MIPS

Assembly Language
Programmer’s Guide

ASM-01-DOC

Your comments on our products and publications

are welcome. A postage-paid form is provided for
this purpose on the last page of this manual.

Part Number 02-00036-005

October 1992
Copyright  1989-1992 Silicon Graphics, Inc. All Rights Reserved.

This manual documents MIPS Pascal version 3.10.

RISComputer, RISCwindows, and RISC/os are trademarks of Silicon Graphics,, Inc.

UNIX is a Trademark of UNIX System Laboratories, Inc.

Silicon Graphics, Inc.

2011 North Shoreline Blvd.
Mountain View, CA 94039-7311

Customer Service
U.S. and Canada: 1 (800) 800-4SGI
International: Contact your local sales representative
Preface: About This Book

This book describes the assembly language supported by the RISCompiler

system, its syntax rules, and how to write assembly programs. For
information on assembling and linking an assembly language program, see
the MIPS RISCompiler and C Programmer’s Guide.
The assembler converts assembly language statements into machine code. In
most assembly languages, each instruction corresponds to a single machine
instruction; however, some assembly language instructions can generate
several machine instructions. This feature results in assembly programs that
can run without modification on future machines, which might have different
machine instructions.
See Appendix B for more information about assembler instructions that
generate multiple machine instructions.

Audience
This book assumes that you are an experienced assembly language
programmer. The assembler produces object modules from the assembly
instructions that the C, Fortran 77, and Pascal compilers generate. It therefore
lacks many functions normally present in assemblers. You should use the
assembler only when you need to:
• Maximize the efficiency of a routine, which might not be possible in
C, Fortran 77, Pascal, or another high-level language; for example, to
write low-level I/O drivers.
• Access machine functions unavailable in high-level languages or
satisfy special constraints such as restricted register usage.
• Change the operating system.
• Change the compiler system.
Further system information can be obtained from the manuals listed at the end
of this section.

Assembly Language Programmer’s Guide iii

About This Book

Topics Covered
This book has these chapters:
• Chapter 1: Registers describes the format for the general registers,
the special registers, and the floating point registers.
• Chapter 2: Addressing describes how addressing works.
• Chapter 3: Exceptions describes exceptions you might encounter
with assembly programs.
• Chapter 4: Lexical Conventions describes the lexical conventions
that the assembler follows.
• Chapter 5: Instruction Set describes the main processor’s
instruction set, including notation, load and store instructions,
computational instructions, and jump and branch instructions.
• Chapter 6: Coprocessor Instruction Set describes the coprocessor
instruction sets.
• Chapter 7: Linkage Conventions describes linkage conventions for
all supported high-level languages. It also discusses memory
allocation and register use.
• Chapter 8: Pseudo-Op-Codes describes the assembler’s pseudo-
operations (directives).
• Chapter 9: MIPSObject File Format provides an overview of the
components comprising the object file and describes the headers and
sections of the object file.
• Chapter 10: Symbol Table describes the purpose of the Symbol
Table and the format of entries in the table. This chapter also lists the
symbol table routines that are supplied.
• Chapter 11: Execution and Linking Format describes Execution
and Linking Format (ELF) for object files. This chapter also
describes the components of an elf object file, symbol table format,
global data area, register information, and relocation.
• Chapter 12: Program Loading and Dynamic Linking describes the
object file structures that relate to program execution. This chapter
also describes how the process image is created from executable files
and object files.
• Appendix A: Instruction Summary summarizes all assembler
instructions.
• Appendix B: Basic Machine Definition describes instructions that
generate more than one machine instruction.
• Index. Contains index entries for this publication.

iv Assembly Language Programmer’s Guide

 About This Book

Special Text Notations

Several special notations are used throughout this manual to differentiate
among the following types of information:

❋ Note
A note presents information of greater-than-normal importance.
☞

Detail
Printed in a sans serif font, a detail presents additional information
that is of ancillary importance.

For More Information

As you use this manual, consult the following book(s):
• RISCompiler and C Programmer’s Guide
(Order number CMP-01-DOC)
• MIPS RISC Architecture
(Order number SYS-02-DOC)
• MIPS RISC/os Programmer’s Reference Manual
(ROS-01-DOC)
• MIPS RISC/os User’s Reference Manual
(ROS-02-DOC)

Assembly Language Programmer’s Guide v

About This Book

vi Assembly Language Programmer’s Guide

Contents

Preface: About This Book

Audience...................................................................................................iii
Topics Covered.........................................................................................iv
Special Text Notations .............................................................................. v
For More Information................................................................................ v

1
Registers
Register Format .................................................................................1-1
Special Registers ...................................................................................1-5

2
Addressing
Address Formats.....................................................................................2-2
Address Descriptions..............................................................................2-3

3
Exceptions
Main Processor Exceptions ....................................................................3-1
Floating-Point Exceptions .....................................................................3-2

4
Lexical Conventions
Tokens ....................................................................................................4-1
Comments...............................................................................................4-2
Identifiers................................................................................................4-2
Constants ................................................................................................4-2
Scalar Constants...............................................................................4-3
Floating Point Constants ..................................................................4-3
String Constants ...............................................................................4-4

Assembly Language Programmer’s Guide vii

 Multiple Lines Per Physical Line ...........................................................4-5
Statements...............................................................................................4-6
Label Definitions .............................................................................4-6
Null Statements................................................................................4-7
Keyword Statements ........................................................................4-7
Expressions.............................................................................................4-7
Precedence .......................................................................................4-7
Expression Operators .......................................................................4-8
Data Types .......................................................................................4-8
Type Propagation in Expressions...................................................4-10

5
Instruction Set
Instruction Classes..................................................................................5-1
Reorganization Constraints and Rules....................................................5-2
Instruction Notation ...............................................................................5-2
Load and Store Instructions ...................................................................5-3
Load and Store Formats ...................................................................5-3
Load Instruction Descriptions..........................................................5-4
Store Instruction Descriptions..........................................................5-7
Computational Instructions ..................................................................5-10
Computational Formats..................................................................5-10
Computational Instruction Descriptions ........................................5-13
Jump and Branch Instructions ..............................................................5-21
Jump and Branch Formats .............................................................5-21
Jump and Branch Instruction Descriptions ....................................5-23
Special Instructions .............................................................................5-25
Special Formats..............................................................................5-25
Special Instruction Descriptions ....................................................5-26
Coprocessor Interface Instructions .......................................................5-27
Coprocessor Interface Formats ......................................................5-27
Coprocessor Interface Instruction Descriptions ............................5-28

6
Coprocessor Instruction Set
Instruction Notation................................................................................6-1
Floating-Point Instructions .....................................................................6-2
Floating-Point Formats ....................................................................6-3
Floating-Point Load and Store Formats...........................................6-3

viii Assembly Language Programmer’s Guide

 Floating-Point Load and Store Descriptions....................................6-4
Floating-Point Computational Formats............................................6-4
Floating-Point Computational Instruction Descriptions ..................6-7
Floating-Point Relational Operations .....................................................6-8
Floating-Point Relational Instruction Formats...............................6-10
Floating-Point Relational Instruction Descriptions .......................6-11
Floating-Point Move Formats ........................................................6-13
Floating-Point Move Instruction Descriptions...............................6-13
System Control Coprocessor Instructions ............................................6-13
System Control Coprocessor Instruction Formats .........................6-13
System Control Coprocessor Instruction Descriptions ..................6-14
Control and Status Register ..........................................................6-15
Floating-Point Rounding................................................................6-20

7
Linkage Conventions
Introduction ............................................................................................7-1
Program Design ......................................................................................7-2
Register Use and Linkage ................................................................7-2
The Stack Frame ..............................................................................7-3
The Shape of Data............................................................................7-7
Examples ................................................................................................7-7
Learning by Doing................................................................................7-11
Calling a High-Level Language Routine .......................................7-11
Calling an Assembly Language Routine ..............................................7-13
Memory Allocation ..............................................................................7-15

8
Pseudo Op-Codes

9
MIPS Object File Format
Overview ................................................................................................9-2
The File Header ......................................................................................9-4
File Header Magic Field (f_magic)..................................................9-5
Flags (f_flags) ..................................................................................9-5
Optional Header......................................................................................9-7
Optional Header Magic Field (magic) .............................................9-8
Section Headers ......................................................................................9-8

Assembly Language Programmer’s Guide ix

 Section Name (s_name) ...................................................................9-9
Flags (s_flags)................................................................................9-10
Global Pointer Tables ....................................................................9-11
Shared Library Information ...........................................................9-12
Section Data..........................................................................................9-12
Section Relocation Information............................................................9-15
Relocation Table Entry ..................................................................9-15
Assembler and Link Editor Processing .........................................9-16
Object Files...........................................................................................9-22
Impure Format (OMAGIC) Files...................................................9-23
Shared Text (NMAGIC) Files .......................................................9-24
Demand Paged (ZMAGIC) Files ...................................................9-25
Target Shared Library (LIBMAGIC) Files....................................9-28
Objects Using Shared Libraries .....................................................9-28
Ucode objects.................................................................................9-29
Loading Object Files......................................................................9-29
Archive files .........................................................................................9-30
Link Editor Defined Symbols...............................................................9-31
Runtime Procedure Table Symbols ...............................................9-32

10
Symbol Table
Overview ..............................................................................................10-2
Format of Symbol Table Entries ..........................................................10-8
Symbolic Header............................................................................10-8
Line Numbers.................................................................................10-9
Procedure Descriptor Table .........................................................10-13
Local Symbols .............................................................................10-13
Optimization Symbols .................................................................10-17
Auxiliary Symbols .......................................................................10-17
File Descriptor Table ...................................................................10-20
External Symbols .........................................................................10-21

11
Execution and Linking Format
Object File Format................................................................................11-2
ELF Header ..........................................................................................11-3
Sections.................................................................................................11-7
Section Header Table.....................................................................11-7

x Assembly Language Programmer’s Guide

 Section Header ...............................................................................11-8
Special Sections ...........................................................................11-14
String Tables ................................................................................11-18
ELF Symbol Table .............................................................................11-18
Symbol Type ......................................................................................11-21
Symbol Values .............................................................................11-22
Global Data Area................................................................................11-23
Register Information...........................................................................11-25
Relocation...........................................................................................11-26

12
Program Loading and Dynamic Linking
Program Header....................................................................................12-2
Base Address..................................................................................12-4
Segment Permissions .....................................................................12-4
Segment Contents ..........................................................................12-5
Program Loading ..................................................................................12-6
Dynamic Linking..................................................................................12-9
Program Interpreter........................................................................12-9
Dynamic Linker .............................................................................12-9
Dynamic Section..........................................................................12-11
Shared Object Dependencies..............................................................12-18
Global Offset Table (GOT) ................................................................12-19
Calling Position Independent Functions ......................................12-20
Symbols........................................................................................12-22
Relocations...................................................................................12-22
Hash table ...........................................................................................12-23
Initialization and Termination Functions ...........................................12-23
Quickstart ...........................................................................................12-24
Shared Object List........................................................................12-24
Conflict Section ...........................................................................12-26
Ordering .......................................................................................12-26

A
Instruction Summary

Assembly Language Programmer’s Guide xi

 B
Basic Machine Definition
Load and Store Instructions...................................................................B-1
Computational Instructions ...................................................................B-2
Branch Instructions................................................................................B-3
Coprocessor Instructions .......................................................................B-3
Special Instructions ...............................................................................B-3

Index

xii Assembly Language Programmer’s Guide

Figures

Figure 1-1: Big-endian Byte Ordering ..................................................... 1-2

