessing 64, 67, 371
initializing 118
naming 119, 352, 372
initializers, as 123
MASM 5.1 behavior 25, 355, 370
memory allocation 117

468 Index
Structures (continued) Trap flag 205

nested 128– 129
new features 345
operators 124
OPTION OLDSTRUCTS 370
redeclaration 124, 355
referencing fields in 126
steps for using 118
string initializers 122, 368
syntax
types 118
variables 121
Structures with LENGTHOF operator See LENGTHOF
operator
Structures with SIZEOF operator See SIZEOF operator
Structures with TYPE operator See TYPE operator
STUB statement 266
SUB instruction 92– 94
Substitution operator (&) 238, 372
SUBSTR directive 245– 246
@SubStr predefined string function 245– 246
SWORD directive 86
Symbol table, listing files 405
Symbols
declaring public and external 214, 220
external 369
naming 346, 368
predefined 9– 11
Symbols, declaring by EXTERNDEF directive See
EXTERNDEF directive
Syntax, MASM 6.1 statements 21
SYSCALL calling convention 308– 311
System date 11
System time 11
T
Tables, lookup 241
Target environment 4
TBYTE directive 86, 159
Terminate-and-Stay-Resident programs See TSRs
TEST instruction 167– 168
Testing for zero 168
Text delimiters See Angle brackets
Text macros See Macros, text
TEXTEQU directive
aliases 369
CATSTR, compared with 247
syntax 226
Time, system 11
Timing (cycle/second) xvii, 399– 400
Tiny model See Memory models, tiny

Index 469
TSRs memory allocation 117

active Unions (continued)
described 275 nested 128– 129
interrupt handlers in 275 operators 125
MS-DOS functions, calling 285 referencing fields in 126
MS-DOS functions, interrupting 286, 302 steps for using 118
deinstalling 292, 305 strings as initializers 122
described 273 types 118
errors, trapping 288– 289 variables 121, 127
examples Unpacked BCD numbers 160
ALARM.ASM 279– 280, 284 Unsegmented architecture 5
SNAP.ASM 293– 305 Unsigned data 91
existing data, preserving 290, 303 .UNTIL directive 173
hardware events, auditing 275– 276, 299 .UNTILCXZ directive 173
interrupt handlers 275 USE16 operand 44– 46
monitoring USE32 operand 44– 46
Critical Error flag 287 USES in PROC statement 184
system status 277, 300 Utilities
MS-DOS internal stacks (lists) 286 IMPLIB 258
multiplex interrupt 290, 304 MASM 342
passive 274 ML xvi
Type conversions See INVOKE directive
Type of strings See TYPE operator
TYPE operator V
and OPATTR 252– 253 VARARG keyword
arrays, with 108 macros, used in 242, 249, 351
compatibility 360, 365 procedures, used with 186– 188, 194
records, with 132 Variables
string, with 110 assembly-time 233
structures, with 124 communal 217
types 86 environment 10, 213, 222
unions, with 125 external 217, 369
TYPEDEF directive floating-point 136– 138
aliases, created by 87, 137 global 211
BNF, from 380 initializing 87
data types, defining 87 integers, allocating memory for 85– 86
indirect operands, defining 163 local address, loading 82
pointers, defined by 15, 75– 78 naming restrictions 9
procedure declarations 193 Virtual memory 5
procedure prototypes 193 Virtual-86 mode 2, 335
qualifiedtypes 16 Visibility
TYPEDEF, used with PTR operator See PTR operator PROC statement 25, 185
Types, data See Data types scope, within 9
U W
Unconditional jumps 162 WDEB386 debugger 264
UNION directive 118– 119, 122, 125– 129 WEP (Windows Exit Procedure) 263– 264, 270
Unions .WHILE directive 173
arrays as initializers 122 WHILE directive 241
arrays of 124 WIDTH operator 133
defined 117 Windows operating system
fields 119, 127– 129 API 257, 262

470 Index
applications 258, 261

DLLs 261
Windows operating system (continued)
exit codes 263
MS-DOS, compared 4
programming for 4
protected mode 2, 6
SDK 268
task header 265, 269
Windows NT 3– 5
WORD align type 45
WORD directive 86
Word size
default 13, 363, 373
expressions, in 13, 26
@WordSize predefined symbol 39
Words, reserved See Reserved words
X
XCHG instruction 90
/X command-line option, ML 213
XLAT instruction 116
XLATB instruction 116
XOR instruction 27, 99– 100
Z
ZERO? flag as operand 178
/Zm command-line option, ML 62, 119
/Zp command-line option, ML 119

Documentation Feedback — Microsoft ® Macro Assembler Version 6.1
Please help us improve our documentation. When you have used MASM 6.1 for a while, please complete and return this form. Use the
back of the form for additional suggestions and comments. Suggestions and comments become the property of Microsoft Corporation.
Rate each component of the document set: What one thing would you like to see added to or removed from
Rate each from 1 (never use) to 5 (often use). each component?
Getting Started _____________________________________
1 2 3 4 5 Getting Started
_________________________________________________
1 2 3 4 5 Programmer’s Guide _________________________________________________
1 2 3 4 5 Environment and Tools _________________________________________________
1 2 3 4 5 Reference Programmer’s Guide _________________________________