Figure 1-2: Little-endian Byte Ordering .................................................. 1-2
Figure 4-1: Section and location counters ................................................ 4-5
Figure 6-1: Floating-point Formats .......................................................... 6-3
Figure 6-2: Floating Control and Status Register 31.............................. 6-15
Figure 7-1: Stack Organization ................................................................ 7-4
Figure 7-2: Stack Example ....................................................................... 7-5
Figure 7-3: Non-Leaf Procedure .............................................................. 7-8
Figure 7-4: Leaf Procedure Without Stack Space for
Local Variables...................................................................... 7-9
Figure 7-5: Leaf Procedure With Stack Space for
Local Variables.................................................................... 7-10
Figure 7-6: Layout of memory (User Program View) ........................... 7-16
Figure 9-1: Object File Format................................................................. 9-3
Figure 9-2: Organization of Section Data .............................................. 9-13
Figure 9-3: Relocation Table Entry for Undefined
External Symbols................................................................. 9-17
Figure 9-4: Relocation Table Entry for a Local
Relocation Entry .................................................................. 9-18
Figure 9-5: Layout of OMAGIC Files in Virtual Memory .................... 9-23
Figure 9-6: Layout of NMAGIC Files in Virtual Memory .................... 9-24
Figure 9-7: Layout of ZMAGIC Files in Virtual Memory..................... 9-26
Figure 9-8: Layout of a ZMAGIC File on Disk ..................................... 9-27
Figure 10-1: The Symbol Table - Overview ............................................ 10-2
Figure 10-2: Functional Overview of the Symbolic Header .................... 10-4
Figure 10-3: Logical Relationship between the File
Descriptor Table and Local Symbols .................................. 10-5
Figure 10-4: Physical Relationship of a File Descriptor
Entry to Other Tables .......................................................... 10-6
Figure 10-5: Logical Relationship between the File
Descriptor Table and Other Tables...................................... 10-7
Figure 10-6: Source Listing for Line Number Example ........................ 10-11
Figure 10-7: Source Listing for Line Number Example ........................ 10-12
Figure 12-1: Example Executable File ..................................................... 12-7

Assembly Language Programmer’s Guide xiii

xiv Assembly Language Programmer’s Guide
Tables

Table 1-1: General (Integer) Registers..............................................1-4

Table 1-2: Special Registers .............................................................1-5
Table 1-3: Floating-Point Registers ..................................................1-6
Table 2-1: Address Formats ..............................................................2-2
Table 2-2: Assembler Addresses.......................................................2-3
Table 4-1: Backslash Conventions....................................................4-4
Table 4-2: Expression Operators.......................................................4-8
Table 4-3: Data Types.......................................................................4-9
Table 5-1: Load and Store Formats...................................................5-3
Table 5-2: Load and Store Formats for mips3
Architecture Only ............................................................5-4
Table 5-3: Load Instruction Descriptions .........................................5-4
Table 5-4: Load Instruction Descriptions for mips3
Architecture Only ............................................................5-6
Table 5-5: Store Instruction Descriptions .........................................5-7
Table 5-6: Store Instruction Descriptions for mips3
Architecture Only ............................................................5-9
Table 5-7: Computational Instruction Formats ...............................5-10
Table 5-8: Computational Instruction Formats for mips3 Architecture
Only ...............................................................................5-12
Table 5-9: Computational Instruction Descriptions........................5-13
Table 5-10: Computational Instruction Descriptions for
mips3 Architecture Only ...............................................5-18
Table 5-11: Jump and Branch Instruction Formats...........................5-21
Table 5-12: Jump and Branch Instruction Descriptions....................5-23
Table 5-13: Special Instruction Formats ...........................................5-25
Table 5-14: Special Instruction Descriptions....................................5-26
Table 5-15: Coprocessor Interface Instruction Formats ...................5-27
Table 5-16: Coprocessor Interface Instruction Descriptions ............5-28
Table 6-1: Floating-Point Load and Store Formats ..........................6-3
Table 6-2: Floating-Point Load and Store Descriptions ...................6-4
Table 6-3: Floating-Point Computational Instruction
Formats ............................................................................6-4

Assembly Language Programmer’s Guide xv

 Table 6-4: Floating-Point Computational Instruction
Formats for mips3 Architecture Only..............................6-6
Table 6-5: Floating-Point Computational Instruction
Descriptions .....................................................................6-7
Table 6-6: Floating-Point Relational Operators................................6-8
Table 6-7: Floating-Point Relational Instruction Formats ..............6-10
Table 6-8: Floating-Point Relational Instruction
Descriptions ...................................................................6-11
Table 6-9: Floating-Point Move Instruction Formats .....................6-13
Table 6-10: Floating-Point Move Instruction Descriptions ..............6-13
Table 6-11: System Control Instruction Formats..............................6-13
Table 6-12: System Control Coprocessor Instruction
Descriptions ...................................................................6-14
Table 7-1: Floating-Point Registers ..................................................7-2
Table 8-1: Pseudo Op-Codes ............................................................8-1
Table 9-1: File Header Format..........................................................9-4
Table 9-2: File Header Magic Numbers ...........................................9-5
Table 9-3: File Header Flags.............................................................9-6
Table 9-4: Optional Header Definition .............................................9-7
Table 9-5: RISC/os Magic Numbers.................................................9-8
Table 9-6: Section Header Format ....................................................9-8
Table 9-7: Section Header Constants for Section Names .................9-9
Table 9-8: Format of s_flags Section Header Entry .......................9-10
Table 9-9: Format of a Relocation Table Entry ..............................9-15
Table 9-10: Section Numbers for Local Relocation Entries .............9-15
Table 9-11: Relocation Types ...........................................................9-16
Table 9-12: Link Editor Defined Symbols........................................9-31
Table 10-1: Format of the Symbolic Header ....................................10-8
Table 10-2: Format of a Line Number Entry ....................................10-9
Table 10-3: Source Listing for Line Number Example ..................10-11
Table 10-4: Format of a Procedure Descriptor Table Entry ...........10-13
Table 10-5: Format of a Local Symbols Entry ...............................10-13
Table 10-6: Index and Value as a Function of Symbol
Type and Storage Class ...............................................10-14
Table 10-7: Symbol Type (st) Constants Supported by
the Compiler ................................................................10-16
Table 10-8: Storage Class Constants Supported by
the Compiler ................................................................10-16
Table 10-9: Storage Class Constants Supported by
the Compiler ................................................................10-17
Table 10-10: Format of an Type Information Record Entry.............10-18

xvi Assembly Language Programmer’s Guide

 Table 10-11: Basic Type (bt) Constants ...........................................10-19
Table 10-12: Type Qualifier (tq) Constants ......................................10-19
Table 10-13: Format of File Descriptor Entry ..................................10-20
Table 10-14: Format an Entry in External Symbols .........................10-21
Table 11-1: sh_link and sh_info values ..........................................11-13
Table 11-2: Special Sections...........................................................11-14
Table 11-3: Relocation Calculations...............................................11-28
Table 12-1: p_flags Values and Interpretation..................................12-5
Table 12-2: Text and Data Segments ................................................12-7
Table 12-3: Example Shared Object Segment Addresses.................12-8
Table 12-4: Dynamic Arrays Tags d_tag ........................................12-15
Table A-1: Main Processor Instruction Summary ............................A-2
Table A-2: System Coprocessor Instruction Summary ....................A-6
Table A-3: Floating Point Instruction Summary ..............................A-6

Assembly Language Programmer’s Guide xvii

xviii Assembly Language Programmer’s Guide
Registers

Chapter 1
1
This chapter describes the organization of data in memory, and the naming
and usage conventions that the assembler applies to the CPU and FPU
registers. See Chapter 7 for information regarding register use and linkage.

Register Format
The CPU’s byte ordering scheme (or endian issues) affects memory
organization and defines the relationship between address and byte position
of data in memory.
The byte ordering is configurable (configuration occurs during hardware
reset) into either big-endian or little-endian byte ordering. When configured
as a big-endian system, byte 0 is always the most-significant (leftmost) byte.
When configured as a little-endian system, byte 0 is always the least-
significant (rightmost byte).
Figure 1-1 and Figure 1-2 illustrate the ordering of bytes within words and the
ordering of halfwords for big and little endian systems.

Assembly Language Programmer’s Guide 1-1

 Chapter 1

Word

Bit: 31 .... 24 23 .... 16 15 ... 8 7 .... 0

Chapter 1

byte 0 byte 1 byte 2 byte 3

sign & most-

significant bits
Halfword

Bit: 15 ... 8 7 .... 0

byte 0 byte1

sign & most-

significant bits

Figure 1-1: Big-endian Byte Ordering

Word
Bit: 31 .... 24 23 .... 16 15 ... 8 7 .... 0
byte 3 byte 2 byte 1 byte 0

sign & most-

significant bits
Halfword
Bit: 15 ... 8 7 .... 0

byte1 byte 0

sign & most-

significant bits

Figure 1-2: Little-endian Byte Ordering

1-2 Assembly Language Programmer’s Guide

 Registers

General Registers
The CPU has thirty-two 32-bit registers. In the mip3 architecture, the size of
each of the thirty two integer registers is 64-bit.

Chapter 1
Table 1-1 summarizes the assembler’s usage and conventions and restrictions
for these registers. The assembler reserves all register names; you must use
lowercase for the names. All register names start with a dollar sign($).
The general registers have the names $0..$31. By including the file regdef.h
(use #include <regdef.h>) in your program, you can use software names for
some general registers. The operating system and the assembler use the
general registers $1, $26, $27, $28, and $29 for specific purposes.
NOTE: Attempts to use these general registers in other ways can produce
unexpected results.) If a program uses the names $1, $26, $27, $28, $29 rather
than the names $at, $kt0, $kt1, $gp, $sp respectively, the assembler issues
warning messages.

Assembly Language Programmer’s Guide 1-3

 Chapter 1

Table 1-1: General (Integer) Registers

Register Software Name

Chapter 1

Use and Linkage

Name (from regdef.h)
$0 Always has the value 0.
$at Reserved for the assembler.
Used for expression evaluations and to hold the integer type
$2..$3 v0-v1 function results. Also used to pass the static link when calling
nested procedures.
Used to pass the first 4 words of integer type actual
$4..$7 a0-a3 arguments, their values are not preserved across procedure
calls.
Temporary registers used for expression evaluations; their
$8..$15 t0-t7
values aren’t preserved across procedure calls.
Saved registers. Their values must be preserved across
$16..$23 s0-s7
procedure calls.
Temporary registers used for expression evaluations; their
$24..$25 t8-t9
values aren’t preserved across procedure calls.
$26..27 or
k0-k1 Reserved for the operating system kernel.
$kt0..$kt1
$28 or $gp gp Contains the global pointer.
$29 or $sp sp Contains the stack pointer.
Contains the frame pointer (if needed); otherwise a saved
$30 or $fp fp
register (like s0-s7).
Contains the return address and is used for expression
$31 ra
evaluation.

NOTE: General register $0 always contains the value 0. All other general
registers are equivalent, except that general register $31 also serves as the
implicit link register for jump and link instructions. See Chapter 7 for a
description of register assignments.

1-4 Assembly Language Programmer’s Guide

 Registers

Special Registers
The CPU defines three 32-bit special registers: PC (program counter), HI and
LO, as shown in Table 1-2. The HI and LO special registers hold the results

Chapter 1
of the multiplication (mult and multu) and division (div and divu)
instructions.
You usually do not need to refer explicitly to these special registers;
instructions that use the special registers refer to them automatically.
Table 1-2: Special Registers

Name Description
PC Program Counter
Multiply/Divide special register holds the most-
HI
significant 32 bits of multiply, remainder of divide
Multiply/Divide special register holds the least-
LO
significant 32 bits of multiply, quotient of divide

NOTE: In mips3 architecture, the HI and Lo registers hold 64-bits.

Assembly Language Programmer’s Guide 1-5

 Chapter 1

Floating-Point Registers
The FPU has sixteen floating-point registers. Each register can hold either a
single-precision (32 bit) or double-precision (64 bit) value. In case of a
Chapter 1

double-precision value, $f0 holds the least-significant half, and $f1 holds the
most-significant half. All references to these registers use an even register
number (e.g., $f4). Table 1-3 summarizes the assembler’s usage conventions
and restrictions for these registers.
Table 1-3: Floating-Point Registers

Register
Use and Linkage
Name
Used to hold floating-point type function results ($f0)
and complex type function results ($f0 has the real part,
$f0..$f2 $f2 has the imaginary part. $f4..$f10 Temporary
registers, used for expression evaluation, whose values
are not preserved across procedure calls.
Used to pass the first two single or double precision
$f12..$f14 actual arguments, whose values are not preserved
across procedure calls.
Temporary registers, used for expression evaluation,
$f16..$f18 whose values are not preserved across procedure
calls.
Saved registers, whose values must be preserved
$f20..$f30
across procedure calls.

1-6 Assembly Language Programmer’s Guide

Addressing

Chapter 2
This chapter describes the formats that you can use to specify addresses. The
machine uses a byte addressing scheme. Access to halfwords requires
alignment on even byte boundaries, and access to words requires alignment
on byte boundaries that are divisible by four. Any attempt to address a data
item that does not have the proper alignment causes an alignment exception.
The unaligned assembler load and store instructions may generate multiple
machine language instructions. They do not raise alignment exceptions.
These instructions load and store unaligned data:
• Load word left (lwl)
• Load word right (lwr)
• Store word left (swl)
• Store word right (swr)
• Unaligned load word (ulw)
• Unaligned load halfword (ulh)
• Unaligned load halfword unsigned (ulhu)
• Unaligned store word (usw)
• Unaligned store halfword (ush)
• These instructions load and store aligned data
• Load word (lw)
• Load halfword (lh)
• Load halfword unsigned (lhu)
• Load byte (lb)

Assembly Language Programmer’s Guide 2-1

 Chapter 2

• Load byte unsigned (lbu)

• Store word (sw)
• Store halfword (sh)
• Store byte (sb)

Address Formats
The assembler accepts these formats shown in Table 2-1 for addresses.
Table 2-1: Address Formats

Format Address
Chapter 2

(base register) Base address (zero Offset assumed)

expression Absolute address
expression (base register) Based address
relocatable-symbol Relocatable address
relocatable-symbol + expression Relocatable address
relocatable-symbol + expression
Indexed relocatable address
(index register)

2-2 Assembly Language Programmer’s Guide

 Addressing

Address Descriptions
The assembler accepts any combination of the constants and operations described
in this chapter for expressions in address descriptions.

Table 2-2: Assembler Addresses

Expression Address Description

Specifies an indexed address, which assumes a zero offset.
( base-register )
The base-register’s contents specify the address.
Specifies an absolute address. The assembler generates the
expression most locally efficient code for referencing a value at the
specified address.

Chapter 2
Specifies a based address. To get the address, the machine
expression (base-register) adds the value of the expression to the contents of the base-
register.
Specifies a relocatable address. The assembler generates the
relocatable-symbol necessary instruction(s) to addressx the item and generates
relocatable information for the link editor.
Specifies a relocatable address. To get the address, the
assembler adds or subtracts the value of the expression, which
has an absolute value, from the relocatable symbol. The
assembler generates the necessary instruction(s) to address
relocatable-symbol + expression
the item and generates relocatable information for the link
editor. If the symbol name does not appear as a label anywhere
in the assembly, the assembler assumes that the symbol is
external.
Specifies an indexed relocatable address. To get the address,
the machine adds the index-register to the relocatable symbol’s
address. The assembler generates the necessary instruction(s)
relocatable-symbol (index register) to address the item and generates relocatable information for
the link editor. If the symbol name does not appear as a label
anywhere in the assembly, the assembler assumes that the
symbol is external.
Specifies an indexed relocatable address. To get the address,
the assembler adds or subtracts the relocatable symbol, the
expression, and the contents of the index-register. The
relocatable + expression assembler generates the necessary instruction(s) to address
the item and generates relocation information for the link editor.
If the symbol does not appear as a label anywhere in the
assembly, the assembler assumes that the symbol is external.

Assembly Language Programmer’s Guide 2-3

 Chapter 2
Chapter 2

2-4 Assembly Language Programmer’s Guide

Exceptions

3
This chapter describes the exceptions that you can encounter while running
assembly programs. The machine detects some exceptions directly, and the
assembler inserts specific tests that signal other exceptions. This chapter lists
only those exceptions that occur frequently.

Main Processor Exceptions

Chapter 3
The following exceptions are the most common to the main processor:
• Address error exceptions, which occur when the machine references a
data item that is not on its proper memory alignment or when an
address is invalid for the executing process.
• Overflow exceptions, which occur when arithmetic operations
compute signed values and the destination lacks the precision to store
the result.
• Bus exceptions, which occur when an address is invalid for the
executing process.
• Divide-by-zero exceptions, which occur when a divisor is zero.

Assembly Language Programmer’s Guide 3-1

 Chapter 3

Floating-Point Exceptions
The following are the most common floating-point exceptions:
• Invalid operation exceptions which include:
– Magnitude subtraction of infinities, for example: ±-1.
– Multiplication of 0 by 1 with any signs.
– Division of 0/0 or 1/1 with any signs.
– Conversion of a binary floating-point number to an integer
format when an overflow or the operand value for the infinity or
NaN precludes a faithful representation in the format (see
Chapter 4).
– Comparison of predicates that have unordered operands, and that
involve Greater Than or Less Than without Unordered.
– Any operation on a signaling NaN.
• Divide-by-zero exceptions.
• Overflow exceptions—these occur when a rounded floating-point
result exceeds the destination format’s largest finite number.
• Underflow exceptions—these occur when a result has lost accuracy
Chapter 3

and also when a nonzero result is between 2Emin (±2 to the minimum
expressible exponent).
• Inexact exceptions.

3-2 Assembly Language Programmer’s Guide

Lexical Conventions

4
This chapter discusses lexical conventions for these topics:
• Tokens
• Comments
• Identifiers
• Constants
• Multiple lines per physical line
• Sections and location counters
• Statements
• Expressions
This chapter uses the following notation to describe syntax:
• | (vertical bar) means “or”

Chapter 4
• [ ] (square brackets) enclose options
• + indicates both addition and subtraction operations

Tokens
The assembler has these tokens:
• Identifiers
• Constants
• Operators

Assembly Language Programmer’s Guide 4-1

 Chapter 4

The assembler lets you put blank characters and tab characters anywhere
between tokens; however, it does not allow these characters within tokens
(except for character constants). A blank or tab must separate adjacent
identifiers or constants that are not otherwise separated.

Comments
The pound sign character (#) introduces a comment. Comments that start with
a # extend through the end of the line on which they appear. You can also use
C-language notation /*...*/ to delimit comments.
The assembler uses cpp (the C language preprocessor) to preprocess
assembler code. Because cpp interprets #s in the first column as pragmas
(compiler directives), do not start a # comment in the first column.

Identifiers
An identifier consists of a case-sensitive sequence of alphanumeric
characters, including these:
• . (period)
• _ (underscore)
• $ (dollar sign)
Identifiers can be up to 31 characters long, and the first character cannot be
numeric.
If an identifier is not defined to the assembler (only referenced), the assembler
assumes that the identifier is an external symbol. The assembler treats the
identifier like a .globl pseudo-operation (see Chapter 8). If the identifier is
defined to the assembler and the identifier has not been specified as global,
Chapter 4

the assembler assumes that the identifier is a local symbol.

Constants
The assembler has these constants:
• Scalar constants
• Floating point constants
• String constants

Scalar Constants
The assembler interprets all scalar constants as twos complement numbers.
Scalar constants can be any of the digits 0123456789abcdefABCDEF.

4-2 Assembly Language Programmer’s Guide

 Lexical Conventions

Scalar constants can be one of these constants:

• Decimal constants, which consist of a sequence of decimal digits
without a leading zero.
• Hexadecimal constants, which consist of the characters 0x (or 0X)
followed by a sequence of digits.
• Octal constants, which consist of a leading zero followed by a
sequence of digits in the range 0..7.

Floating Point Constants

Floating point constants can appear only in .float and .double pseudo-
operations (directives), see Chapter 8, and in the floating point Load
Immediate instructions, see Chapter 6. Floating point constants have this
format:
+d1[.d2][e|E+d3]
Where:
• d1 is written as a decimal integer and denotes the integral part of the
floating point value.
• d2 is written as a decimal integer and denotes the fractional part of
the floating point value.
• d3 is written as a decimal integer and denotes a power of 10.
• The “+” symbol is optional.
For example:
21.73E–3
represents the number .02173.

Chapter 4
.float and .double directives may optionally use hexadecimal floating point
constants instead of decimal ones. A hexadecimal floating point constant
consists of:
<+ or –> 0x <1 or 0 or nothing> . <hex digits> H 0x <hex
digits>
The assembler places the first set of hex digits (excluding the 0 or 1 preceding
the decimal point) in the mantissa field of the floating point format without
attempting to normalize it. It stores the second set of hex digits into the
exponent field without biasing them. It checks that the exponent is
appropriate if the mantissa appears to be denormalized. Hexadecimal
floating point constants are useful for generating IEEE special symbols, and
for writing hardware diagnostics.
For example, either of the following generates a single-precision “1.0”:

Assembly Language Programmer’s Guide 4-3

 Chapter 4

.float 1.0e+0
.float 0x1.0h0x7f

String Constants
String constants begin and end with double quotation marks (”).
The assembler observes C language backslash conventions. For octal
notation, the backslash conventions require three characters when the next
character could be confused with the octal number. For hexadecimal
notation, the backslash conventions require two characters when the next
character could be confused with the hexadecimal number (i.e., use a 0 for the
first character of a single character hex number).
The assembler follows the backslash conventions shown in Table 4-1.

Table 4-1: Backslash Conventions

Convention Meaning
\a Alert (0x07)
\b Backspace (0x08)
\f Form feed (0x0c)
\n Newline (0x0a)
\r Carriage return (0x0d)
\t horizontal tab (0x09)
\v Vertical feed (0x0b)
\\ Backslash (0x5c)
\" Quotation mark (0x22)
\’ Single quote (0x27)
\000 Character whose octal value is 000.
Chapter 4

\Xnn Character whose hexadecimal value is nn

Multiple Lines Per Physical Line

You can include multiple statements on the same line by separating the
statements with semicolons. The assembler does not recognize semicolons as
separators when they follow comment symbols (# or /*).

Sections and Location Counters

Assembled code and data fall in one of the sections shown in Figure 4-1.

4-4 Assembly Language Programmer’s Guide

 Lexical Conventions

.text Text section.

.rdata Read-only data section.

.data

.lit8 Data sections.

.lit4

.sdata Small data section, addressed

through register $gp.

.sbss Small bss section, addressed

through register $gp.

bss (block started by storage)

.bss section, which holds zero-initialized
data.

Figure 4-1: Section and location counters

(For more information on section data, see Chapter 9 of this manual.)

Chapter 4
The assembler always generates the text section before other sections.
Additions to the text section happen in four-byte units. Each section has an
implicit location counter, which begins at zero and increments by one for each
byte assembled in the section.
The bss section holds zero-initialized data. If a .lcomm pseudo-op defines a
variable (see Chapter 8), the assembler assigns that variable to the bss (block
started by storage) section or to the sbss (short block started by storage)
section depending on the variable’s size. The default variable size for sbss is
8 or fewer bytes.
The command line option –G for each compiler (C, Pascal, Fortran 77, or the
assembler), can increase the size of sbss to cover all but extremely large data
items. The link editor issues an error message when the –G value gets too

Assembly Language Programmer’s Guide 4-5

 Chapter 4

large. If a –G value is not specified to the compiler, 8 is the default. Items

smaller than, or equal to, the specified size go in sbss. Items greater than the
specified size go in bss.
Because you can address items much more quickly through $gp than through
a more general method, put as many items as possible in sdata or sbss. The
size of sdata and sbss combined must not exceed 64K bytes.

Statements
Each statement consists of an optional label, an operation code, and the
operand(s). The machine allows these statements:
• Null statements
• Keyword statements

Label Definitions
A label definition consists of an identifier followed by a colon. Label
definitions assign the current value and type of the location counter to the
name. An error results when the name is already defined, the assigned value
changes the label definition, or both conditions exists.
Label definitions always end with a colon. You can put a label definition on
a line by itself.
A generated label is a single numeric value (1...255). To reference a generated
label, put an f (forward) or a b (backward) immediately after the digit. The
reference tells the assembler to look for the nearest generated label that
corresponds to the number in the lexically forward or backward direction.
Chapter 4

Null Statements
A null statement is an empty statement that the assembler ignores. Null
statements can have label definitions. For example, this line has three null
statements in it:
label: ; ;

Keyword Statements
A keyword statement begins with a predefined keyword. The syntax for the
rest of the statement depends on the keyword. All instruction opcodes are
keywords. All other keywords are assembler pseudo-operations (directives).

4-6 Assembly Language Programmer’s Guide

 Lexical Conventions

Expressions
An expression is a sequence of symbols that represent a value. Each
expression and its result have data types. The assembler does arithmetic in
twos complement integers with 32 bits of precision. Expressions follow
precedence rules and consist of:
• Operators
• Identifiers
• Constants
Also, you may use a single character string in place of an integer within an
expression. Thus:
.byte “a” ; .word “a”+0x19
is equivalent to:
.byte 0x61 ; .word 0x7a

Precedence
Unless parentheses enforce precedence, the assembler evaluates all operators
of the same precedence strictly from left to right. Because parentheses also
designate index-registers, ambiguity can arise from parentheses in
expressions. To resolve this ambiguity, put a unary + in front of parentheses
in expressions.
The assembler has three precedence levels, which are listed here from lowest
to highest precedence:

least binding,
lowest precedence: binary +. -
..

Chapter 4
. binary *, /, %, <<, >>, ^, &, |
most binding
highest precedence: unary -, +, ~

NOTE: The assembler’s precedence scheme differs from that of the C

language.