_________________________________________________
1 2 3 4 5 Online Help _________________________________________________
List what stands out most about each component: _________________________________________________
List the one or two things that you like, dislike, or both. Or, use Environment and Tools _______________________________
the space to comment on the rating you gave the component in _________________________________________________
the previous section. _________________________________________________
Getting Started______________________________________ _________________________________________________
__________________________________________________ Reference__________________________________________
__________________________________________________ _________________________________________________
__________________________________________________ _________________________________________________
Programmer’s Guide _________________________________ _________________________________________________
__________________________________________________ Online Help ________________________________________
__________________________________________________ _________________________________________________
__________________________________________________ _________________________________________________
Environment and Tools _______________________________ _________________________________________________
__________________________________________________ How many years of programming experience do you have?
__________________________________________________ With assembly language
__________________________________________________
With other programming languages (including application
Reference __________________________________________ macro languages)
__________________________________________________
__________________________________________________
__________________________________________________
Online Help_________________________________________
__________________________________________________
__________________________________________________
__________________________________________________
Filename: LMAPGDFD.DOC Project: Documentation Feedback Form MASM 6.1

Template: FEEDBACK.DOT Author: Ruth L Silverio Last Saved By: Mike Eddy
Name _____________________________________________
Address ___________________________________________
City/State/Zip _______________________________________
Home Phone ( _____ ) ________________________________
Work Phone ( _____ ) ________________________________
May we contact you for additional information about your comments? Yes No
Additional comments:
......................................................................................................................Fold......................................................................................................................
Place stamp
here.
Post Office
will not deliver
without proper
postage.
Microsoft Corporation
Languages — MASM 6.1
One Microsoft Way
Redmond WA 98052-9953
Filename: LMAPGDFD.DOC Project: Documentation Feedback Form MASM 6.1

Template: FEEDBACK.DOT Author: Ruth L Silverio Last Saved By: Mike Eddy

MASM611 - Programmer's Guide

Uploaded by

MASM611 - Programmer's Guide

Uploaded by

Programmer’s Guide

Assembly-Language Development System

For MS-DOS ® and Windows ™ Operating Systems

Filename: LMAPGTTL.DOC Project:

©1992 Microsoft Corporation. All rights reserved.

U.S. Patent No. 4,955,066

Hercules is a registered trademark of Hercules Computer Technology.

Document No. DB35747-1292

Filename: LMAPGCPY.DOC Project:

Chapter 1 Understanding Global Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 2 Organizing Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Filename: LMAPGTOC.DOC Project:

Defining Basic Attributes with .MODEL . . . . . . . . . . . . . . . . . . . . . . . . 34

Chapter 3 Using Addresses and Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Chapter 4 Defining and Using Simple Data Types . . . . . . . . . . . . . . . . . . . . . . . 85

Filename: LMAPGTOC.DOC Project:

Multiplying and Dividing with Shift Instructions . . . . . . . . . . . . . . . . . . 102

Chapter 5 Defining and Using Complex Data Types. . . . . . . . . . . . . . . . . . . . . 105

Chapter 6 Using Floating-Point and Binary Coded Decimal Numbers . . . . . . . 135

Chapter 7 Controlling Program Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Filename: LMAPGTOC.DOC Project:

Passing Arguments on the Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

Chapter 9 Using Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Filename: LMAPGTOC.DOC Project:

Returning Values with Macro Functions. . . . . . . . . . . . . . . . . . . . . . . . . . 248

Chapter 10 Writing a Dynamic-Link Library For Windows . . . . . . . . . . . . . . . . 257

Chapter 11 Writing Memory-Resident Software . . . . . . . . . . . . . . . . . . . . . . . . 273

Filename: LMAPGTOC.DOC Project:

Communicating Through the Multiplex Interrupt . . . . . . . . . . . . . . . . . . . . 290

Chapter 12 Mixed-Language Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Chapter 13 Writing 32-Bit Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

Filename: LMAPGTOC.DOC Project:

Rewriting Code for Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

Appendix B BNF Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

Appendix Generating and Reading Assembly Listings. . . . . . . . . . . . . . . . . . 397

Appendix D MASM Reserved Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

Appendix E Default Segment Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417

Filename: LMAPGTOC.DOC Project:

Figures and Tables

Filename: LMAPGTOC.DOC Project:

Filename: LMAPGTOC.DOC Project:

New and Extended Features in MASM 6.1

Filename: LMAPGINT.DOC Project:

MASM Features New Since Version 5.1

Filename: LMAPGINT.DOC Project:

MASM 6.1 has other improved capabilities, such as:

MASM Features New Since Version 6.0

Filename: LMAPGINT.DOC Project:

u Program listings now show instruction timings. The number of required

ML and MASM Command Lines

Compatibility with Earlier Versions of MASM

Filename: LMAPGINT.DOC Project:

A Word About Instruction Timings

Filename: LMAPGINT.DOC Project:

Books for Further Reading

Books About Programming in Assembly Language

Books About MS-DOS and BIOS

Filename: LMAPGINT.DOC Project:

Books and Articles About Windows

Books About Other Topics

Filename: LMAPGINT.DOC Project:

Getting Assistance and Reporting Problems

Filename: LMAPGINT.DOC Project:

Understanding Global Concepts

The Processing Environment

Filename: LMAPGC01.DOC Project:

Filename: LMAPGC01.DOC Project:

8086 and 8088

80186 and 80188

8087, 80287, and 80387

Filename: LMAPGC01.DOC Project:

Filename: LMAPGC01.DOC Project:

Filename: LMAPGC01.DOC Project:

With the advent of protected-mode processors such as the 80286, segmented

Figure 1.1 Segment Allocation

Segment protection makes software development easier and more reliable in