Assembly Language Programmer’s Guide 4-7

 Chapter 4

Expression Operators
For expressions, you can rely on the precedence rules, or you can group
expressions with parentheses. The assembler has the operators listed in Table
4-2.

Table 4-2: Expression Operators

Operator Meaning
+ Addition
- Subtraction
* Multiplication
/ Division
% Remainder
<< Shift Left
>> Shift Right (sign NOT extended)
^ Bitwise Exclusive-OR
& Bitwise AND
| Bitwise OR
- Minus (unary)
+ Identity (unary)
~ Complement

Data Types
The assembler manipulates several types of expressions. Each symbol you
reference or define belongs to one of the categories shown in Table 4-3.
Chapter 4

Table 4-3: Data Types

Type Description
Any symbol that is referenced but not defined becomes global undefined,
and this module will attempt to import it. The assembler uses 32-bit
undefined
addressing to access these symbols. (Declaring such a symbol in a. globl
pseudo-op merely makes its status clearer).
A symbol defined by a .extern pseudo-op becomes global small undefined
if its size is greater than zero but less than the number of bytes specified by
sundefined the –G option on the command line (which defaults to 8). The linker places
these symbols within a 64k byte region pointed to by the $gp register, so
that the assembler can use economical 16-bit addressing to access them.

4-8 Assembly Language Programmer’s Guide

 Lexical Conventions

Table 4-3: Data Types

Type Description
absolute A constant defined in an “=” expression.
The text section contains the program’s instructions, which are not
text modifiable during execution. Any symbol defined while the .text pseudo-op
is in effect belongs to the text section.
The data section contains memory which the linker can initialize to nonzero
values before your program begins to execute. Any symbol defined while
data
the .data pseudo-op is in effect belongs to the data section. The assembler
uses 32-bit addressing to access these symbols.
This category is similar to data, except that defining a symbol while the
.sdata (“small data”) pseudo-op is in effect causes the linker to place it
sdata
within a 64k byte region pointed to by the $gp register, so that the
assembler can use economical 16-bit addressing to access it.
Any symbol defined while the .rdata pseudo-op is in effect belongs to this
rdata
category, which is similar to data, but may not be modified during execution.
The bss and sbss sections consist of memory which the kernel loader
initializes to zero before your program begins to execute. Any symbol
defined in a .comm or .lcomm pseudo-op belongs to these sections (except
that a .data, .sdata, or .rdata pseudo-op can override a .comm directive). If
its size is less than the number of bytes specified by the –G option on the
command line (which defaults to 8), it belongs to sbss (“small bss”), and the
bss and sbss linker places it within a 64k byte region pointed to by the $gp register so that
the assembler can use economical 16-bit addressing to access it.
Otherwise, it belongs to bss and the assembler uses 32-bit addressing.
Local symbols in bss or sbss defined by .lcomm are allocated memory by
the assembler; global symbols are allocated memory by the link editor; and
symbols defined by .comm are overlaid upon like-named symbols (in the

Chapter 4
fashion of Fortran “COMMON” blocks) by the link editor.

Symbols in the undefined and small undefined categories are always global
(that is, they are visible to the link editor and can be shared with other
modules of your program). Symbols in the absolute, text, data, sdata, rdata,
bss, and sbss categories are local unless declared in a .globl pseudo-op.

Type Propagation in Expressions

When expression operators combine expression operands, the result’s type
depends on the types of the operands and on the operator. Expressions follow
these type propagation rules:
• If an operand is undefined, the result is undefined.

Assembly Language Programmer’s Guide 4-9

 Chapter 4

• If both operands are absolute, the result is absolute.

• If the operator is + and the first operand refers to a relocatable text-
section, data-section, bss-section, or an undefined external, the result
has the postulated type and the other operand must be absolute.
• If the operator is – and the first operand refers to a relocatable text-
section, data-section, or bss-section symbol, the second operand can
be absolute (if it previously defined) and the result has the first
operand’s type; or the second operand can have the same type as the
first operand and the result is absolute. If the first operand is external
undefined, the second operand must be absolute.
• The operators * , /, % , << , >> , ~, ^ , & , and | apply only to
absolute symbols.
Chapter 4

4-10 Assembly Language Programmer’s Guide

Instruction Set

5
This chapter describes instruction notation and discusses assembler
instructions for the main processor. Chapter 6 describes coprocessor notation
and instructions.

Instruction Classes
The assembler has these classes of instructions for the main processor:
• Load and Store Instructions. These instructions load immediate
values and move data between memory and general registers.
• Computational Instructions. These instructions do arithmetic and
logical operations for values in registers.
• Jump and Branch Instructions. These instructions change program
control flow.
• Coprocessor Interface. These instructions provide standard
interfaces to the coprocessors.
• Special Instructions.These instructions do miscellaneous tasks.

Chapter 5

Assembly Language Programmer’s Guide 5-1

 Chapter 5

Reorganization Constraints and Rules

To maximize performance, the goal of RISC designs is to achieve an
execution rate of one machine cycle per instruction. In writing assembly
language instructions, you must be aware of the rules to achieve this goal.
This information is given in MIPS RISC Architecture (published by Prentice
Hall). You should refer to the following sections in this book for more
information:
Chapter Section Title
1 Cycles/Instruction
1 Instruction Pipelines
1 Instruction Operation Time
1 Instruction Access Time
3 The Delayed Instruction Slot
3 Delayed Loads
3 Delayed Jumps and Branches
C Filling the Branch Delay Slot
Refer also to Table 8-6 “Floating-Point Operation Latencies” in Chapter 8 of
the same book.

Instruction Notation
The tables in this chapter list the assembler format for each load, store,
computational, jump, branch, coprocessor, and special instruction. The
format consists of an op-code and a list of operand formats. The tables list
groups of closely related instructions; for those instructions, you can use any
op-code with any specified operand.
Operands can take any of these formats:
• Memory references. For example, a relocatable symbol +/– an
expression(register).
• Expressions (for immediate values).
• Two or three operands. For example, add $3,$4 is the same as add
$3,$3,$4.
Chapter 5

5-2 Assembly Language Programmer’s Guide

 Instruction Set

Load and Store Instructions

The machine has general-purpose load and store instructions.

Load and Store Formats

The operands in Table 5-1 have the following meanings:

Operand Description
destination Destination register
address Symbolic expression (see Chapter2)
source Source register
expression Absolute value

Table 5-1: Load and Store Formats

Description Op-code Operands
Load Address la destination, address
Load Byte lb
Load Byte Unsigned lbu
Load Halfword lh
Load Halfword Unsigned lhu
Load Linked* ll
Load Word lw
Load Word Left lwl
Load Word Right lwr
Load Double ld
Unaligned Load Halfword ulh
Unaligned Load Halfword
ulhu
Unsigned
Unaligned Load Word ulw
Load Immediate li destination, expression
Load Upper Immediate lui
Store Byte sb source, address
Store Conditional * sc
Store Double sd
Store Halfword sh
Store Word Left swl
Store Word Right swr
Chapter 5

Store Word sw
Unaligned Store Halfword ush
Unaligned Store Word usw
* Not valid in mips1 architectures

Assembly Language Programmer’s Guide 5-3

 Chapter 5

Table 5-2: Load and Store Formats for mips3 Architecture Only
Description Op-code Operands
Load Doubleword ld destination, address
Load Linked Doubleword lld
Load Word Unsigned lwu
Load Doubleword Left ldl
Load Doubleword Right ldr
Unaligned Load Double uld
Store Doubleword sd source, address
Store Conditional
scd
Doubleword
Store Double Left sdl
Store Double Right sdr
Unaligned Store Doubleword usd

Load Instruction Descriptions

For all machine load instructions, the effective address is the 32-bit twos-
complement sum of the contents of the index-register and the (sign-extended)
16-bit offset. Instructions that have symbolic labels imply an index-register,
which the assembler determines. The assembler supports additional load
instructions, which can produce multiple machine instructions.
Note: Load instructions can generate many code sequences for which the link
editor must fix the address by resolving external data items.

Table 5-3: Load Instruction Descriptions

Instruction Name Description

Load Address (la) Loads the destination register with the effective address of the specified data
item.
Load Byte (lb) Loads the least-significant byte of the destination register with the contents of
the byte that is at the memory location specified by the effective address. The
machine treats the loaded byte as a signed value: bit seven is extended to fill
the three most-significant bytes.
Load Byte Loads the least-significant byte of the destination register with the contents of
Unsigned (lbu) the byte that is at the memory location specified by the effective address.
Because the machine treats the loaded byte as an unsigned value, it fills the
three most-significant bytes of the destination register with zeros.
Chapter 5

Load Halfword (lh) Loads the two least-significant bytes of the destination register with the contents
of the halfword that is at the memory location specified by the effective address.
The machine treats the loaded halfword as a signed value. If the effective
address is not even, the machine signals an address error exception.

5-4 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-3: Load Instruction Descriptions

Instruction Name Description

Load Halfword Loads the least-significant bits of the destination register with the contents of the
Unsigned (lhu) halfword that is at the memory location specified by the effective address.
Because the machine treats the loaded halfword as an unsigned value, it fills
the two most-significant bytes of the destination register with zeros. If the
effective address is not even, the machine signals an address error exception.
Load Linked (li) * Loads the destination register with the contents of the word that is at the
memory location. This instruction implicitly performs a SYNC operation; all
loads and stores to shared memory fetched prior to the ll must access memory
before the ll, and loads and stores to shared memory fetched subsequent to the
ll must access memory after the ll. Load Linked and Store Conditional can be
use to automatically update memory locations. *This instruction is not valid in
the mips1 architectures. The machine signals an address exception when the
effective address is not divisible by four.
Load Word (lw) Loads the destination register with the contents of the word that is at the
memory location. The machine replaces all bytes of the register with the
contents of the loaded word. The machine signals an address error exception
when the effective address is not divisible by four.
Load Word Left (lwl) Loads the sign; that is, Load Word Left loads the destination register with the
most-significant bytes of the word specified by the effective address. The
effective address must specify the byte containing the sign. In a big-endian
machine, the effective address specifies the lowest numbered byte, and in a
little-endian machine the effective address specifies the highest numbered byte.
Only the bytes which share the same aligned word in memory are merged into
the destination register.
Load Word Right Loads the lowest precision bytes; that is, Load Word Right loads the destination
(lwr) register with the least-significant bytes of the word specified by the effective
address. The effective address must specify the byte containing the least-
significant bits. In a big-endian machine, the effective address specifies the
highest numbered byte, and in a little-endian machine the effective address
specifies the lowest numbered byte. Only the bytes which share the same
aligned word in memory are merged into the destination register.
Load Doubleword ld is a machine instruction in the mips3 architecture. For the -mips1 [default] and
(ld) -mips2 option: Loads the register pair (destination and destination +1) with the
two successive words specified by the address. The destination register must
be the even register of the pair. When the address is not on a word boundary,
the machine signals an address error exception. Note: This is retained for use
with the -mips1 and -mips2 options to provide backward compatibility only.
Unaligned Load Loads a halfword into the destination register from the specified address and
Chapter 5

Halfword (ulh) extends the sign of the halfword. Unaligned Load Halfword loads a halfword
regardless of the halfword’s alignment in memory.
Unaligned Load Loads a halfword into the destination register from the specified address and
Halfword Unsigned zero extends the halfword. Unaligned Load Halfword Unsigned loads a halfword
(ulhu) regardless of the halfword’s alignment in memory.

Assembly Language Programmer’s Guide 5-5

 Chapter 5

Table 5-3: Load Instruction Descriptions

Instruction Name Description

Unaligned Load Loads a word into the destination register from the specified address. Unaligned
Word (ulw) Load Word loads a word regardless of the word’s alignment in memory.
Load Immediate (li) Loads the destination register with the value of an expression that can be
computed at assembly time. Note: Load Immediate can generate any efficient
code sequence to put a desired value in the register.
Load Upper Loads the most-significant half of a register with the expression’s value, The
Immediate (lui) machine fills the least-significant half of the register with zeros. The
expression’s value must be in the range –32768...65535.

Table 5-4: Load Instruction Descriptions for mips3 Architecture Only

Instruction Name Description

Load Doubleword Loads the destination register with the contents of the double word that is at the
(ld) memory location. The machine replaces all bytes of the register with the
contents of the loaded double word. The machine signals an address error
exception when the effective address is not divisible by eight.
Load Linked Loads the destination register with the contents of the doubleword that is
Doubleword (lld) currently in the memory location. This instruction implicitly performs a SYNC
operation. Load Linked Doubleword and Store Conditional Doubleword can be
used to atomically update memory locations.
Load Word Loads the least-significant bits of the destination register with the contents of the
Unsigned (lwu) word (32 bits) that is at the memory location specified by the effective address.
Because the machine treats the loaded word as an unsigned value, it fills the
four most-significant bytes of the destination register with zeros. If the effective
address is not divisible by four, the machine signals an address error exception.
Load Doubleword Loads the destination register with the most-significant bytes of the doubleword
Left (ldl) specified by the effective address. The effective address must specify the byte
containing the sign. In a big-endian machine, the effective address specifies the
lowest numbered byte, and in a little-endian machine, the effective address
specifies the highest numbered byte. Only the bytes which share the same
aligned doubleword in memory are merged into the destination register.
Load Doubleword Loads the destination register with the least-significant bytes of the doubleword
Right (ldr) specified by the effective address. The effective address must specify the byte
containing the least-significant bits. In a bid-endian machine, the effective
address specifies the highest numbered byte. In a little-endian machine, the
Chapter 5

effective address specifies the lowest numbered byte. Only the bytes which
share the same aligned doubleword in memory are merged into the destination
register.
Unaligned Load Loads a doubleword into the destination register from the specified address. uld
Doubleword (uld) loads a doubleword regardless of the doubleword’s alignment in memory.

5-6 Assembly Language Programmer’s Guide

 Instruction Set

Store Instruction Descriptions

For all machine store instructions, the effective address is the 32-bit twos-
complement sum of the contents of the index-register and the (sign-extended)
16-bit offset. The assembler supports additional store instructions, which can
produce multiple machine instructions. Instructions that have symbolic labels
imply an index-register, which the assembler determines.

Table 5-5: Store Instruction Descriptions

Instruction Name Description

Store Byte (sb) Stores the contents of the source register’s least-significant byte in the byte
specified by the effective address.
Store Conditional Stores the contents of a word from the source register into the memory location
(sc) specified by the effective address. This instruction implicitly performs a SYNC
operation; all loads and stores to shared memory fetched prior to the sc must
access memory before the sc, and loads and stores to shared memory fetched
subsequent to the sc must access memory after the sc. If any other processor
or device has modified the physical address since the time of the previous Load
Linked instruction, or if an RFE or ERET instruction occurs between the Load
Linked and this store instruction, the store fails. The success or failure of the
store operation (as defined above) is indicated by the contents of the source
register after execution of the instruction. A successful store sets it to 1; and a
failed store sets it to 0. This instruction is not valid in the mips1 architectures.
The machine signals an address exception when the effective address is not
divisible by four.
Store Doubleword sd is a machine instruction in the mips3 architecture. For the -mips1 [default]
(sd) and -mips2 options: Stores the contents of the register pair in successive words,
which the address specifies. The source register must be the even register of
the pair, and the storage address must be word aligned. Note: This is retained
for use with the -mips1 and -mips2 options to provide backward compatibility
only.
Store Halfword (sh) Stores the two least-significant bytes of the source register in the halfword that
is at the memory location specified by the effective address. The effective
address must be divisible by two, otherwise the machine signals an address
error exception.
Store Word Left Stores the most-significant bytes of a word in the memory location specified by
Chapter 5

(swl) the effective address. The contents of the word at the memory location,
specified by the effective address, are shifted right so that the leftmost byte of
the unaligned word is in the addressed byte position. The stored bytes replace
the corresponding bytes of the effective address. The effective address’s last
two bits determine how many bytes are involved.

Assembly Language Programmer’s Guide 5-7

 Chapter 5

Table 5-5: Store Instruction Descriptions

Instruction Name Description

Store Word Right Stores the least-significant bytes of a word in the memory location specified by
(swr) the effective address. The contents of the word at the memory location,
specified by the effective address, are shifted left so that the right byte of the
unaligned word is in the addressed byte position. The stored bytes replace the
corresponding bytes of the effective address. The effective address’s last two
bits determine how many bytes are involved.
Store Word (sw) Stores the contents of a word from the source register in the memory location
specified by the effective address. The effective address must be divisible by
four, otherwise the machine signals an address error exception.
Unaligned Store Stores the contents of the two least-significant bytes of the
Halfword (ush) source register in a halfword that the address specifies. The machine does not
require alignment for the storage address.
Unaligned Store Stores the contents of the source register in a word specified by
Word (usw) the address. The machine does not require alignment for the storage address.
Chapter 5

5-8 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-6: Store Instruction Descriptions for mips3 Architecture Only

Instruction Name Description

Stores the contents of a doubleword from the source register in the memory
Store Doubleword
location specified by the effective address. The effective address must be
(sd)
divisible by eight, otherwise the machine signals an address error exception.
Stores the contents of a doubleword from the source register into the memory
locations specified by the effective address. This instruction implicitly performs a
SYNC operation. If any other processor or device has modified the physical
address since the time of the previous Load Linked instruction, or if an ERET
Store Conditional instruction occurs between the Load Linked instruction and this store
Doubleword (scd) instruction, the store fails and is inhibited from taking place. The success or
failure of the store operation (as defined above) is indicated by the contents of
the source register after execution of this instruction. A successful store sets it
to 1; and a failed store sets it to 0. The machine signals an address exception
when the effective address is not divisible by eight.
Stores the most-significant bytes of a doubleword in the memory location
Store Doubleword
specified by the effective address. It alters only the doubleword in memory
Left (sdl)
which contains the byte indicated by the effective address.
Stores the least-significant bytes of a doubleword in the memory location
Store Doubleword
specified by the effective address. It alters only the doubleword in memory
Right (sdr)
which contains the byte indicated by the effective address.
Unaligned Store Stores the contents of the source register in a doubleword specified by the
Doubleword (usd) address. The machine does not require alignment for the storage address.

Chapter 5

Assembly Language Programmer’s Guide 5-9

 Chapter 5

Computational Instructions
The machine has general-purpose and coprocessor-specific computational
instructions (for example, the floating-point coprocessor). This part of the
book describes general-purpose computational instructions.

Computational Formats
In the Table 5-7, operands have the following meanings:

Operand Description
destination/src1 Destination register is also source register 1
destination Destination register
immediate Immediate value
src1,src2 Source registers

Table 5-7: Computational Instruction Formats

Description Op-code Operand
Add with Overflow add destination, src1, src2
Add without Overflow addu destination, src1, src2
AND and destination, src1, immediate
Divide Signed div destination/src1, immediate
Divide Unsigned divu
Exclusive-OR xor
Multiply mul
Multiply with Overflow mulo
Multiply with Overflow Unsigned mulou
NOT OR nor
OR or
Set Equal seq
Set Greater sgt
Set Greater/Equal sge
Set Greater/Equal Unsigned sgeu
Set Greater Unsigned sgtu
Set Less slt
Set Less/Equal sle
Set Less/Equal Unsigned sleu
Set Less Unsigned sltu
Chapter 5

Set Not Equal sne

Subtract with Overflow sub
Subtract without Overflow subu
Remainder Signed rem
Remainder Unsigned remu

5-10 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-7: Computational Instruction Formats

Description Op-code Operand
Rotate Left rol
Rotate Right ror
Shift Right Arithmetic sra
Shift Left Logical sll
Shift Right Logical srl
Absolute Value abs destination,src1
Negate with Overflow neg destination/src1
Negate without Overflow negu
NOT not
Move move destination,src1
Multiply mult src1,src2
Multiply Unsigned multu
Trap if Equal teq src1, src2
Trap if not Equal tne src1, immediate
Trap if Less Than tlt
Trap if Less than, Unsigned tltu
Trap if Greater Than or Equal tge
Trap if Greater than or Equal, tgeu
Unsigned

Chapter 5

Assembly Language Programmer’s Guide 5-11

 Chapter 5

Table 5-8: Computational Instruction Formats for mips3 Architecture Only

Description Op-code Operand
Doubleword Add with Overflow dadd destination,src1, src2
destination/src1,src2
Doubleword Add without Overflow daddu destination,src1, immediate
destination/src1, immediate
Doubleword Divide Signed ddiv
Doubleword Divide Unsigned ddivu
Doubleword Multiply dmul
Doubleword Multiply with dmulo
Overflow
Doubleword Multiply with dmulou
Overflow Unsigned
Doubleword Subtract with dsub
Overflow
Doubleword Subtract without dsubu
Overflow
Doubleword Remainder Signed drem
Doubleword Remainder Unsigned dremu
Doubleword Rotate Left drol
Doubleword Rotate Right dror
Doubleword Shift Right Arithmetic dsra
Doubleword Shift Left Logical dsll
Doubleword Shift Right Logical dsrl
Doubleword Absolute Value dabs destination,src1
Doubleword Negate with dneg destination/src1
Overflow
Doubleword Negate without dnegu
Overflow
Doubleword Multiply dmult src1, src2
Doubleword Multiply Unsigned dmultu src1, immediate
Chapter 5

5-12 Assembly Language Programmer’s Guide

 Instruction Set

Computational Instruction Descriptions

Table 5-9: Computational Instruction Descriptions

Instruction Name Description

Absolute Value Computes the absolute value of the contents of src1 and puts the result in the
(abs) destination register. If the value in src1 is –2147483648, the machine signals an
overflow exception.
Add with Overflow Computes the twos complement sum of two signed values. This instruction
(add) adds the contents of src1 to the contents of src2, or it can add the contents of
src1 to the immediate value. Add (with overflow) puts the result in the
destination register. When the result cannot be extended as a 32-bit number,
the machine signals an overflow exception.
Add without Computes the twos complement sum of two 32-bit values. This instruction adds
Overflow (addu) the contents of src1 to the contents of src2, or it can add the contents of src1 to
the immediate value. Add (without overflow) puts the result in the destination
register. Overflow exceptions never occur.
AND (and) Computes the Logical AND of two values. This instruction ANDs (bit-wise) the
contents of src1 with the contents of src2, or it can AND the contents of src1
with the immediate value. The immediate value is not sign extended. AND puts
the result in the destination register.
Divide Signed (div) Computes the quotient of two values. Divide (with overflow) treats src1 as the
dividend. The divisor can be src2 or the immediate value. The instruction
divides the contents of src1 by the contents of src2, or it can divide src1 by the
immediate value. It puts the quotient in the destination register. If the divisor is
zero, the machine signals an error and may issue a break instruction. The div
instruction rounds toward zero. Overflow is signaled when dividing –
2147483648 by –1. The machine may issue a break instruction for divide-by-
zero or for overflow. Note: The special case div $0,src1,src2 generates the real
machine divide instruction and leaves the result in the hi/lo register. The hi
register contains the remainder and the lo register contains the quotient. No
checking for divide-by-zero is performed.
Divide Unsigned Computes the quotient of two unsigned 32-bit values. Divide (unsigned) treats
(divu) src1 as the dividend. The divisor can be src2 or the immediate value. This
instruction divides the contents of src1 by the contents of src2, or it can divide
the contents of src1 by the immediate value. Divide (unsigned) puts the quotient
in the destination register. If the divisor is zero, the machine signals an
exception and may issue a break instruction. See the note for div concerning $0
as a destination. Overflow exceptions never occur.
Chapter 5

Exclusive-OR (xor) Computes the XOR of two values. This instruction XORs (bit-wise) the contents
of src1 with the contents of src2, or it can XOR the contents of src1 with the
immediate value. The immediate value is not sign extended. Exclusive-OR puts
the result in the destination register.
Move (move) Moves the contents of src1 to the destination register.

Assembly Language Programmer’s Guide 5-13

 Chapter 5

Table 5-9: Computational Instruction Descriptions

Instruction Name Description

Multiply (mul) Computes the product of two values. This instruction puts the 32-bit product of
src1 and src2, or the 32-bit product of src1 and the immediate value, in the
destination register. The machine does not report overflow. Note: Use mul
when you do not need overflow protection: it’s often faster than mulo and mulou.
For multiplication by a constant, the mul instruction produces faster machine
instruction sequences than mult or multu instructions can produce.
Multiply (mult) Computes the 64-bit product of two 32-bit signed values. This instruction
multiplies the contents of src1 by the contents of src2 and puts the result in the
hi and lo registers (see Chapter 1). No overflow is possible. Note: The mult
instruction is a real machine language instruction
Multiply Unsigned Computes the product of two unsigned 32-bit values. It multiplies the contents of
(multu) src1 and the contents of src2 and puts the result in the hi and lo registers (see
Chapter 1). No overflow is possible. Note: The multu instruction is a real
machine language instruction.
Multiply with Computes the product of two 32-bit signed values. Multiply with Overflow puts
Overflow (mulo) the 32-bit product of src1 and src2, or the 32-bit product of src1 and the
immediate value, in the destination register. When a overflow occurs, the
machine signals an overflow exception and may execute a break instruction.
Note: For multiplication by a constant, mulo produces faster machine instruction
sequences than mult or multu can produce; however, if you do not need
overflow detection, use the mul instruction. It’s often faster than mulo.
Multiply with Computes the product of two 32-bit unsigned values. Multiply with Overflow
Overflow Unsigned Unsigned puts the 32-bit product of src1 and src2, or the product of src1 and the
(mulou) immediate value, in the destination register. This instruction treats the multiplier
and multiplicand as 32-bit unsigned values. When an overflow occurs, the
machine signals an overflow exception and may issue an break instruction.
Note: For multiplication by a constant, mulou produces faster machine
instruction sequences than mult or multu can reproduce; however, if you do not
need overflow detection, use the mul instruction. It’s often faster than mulou.

Negate with Computes the negative of a value. This instruction negates the contents of src1
Overflow (neg) and puts the result in the destination register. If the value in src1 is –
2147483648, the machine signals an overflow exception.
Negate without Negates the integer contents of src1 and puts the result in the destination
Overflow (negu) register. The machine does not report overflows.
NOT (not) Computes the Logical NOT of a value. This instruction complements (bit-wise)
the contents of src1 and puts the result in the destination register.
Chapter 5

NOT OR (nor) Computes the NOT OR of two values. This instruction combines the contents of
src1 with the contents of src2 (or the immediate value). NOT OR complements
the result and puts it in the destination register.

5-14 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-9: Computational Instruction Descriptions

Instruction Name Description

OR (or) Computes the Logical OR of two values. This instruction ORs (bit-wise) the
contents of src1 with the contents of src2, or it can OR the contents of src1 with
the immediate value. The immediate value is not sign extended. OR puts the
result in the destination register.
Remainder Signed Computes the remainder of the division of two unsigned 32-bit values. The
(rem) machine defines the remainder rem(i,j) as i–(j*div(i,j)) where j · 0. Remainder
(with overflow) treats src1 as the dividend. The divisor can be src2 or the
immediate value. This instruction divides the contents of src1 by the contents of
src2, or it can divide the contents of src1 by the immediate value. It puts the
remainder in the destination register. The rem instruction rounds toward zero,
rather than toward negative infinity. For example, div(5,–3)=–1, and rem(5,–
3)=2. For divide-by-zero, the machine signals an error and may issue a break
instruction.
Remainder Computes the remainder of the division of two unsigned 32-bit values. The
Unsigned (remu) machine defines the remainder rem(i,j) as i–(j*div(i,j)) where j · 0. Remainder
(unsigned) treats src1 as the dividend. The divisor can be src2 or the immediate
value. This instruction divides the contents of src1 by the contents of src2, or it
can divide the contents of src1 by the immediate value. Remainder (unsigned)
puts the remainder in the destination register. For divide-by-zero, the machine
signals an error and may issue a break instruction.
Rotate Left (rol) Rotates the contents of a register left (toward the sign bit). This instruction
inserts in the least-significant bit any bits that were shifted out of the sign bit.
The contents of src1 specify the value to shift, and the contents of src2 (or the
immediate value) specify the amount to shift. Rotate Left puts the result in the
destination register. If src2 (or the immediate value) is greater than 31, src1
shifts by (src2 MOD 32).
Rotate Right (ror) Rotates the contents of a register right (toward the least-significant bit). This
instruction inserts in the sign bit any bits that were shifted out of the least-
significant bit. The contents of src1 specify the value to shift, and the contents of
src2 (or the immediate value) specify the amount to shift. Rotate Right puts the
result in the destination register. If src2 (or the immediate value) is greater than
32, src1 shifts by src2 MOD 32.
Set Equal (seq) Compares two 32-bit values. If the contents of src1 equal the contents of src2
(or src1 equals the immediate value) this instruction sets the destination register
to one; otherwise, it sets the destination register to zero.
Set Greater (sgt) Compares two signed 32-bit values. If the contents of src1 are greater than the
contents of src2 (or src1 is greater than the immediate value), this instruction
sets the destination register to one; otherwise, it sets the destination register to
Chapter 5

zero.
Set Greater/Equal Compares two signed 32-bit values. If the contents of src1 are greater than or
(sge) equal to the contents of src2 (or src1 is greater than or equal to the immediate
value), this instruction sets the destination register to one; otherwise, it sets the
destination register to zero.

Assembly Language Programmer’s Guide 5-15

 Chapter 5

Table 5-9: Computational Instruction Descriptions

Instruction Name Description

Set Greater/Equal Compares two unsigned 32-bit values. If the contents of src1 are greater than or
Unsigned (sgeu) equal to the contents of src2 (or src1 is greater than or equal to the immediate
value), this instruction sets the destination register to one; otherwise, it sets the
destination register to zero.
Set Greater Compares two unsigned 32-bit values. If the contents of src1 are greater than
Unsigned (sgtu) the contents of src2 (or src1 is greater than the immediate value), this
instruction sets the destination register to one; otherwise, it sets the destination
register to zero.
Set Less (slt) Compares two signed 32-bit values. If the contents of src1 are less than the
contents of src2 (or src1 is less than the immediate value), this instruction sets
the destination register to one; otherwise, it sets the destination register to zero.
Set Less/Equal (sle) Compares two signed 32-bit values. If the contents of src1 are less than or
equal to the contents of src2 (or src1 is less than or equal to the immediate
value), this instruction sets the destination register to one; otherwise, it sets the
destination register to zero.
Set Less/Equal Compares two unsigned 32-bit values. If the contents of src1 are less than or
Unsigned (sleu) equal to the contents of src2 (or src1 is less than or equal to the immediate
value) this instruction sets the destination register to one; otherwise, it sets the
destination register to zero.
Set Less Unsigned Compares two unsigned 32-bit values. If the contents of src1 are less than the
(sltu) contents of src2 (or src1 is less than the immediate value), this instruction sets
the destination register to one; otherwise, it sets the destination register to zero.
Set Not Equal (sne) Compares two 32-bit values. If the contents of scr1 do not equal the contents of
src2 (or src1 does not equal the immediate value), this instruction sets the
destination register to one; otherwise, it sets the destination register to zero.
Shift Left Logical Shifts the contents of a register left (toward the sign bit) and inserts zeros at the
(sll) least-significant bit. The contents of src1 specify the value to shift, and the
contents of src2 or the immediate value specify the amount to shift. If src2 (or
the immediate value) is greater than 31 or less than 0, src1 shifts by src2 MOD
32.
Shift Right Shifts the contents of a register right (toward the least-significant bit) and inserts
Arithmetic (sra) the sign bit at the most-significant bit. The contents of src1 specify the value to
shift, and the contents of src2 (or the immediate value) specify the amount to
shift. If src2 (or the immediate value) is greater than 31 or less than 0, src1 shifts
by the result of src2 MOD 32.
Shift Right Logical Shifts the contents of a register right (toward the least-significant bit) and inserts
Chapter 5

(srl) zeros at the most-significant bit. The contents of src1 specify the value to shift,
and the contents of src2 (or the immediate value) specify the amount to shift. If
src2 (or the immediate value) is greater than 31 or less than 0, src1 shifts by the
result of src2 MOD 32.

5-16 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-9: Computational Instruction Descriptions

Instruction Name Description

Subtract with Computes the twos complement difference for two signed values. This
overflow (sub) instruction subtracts the contents of src2 from the contents of src1, or it can
subtract the contents of the immediate from the src1 value. Subtract puts the
result in the destination register. When the true result’s sign differs from the
destination register’s sign, the machine signals an overflow exception.
Subtract without Computes the twos complement difference for two 32-bit values. This
overflow (subu) instruction subtracts the contents of src2 from the contents of src1, or it can
subtract the contents of the immediate from the src1 value. Subtract (without
overflow) puts the result in the destination register. Overflow exceptions never
happen.
Trap if Equal (teq) Compares two 2-bit values. If the contents of src1 equal the contents of src2 (or
src1 equals the immediate value), a trap exception occurs.
Trap if not Equal Compares two 32-bit values. If the contents of src1 do not equal the contents of
(tne) src2 (or src1 does not equal the immediate value), a trap exception occurs.
Trap if Less Than Compares two signed 32-bit values. If the contents of src1 are less than the
(tlt) contents of src2 (or src1 is less than the immediate value), a trap exception
occurs.
Trap if Less Than Compares two unsigned 32-bit values. If the contents of src1 are less than the
Unsigned (tltu) contents of src2 (or src1 is less than the immediate value), a trap exception
occurs.
Trap if Greater than Compares two signed 32-bit values. If the contents of src1 are greater than the
or Equal (tge) contents of src2 (or src1 is greater than the immediate value), a trap exception
occurs.
Trap if Greater than Compares two unsigned 32-bit values. If the contents of src1 are greater than
or Equal Unsigned the contents of src2 (or src1 is greater than the immediate value), a trap
(tgeu) exception occurs.

Chapter 5

Assembly Language Programmer’s Guide 5-17

 Chapter 5

Table 5-10: Computational Instruction Descriptions for mips3

Architecture Only

Instruction Name Description

Doubleword Computes the absolute value of the contents of src1, treated as a 64-bit signed
Absolute Value value, and puts the result in the destination register. If the value in src1 is -
(dabs) 2**63, the machine signals an overflow exception.
Doubleword Add Computes the twos complement sum of two 64-bit signed values. The
with overflow (dadd) instruction adds the contents of src1 to the contents of src2, or it can add the
contents of src1 to the immediate value. When the result cannot be extended as
a 64-bit number, the machine signals an overflow exception.
Doubleword Add Computes the twos complement sum of two 64-bit values. The instruction adds
without overflow the contents of src1 to the contents of src2, or it can add the contents of src1 to
(daddu the immediate value. Overflow exceptions never occur.
Doubleword Divide Computes the quotient of two 64-bit values. ddiv treats src1 as the dividend.
Signed (ddiv) The divisor can be src2 or the immediate value. It puts the quotient in the
destination register. If the divisor is zero, the machine signals an error and may
issue a break instruction. The ddiv instruction rounds towards zero. Overflow is
signaled when dividing -2**63 by -1. Note: The special case ddiv $0,src1,src2
generates the real machine doubleword divide instruction and leaves the result
in the hi/lo register. The hi register contains the quotient. No checking for divide-
by-zero is performed.
Doubleword Divide Computes the quotient of two unsigned 64-bit values. ddivu treats src1 as the
Unsigned (ddivu) dividend. The divisor can be src2 or the immediate value. It puts the quotient in
the destination register. If the divisor is zero, the machine signals an exception
and may issue a break instruction. See note for ddiv concerning $0 as a
destination. Overflow exceptions never occur.
Doubleword Computes the product of two values. This instruction puts the 64-bit product of
Multiply (dmul) src1 and src2, or the 64-bit product of src1 and the immediate value, in the
destination register. The machine does not report overflow. Note: Use dmul
when you do not need overflow protection. It is often faster than dmulo and
dmulou. For multiplication by a constant, the dmul instruction produces faster
machine instruction sequences than dmult or dmultu can produce.
Doubleword Computes the 128-bit product of two 64-bit signed values. This instruction
Multiply (dmult) multiplies the contents of src1 by the contents of src2 and puts the result in the
hi and lo registers. No overflow is possible. Note: The dmult instruction is a real
machine language instruction.
Doubleword Computes the product of two unsigned 64-bit values. It multiplies the contents of
Multiply Unsigned src1 and the contents of src2, putting the result in the hi and lo registers. No
Chapter 5

(dmultu) overflow is possible. Note: The dmultu instruction is a real machine language
instruction.

5-18 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-10: Computational Instruction Descriptions for mips3

Architecture Only

Instruction Name Description

Doubleword Computes the product of two 64-bit signed values. It puts the 64-bit product of
Multiply with src1 and src2, or the 64-bit product of src1 and the immediate value, in the
Overflow (dmulo) destination register. When an overflow occurs, the machine signals an overflow
exception and may execute a break instruction. Note: For multiplication by a
constant, dmulo produces faster machine instruction sequences than dmult or
dmultu can produce; however, if you do not need overflow detection, use the
dmul instruction. It is often faster than dmulo.
Doubleword Computes the product of two 64-bit unsigned values. It puts the 64-bit product of
Multiply with src2 and src2, or the 64-bit product of src1 and the immediate value, into the
Overflow Unsigned destination register. When an overflow occurs, the machine signals an overflow
(dmulou) exception and may issue a break instruction. Note: For multiplication by a
constant, dmulou produces faster machine instruction sequences than dmult or
dmultu produces; however, if you do not need overflow detection, use the dmul
instruction. It is often faster than dmulou.
Doubleword Negate Computes the negative of a 64-bit value. The instruction negates the contents of
with Overflow (dneg) src1 and puts the result in the destination register. If the value of src1 is -2**63,
the machine signals an overflow exception.
Doubleword Negate
Negates the 64-bit contents of src1 and puts the result in the destination
without Overflow
register. The machine does not report overflow.
(dnegu)
Doubleword Computes the remainder of the division of two signed 64-bit values. It treats
Remainder Signed src1 as the dividend. The divisor can be src2 or the immediate value. The
(drem) dremu instruction puts the remainder in the destination register. If the divisor is
zero, the machine signals an error and may issue a break instruction.
Doubleword Computes the remainder of the division of two unsigned 64-bit values. It treats
Remainder src1 as the dividend. The divisor can be src2 or the immediate value. The
Unsigned (dremu) dremu instruction puts the remainder in the destination register. If the divisor is
zero, the machine signals an error and may issue a break instruction.
Doubleword Rotate Rotates the contents of a 64-bit register left (towards the sign bit). This
Left (drol) instruction inserts in the least-significant bit any bits that were shifted out of the
sign bit. The contents of src1 specify the value to shift, and contents of src2 (or
the immediate value) specify the amount to shift. If src2 (or the immediate value)
is greater than 63, src1 shifts by src2 MOD 64.
Doubleword Rotate Rotates the contents of a 63-bit register right (towards the least-significant bit).
Right (dror) This instruction inserts in the sign bit any bits that were shifted out of the least-
significant bit. The contents of src1 specify the value to shift, and the contents of
Chapter 5

src2 (or the immediate value) specify the amount to shift. If src2 (or the
immediate value is greater than 63, src1 shifts by src2 MOD 64.

Assembly Language Programmer’s Guide 5-19

 Chapter 5

Table 5-10: Computational Instruction Descriptions for mips3

Architecture Only

Instruction Name Description

Doubleword Shift Shifts the contents of a 64-bit register left (towards the sign bit) and inserts
Left Logical (dsll) zeros at the least-significant bit. The contents of src1 specify the value to shift,
and the contents of src2 (or the immediate value) specify the amount to shift. If
src2 (or the immediate value) is greater than 63, src1 shifts by src2 MOD 64.
Doubleword Shift Shifts the contents of a 64-bit register right (towards the least-significant bit) and
Right Arithmetic inserts the sign bit at the most-significant bit. The contents of src2 (or the
(dsra) immediate value) specify the amount to shift. If src2 (or the immediate value) is
greater than 63, src1 shifts by src2 MOD 64.
Doubleword Shift Shifts the contents of a 64-bit register right (towards the least-significant bit) and
Right Logical (dsrl) inserts zeros at the most-significant bit. The contents of src1 specify the value to
shift, and the contents of src2 (or the immediate value) specify the amount to
shift. If src2 (or the immediate value) is greater than 63, src1 shifts by src2 MOD
64.
Doubleword Computes the twos complement difference for two signed 64-bit values. This
Subtract with instruction subtracts the contents of src2 from the contents of src1, or it can
Overflow (dsub) subtract the immediate value from the contents of src1. It puts the result in the
destination register. When the true result’s sign differs from the destination
register’s sign, the machine signals an overflow exception.
Doubleword Computes the twos complement difference for two unsigned 64-bit values. This
Subtract without instruction subtracts the contents of src2 from the contents of src1, or it can
Overflow (dsub) subtract the immediate value from the contents of src1. It puts the result in the
destination register. Overflow exceptions never happen.
Chapter 5

5-20 Assembly Language Programmer’s Guide

 Instruction Set

Jump and Branch Instructions

The jump and branch instructions let you change an assembly program’s
control flow. This section of the book describes jump and branch instructions.

Jump and Branch Formats

In Table 5-11 below, the operands have the following meanings:

Operand Description
address An expression.
immediate An expression with an absolute value.
label A symbol label.
return Register containing the return address.
src1, src2 The source registers.
target Register containing the target.

Table 5-11: Jump and Branch Instruction Formats

Description Op-code Operand
Jump j address
Jump and Link jal address
target
return,target
Branch on Equal beq src1,src2,label
Branch on Greater bgt src1,immediate,label
Branch on Greater/Equal bge
Branch on Greater/Equal Unsigned bgeu
Branch on Greater Unsigned bgtu
Branch on Less blt
Branch on Less/Equal ble
Branch on Less/Equal Unsigned bleu
Branch on Less Unsigned bltu
Branch on Not Equal bne
Branch b label
Branch and Link bal
Branch on Equal Likely* beql src1,src2,label
Branch on Greater Likely* bgtl src1,
immediate,label
Branch on Greater/Equal Likely * bgel
Chapter 5

Branch on Greater/Equal Unsigned bgeul

Likely*
Branch on Greater Unsigned Likely* bgtul
Branch on Less Likely* bltl
Branch on Less/Equal Likely * blel

Assembly Language Programmer’s Guide 5-21

 Chapter 5

Table 5-11: Jump and Branch Instruction Formats

Description Op-code Operand
Branch on Less/Equal Unsigned Likely* bleul
Branch on Less Unsigned Likely* bltul
Branch on Not Equal Likely* bnel
Branch on Equal to Zero beqz src1,label
Branch on Greater/Equal Zero bgez
Branch on Greater Than Zero bgtz
Branch on Greater or Equal to Zero and bgezal
Link
Branch on Less Than Zero and Link bltzal
Branch on Less/Equal Zero blez
Branch on Less Than Zero bltz
Branch on Not Equal to Zero bnez
Branch on Equal to Zero Likely* beqzl src1,label
Branch on Greater/Equal Zero Likely* bgezl
Branch on Greater Than Zero Likely* bgtzl
Branch on Greater or Equal to Zero and bgezall
Link Likely*
Branch on Less Than Zero and Link bltzall
Likely*
Branch on Less/Equal Zero Likely* blezl
Branch on Less Than Zero Likely* bltzl
Branch on Not Equal to Zero Likely* bnezl
* Not valid in mips1 architecture.
Chapter 5

5-22 Assembly Language Programmer’s Guide

 Instruction Set

Jump and Branch Instruction Descriptions

In the following branch instructions, branch destinations must be defined in
the source being assembled.

Table 5-12: Jump and Branch Instruction Descriptions

Instruction Name Description

Branch (b) Branches unconditionally to the specified label.
Branch and Link Branches unconditionally to the specified label and puts the return address in
(bal) general register $31.
Branch on Equal Branches to the specified label when the contents of src1 equal the contents of
(beq) src2, or it can branch when the contents of src1 equal the immediate value.
Branch on Equal
Branches to the specified label when the contents of src1 equal zero.
to Zero (beqz)
Branch on Branches to the specified label when the contents of src1 are greater than the
Greater (bgt) contents of src2, or it can branch when the contents of src1 are greater than the
immediate value. The comparison treats the comparands as signed 32-bit values.
Branch on Branches to the specified label when the contents of src1 are greater than or
Greater/Equal equal to the contents of src2, or it can branch when the contents of src1 are
Unsigned (bgeu) greater than or equal to the immediate value. The comparison treats the
comparands as unsigned 32-bit values.
Branch on
Branches to the specified label when the contents of src1 are greater than or
Greater/Equal
equal to zero.
Zero (bgez)
Branch on Branches to the specified label when the contents of src1 are greater than or
Greater/Equal equal to zero and puts the return address in general register $31. When this write
Zero and Link is done, it destroys the contents of the register. See the MIPS RISC Architecture
(bgezal) book for more information. Do not use bgezal $31.
Branch on Branches to the specified label when the contents of src1 are greater than or
Greater or Equal equal to the contents of src2, or it can branch when the contents of src1 are
(bge) greater than or equal to the immediate value. The comparison treats the
comparands as signed 32-bit values.
Branch on Branches to the specified label when the contents of src1 are greater than the
Greater Than contents of src2, or it can branch when the contents of src1 are greater than the
Unsigned (bgtu) immediate value. The comparison treats the comparands as unsigned 32-bit values.
Branch on
Greater Than Branches to the specified label when the contents of src1 are greater than zero.
Zero (bgtz)
Chapter 5

Branch on Less Branches to the specified label when the contents of src1 are less than zero. The
Than Zero (bltz) program must define the destination.
Branch on Less Branches to the specified label when the contents of src1 are less than the
(blt) contents of src2, or it can branch when the contents of src1 are less than the
immediate value. The comparison treats the comparands as signed 32-bit values.

Assembly Language Programmer’s Guide 5-23

 Chapter 5

Table 5-12: Jump and Branch Instruction Descriptions

Instruction Name Description

Branch on Less/ Branches to the specified label when the contents of src1 are less than or equal to
Equal Unsigned the contents of src2, or it can branch when the contents of src1 are less than or
(bleu) equal to the immediate value. The comparison treats the comparands as
unsigned 32-bit values.
Branch on Less/ Branches to the specified label when the contents of src1 are less than or equal to
Equal Zero (blez) zero. The program must define the destination.
Branch on Less or Branches to the specified label when the contents of src1 are less than or equal to
Equal (ble) the contents of src2, or it can branch when the contents of src1 are less than or
equal to the immediate value. The comparison treats the comparands as signed
32-bit values.
Branch on Less Branches to the specified label when the contents of src1 are less than the
Than Unsigned contents of src2, or it can branch when the contents of src1 are less than the
(bltu) immediate value. The comparison treats the comparands as unsigned 32-bit
values.
Branch on Less Branches to the specified label when the contents of src1 are less than zero and
Than Zero and puts the return address in general register $31. Because the value is always
Link (bltzal) stored in register 31, there is a chance of a stored value being overwritten before
it is used. See the MIPS RISC Architecture book for more information. Do not use
bgezal $31
Branch on Not Branches to the specified label when the contents of src1 do not equal the
Equal (bne) contents of src2, or it can branch when the contents of src1 do not equal the
immediate value.
Branch on Not
Equal to Zero Branches to the specified label when the contents of src1 do not equal zero.
(bnez)
Jump (j) Unconditionally jumps to a specified location. A symbolic address or a general
register specifies the destination. The instruction j $31 returns from the a jal call
instruction.
Jump And Link Unconditionally jumps to a specified location and puts the return address in a
(jal) general register. A symbolic address or a general register specifies the target
location. By default, the return address is placed in register $31. If you specify a
pair of registers, the first receives the return address and the second specifies the
target. The instruction jal procname transfers to procname and saves the return
address. For the two-register form of the instruction, the target register may not
be the same as the return-address register. For the one-register form, the target
may not be $31.
Chapter 5

* Likely Same an the ordinary branch instruction (without the "Likely"), except in a branch
likely instruction, the instruction in the delay slot is nullified if the conditional
branch is not taken. Note: The branch likely instructions should be used only
inside a .set noreorder schedule in an assembly program. The assembler does
not attempt to schedule the delay slot of a branch likely instruction.

5-24 Assembly Language Programmer’s Guide

 Instruction Set

Special Instructions
The main processor’s special instructions do miscellaneous tasks.

Special Formats
In Table 5-13, operands have the following meanings:

Operand Description
register Destination or source register
breakcode Value that determines the break type

Table 5-13: Special Instruction Formats

Description Op-code Operand
Break break breakcode
breakcode1,breakcode2
Exception Return eret*
Restore From Exception rfe**
Syscall syscall
Move From HI Register mfhi register
Move To HI Register mthi
Move From LO Register mflo
Move To LO Register mtlo
* Not available in mips1 and mips2 architectures.
** Not available in R4000. Use the eret instruction instead.

Chapter 5

Assembly Language Programmer’s Guide 5-25

 Chapter 5

Special Instruction Descriptions

Table 5-14: Special Instruction Descriptions

Instruction Name Description

Break (break) Unconditionally transfers control to the exception handler. The breakcode
operand is interpreted by software conventions. The breakcode1 operand is
used to fill the low-order 10 bits of the 20 bit immediate field in the break
instruction. The optional second operand, breakcode2, fills the high-order 10
bits.
Exception Return Returns from an interrupt, exception or error trap. Similar to a branch or jump
(eret) instruction, eret executes the next instruction before taking effect. Use this on
R4000 processor machines in place of rfe.
Move From HI
Moves the contents of the hi register to a general-purpose register.
Register (mfhi)
Move From LO
Moves the contents of the lo register to a general-purpose register.
Register (mflo)
Move To HI Register
Moves the contents of a general-purpose register to the hi register.
(mthi)
Move To LO
Moves the contents of a general-purpose register to the lo register.
Register (mtlo)
Restore From Restores the previous interrupt called and user/kernel state. This instruction can
Exception (rfe) execute only in kernel state and is unavailable in user mode.
Syscall (syscall) Causes a system call trap. The operating system interprets the information set
in registers to determine what system call to do.
Chapter 5

5-26 Assembly Language Programmer’s Guide

 Instruction Set

Coprocessor Interface Instructions

The coprocessor interface instructions provide standard ways to access the
machine’s coprocessors.

Coprocessor Interface Formats

In Table 5-15, the operands have the following meanings:

Operand Description
address A symbolic expression
destination The destination coprocessor register
dest-gpr The destination general register
label A symbolic label
operation The coprocessor specific operation
source A coprocessor register from which values are assigned
src-gpr A general register from which values are assigned
z A coprocessor number in the range 0...2

Table 5-15: Coprocessor Interface Instruction Formats

Description Op-code Operand
Load Word Coprocessor z lwcz destination,address
Load Double Coprocessor z* ldcz
Store Word Coprocessor z swcz source, address
Store Double Coprocessor z* sdcz
Move From Coprocessor z mfcz dest-gpr, source
Move To Coprocessor z mtcz src-gpr, destination
Doubleword Move From Coprocessor z ** dmfcz
Doubleword Move To Coprocessor z ** dmtcz
Branch Coprocessor z False bczf label
Branch Coprocessor z True bczt
Branch Coprocessor z False Likely* bczfl
Branch Coprocessor z True Likely* bcztl
Coprocessor z Operation cz expression
Control From Coprocessor z cfcz dest-gpr, source
Control To Coprocessor z ctcz src-gpr, destination
Chapter 5

* Not valid in mips1 architectures.

** Not valid in mips1 and mips2 architectures.

NOTE: You cannot use coprocessor load and store instructions with the
system control coprocessor (cp0).

Assembly Language Programmer’s Guide 5-27

 Chapter 5

Coprocessor Interface Instruction Descriptions

Table 5-16: Coprocessor Interface Instruction Descriptions

Instruction Name Description

Branch Coprocessor Branches to the specified label when the specified coprocessor asserts a true
z True (bczt) condition. The z selects one of the coprocessors. A previous coprocessor
operation sets the condition.
Branch Coprocessor Branches to the specified label when the specified coprocessor asserts a false
z False (bczf) condition. The z selects one of the coprocessors. A previous coprocessor
operation sets the condition.
Branch Coprocessor Branches to the specified label when the specified coprocessor asserts a true
z True Likely (bcztl) condition. If the conditional branch is not taken, the instruction in the branch
delay slot is nullified. Note: The branch likely instructions should be used only
within a .set noreorder block. The assembler does not attempt to schedule the
delay slot of a branch likely instruction.
Branch Coprocessor Branches to the specified label when the specified coprocessor asserts a false
z False Likely (bczfl) condition. If the conditional branch is not taken, the instruction in the branch
delay slot is nullified. Note: The branch likely instructions should be used only
within a .set noreorder block. The assembler does not attempt to schedule the
delay slot of a branch likely instruction.
Control From Stores the contents of the coprocessor control register specified by the source
Coprocessor z (cfcz) in the general register specified by dest-gpr.
Control To Stores the contents of the general register specified by src-gpr in the
Coprocessor (ctcz) coprocessor control register specified by the destination.
Coprocessor z Executes a coprocessor-specific operation on the specified coprocessor. The z
Operation (cz) selects one of four distinct coprocessors.
Load Word Loads the destination with the contents of a word that is at the memory location
Coprocessor z specified by the effective address. The z selects one of four distinct
(lwcz) coprocessors. Load Word Coprocessor replaces all register bytes with the
contents of the loaded word. If bits 0 and 1 of the effective address are not zero,
the machine signals an address exception.
Load Double Loads a doubleword from the memory location specified by the effective
Coprocessor z (ldcz) address and makes the data available to coprocessor unit z. The manner in
which each coprocessor uses the data is defined by the individual coprocessor
specifications. This instruction is not valid in mips1 architectures. If any of the
three least-significant bits of the effective address are non-zero, the machine
signals an address error exception.
Chapter 5

Move From
Stores the contents of the coprocessor register specified by the source in the
Coprocessor z
general register specified by dest-gpr.
(mfcz)

5-28 Assembly Language Programmer’s Guide

 Instruction Set

Table 5-16: Coprocessor Interface Instruction Descriptions

Instruction Name Description

Move To
Stores the contents of the general register specified by src-gpr in the
Coprocessor z
coprocessor register specified by the destination.
(mtcz)
Doubleword Move
Stores the 64-bit contents of the coprocessor register specified by the source
From Coprocessor
into the general register specified by dest-gpr.
z (dmfcz)
Doubleword Move
Stores the 64-bit contents of the general register src-gpr into the coprocessor
To Coprocessor z
register specified by the destination.
(dmtcz)
Store Word Stores the contents of the coprocessor register in the memory location specified
Coprocessor z by the effective address. The z selects one of four distinct coprocessors. If bits 0
(swcz) and 1 of the effective address are not zero, the machine signals an address
error exception.
Store Double Coprocessor z sources a doubleword, which the processor writes the memory
Coprocessor z location specified by the effective address. The data to be stored is defined by
(sdcz) the individual coprocessor specifications. This instruction is not valid in mips1
architecture. If any of the three least-significant bits of the effective address are
non-zero, the machine signals an address error exception.

Chapter 5

Assembly Language Programmer’s Guide 5-29

 Chapter 5
Chapter 5

5-30 Assembly Language Programmer’s Guide

Coprocessor Instruction Set

Chapter 6
6
This chapter describes the coprocessor instructions for these coprocessors:
• System control coprocessor (cp0) instructions
• Floating-point coprocessor instructions

See Chapter 5 for a description of the main processor’s instructions and the
coprocessor interface instructions.

Instruction Notation
The tables in this chapter list the assembler format for each coprocessor’s
load, store, computational, jump, branch, and special instructions. The format
consists of an op-code and a list of operand formats. The tables list groups of
closely related instructions; for those instructions, you can use any op-code
with any specified operand.
NOTE: The system control coprocessor instructions do not have operands.
Operands can have any of these formats:
• Memory references: for example, a relocatable symbol +/– an
expression(register)
• Expressions (for immediate values)
• Two or three operands: for example, add $3,$4 is the same as add
$3,$3,$4

Assembly Language Programmer’s Guide 6-1

 Chapter 6

The following terms are used to discuss floating-point operations:

• infinite: A value of +1 or –1.
• infinity: A symbolic entity that represents values with magnitudes
Chapter 6

greater than the largest value in that format.

• ordered: The usual result from a comparison, namely: <,=, or >.
• NaN: Symbolic entities that represent values not otherwise available
in floating-point formats. There are two kinds of NaNs. Quiet NaNs
represent unknown or uninitialized values. Signaling NaNs represent
symbolic values and values that are too big or too precise for the
format. Signaling NaNs raise an invalid operation exception
whenever an operation is attempted on them.
• unordered: The condition that results from a floating-point
comparison when one or both operands are NaNs.

Floating-Point Instructions
The floating-point coprocessor has these classes of instructions:
• Load and Store Instructions: Load values and move data between
memory and coprocessor registers.
• Move Instructions: Move data between registers.
• Computational Instructions: Do arithmetic and logical operations
on values in coprocessor registers.
• Relational Instructions: Compare two floating-point values.
A particular floating-point instruction may be implemented in hardware,
software, or a combination of hardware and software.

6-2 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Floating-Point Formats
The formats for the single- and double-precision floating-point constants are
shown below:

Chapter 6
0 1 89 31 (big-endian)

1 8 bits 23 bits Single-Precision

31 30 23 22 0 (little-endian)
(big-endian)
0 1 1112 63

1 11 bits 52 bits

63 62 52 51 0
Double-Precision (little-endian)

Figure 6-1: Floating-point Formats

Floating-Point Load and Store Formats

Floating-point load and store instructions must use even registers. The
operands in Table 6-1 have the following meanings:
Operand Meaning
address Offset (base)
destination Destination register
source Source register

Table 6-1: Floating-Point Load and Store Formats

Description Op-code Operand

Load Fp
Double l.d destination, address
Single l.s
Load Immediate Fp
Double li.d destination, floating-point constant
Single li.s
Store Fp
Double s.d source, address
Single s.s

Assembly Language Programmer’s Guide 6-3

 Chapter 6

Floating-Point Load and Store Descriptions

This part of Chapter 6 groups the instructions by function. Please consult
Table 6-1 for the op-codes. Table 6-2 describes the floating-point Load and
Chapter 6

Store instructions.

Table 6-2: Floating-Point Load and Store Descriptions

Instruction Description
Load eight bytes for double-precision and four bytes for single-precision from
the specified effective address into the destination register, which must be an
Load Fp Instructions even register. The bytes must be word aligned. Note: We recommend that
you use doubleword alignment for double-precision operands. is required in
the mips2 architecture (R4000 & R6000).
Stores eight bytes for double-precision and four bytes for single-precision
from the source floating-point register in the destination register, which must
Store Fp Instructions be an even register. Note: We recommend that you use doubleword
alignment for double-precision operands. It is required in the mips2
architecture (R4000 & R6000).

Floating-Point Computational Formats

This part of Chapter 6 describes floating-point computational instructions.
The operands in Table 6-3 and Table 6-4 have the following meaning:
Operand Meaning
destination Destination register
gpr General-purpose register
source Source register

Table 6-3: Floating-Point Computational Instruction Formats

Description Op-code Operand

Absolute Value Fp
Double abs.d destination, src1
Single abs.s
Negate Fp
Double neg.d
Single neg.s
Add Fp
Double add.d destination, src1, src2
Single add.s
Divide Fp

6-4 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Table 6-3: Floating-Point Computational Instruction Formats

Description Op-code Operand

Double div.d

Chapter 6
Single div.s
Multiply Fp
Double mul.d
Single mul.s
Subtract Fp
Double sub.d
Single sub.s
Convert Source to Specified Fp Precision
Double to Single Fp cvt.s.d destination, src1
Fixed Point to Single Fp cvt.s.w
Single to Double Fp cvt.d.s
Fixed Point to Double Fp cvt.d.w
Single to Fixed Point Fp cvt.w.s
Double to Fixed Point Fp cvt.w.d
Truncate and Round Operations
Truncate to Single Fp trunc.w.s destination, src, gpr
Truncate to Double Fp trunc.w.d
Round to Single Fp round.w.s
Round to Double Fp round.w.d
Ceiling to Double Fp ceil.w.d
Ceiling to Single Fp ceil.w.s
Ceiling to Double Fp, Unsigned ceilu.w.d
Ceiling to Single Fp, Unsigned ceilu.w.s
Floor to Double Fp floor.w.d
Floor to Single Fp floor.w.s
Floor to Double Fp, Unsigned flooru.w.d
Floor to Single Fp, Unsigned flooru.w.s
roundu.w.
Round to Double Fp, Unsigned
d
Round to Single Fp, Unsigned roundu.w.s
Truncate to Double Fp,
truncu.w.d
Unsigned
Truncate to Single Fp, Unsigned truncu.w.s

Assembly Language Programmer’s Guide 6-5

 Chapter 6

Table 6-4: Floating-Point Computational Instruction Formats for mips3

Architecture Only
Chapter 6

Description Op-code Operand

Convert Source to Specified Fp Precision
Long Fixed Point to Single Fp cvt.s.l destination, src1
Long Fixed Point to Double FP cvt.d.l
Single to Long Fixed Point FP cvt.l.s
Double to Long Fixed Point FP cvt.l.d
Truncate and Round Operations
Truncate Single to Long Fixed Point trunc.l.s destination, src, gpr
Truncate Double to Long Fixed Point trunc.l.d
Round Single to Long Fixed Point round.l.s
Round Double to Long Fixed Point round.l.d
Ceiling Single to Long Fixed Point ceil.l.s
Ceiling Double to Long Fixed Point ceil.l.d
Floor Single to Long Fixed Point floor.l.s
Floor Double to Long Fixed Point floor.l.d

6-6 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Floating-Point Computational Instruction Descriptions

This part of Chapter 6 groups the instructions by function. Refer to Table 6-
3 and Table 6-4 for the op-code names. Table 6-5 describes the floating-point

Chapter 6
Computational instructions.

Table 6-5: Floating-Point Computational Instruction Descriptions

Instruction Description
Absolute Value Fp Compute the absolute value of the contents of src1 and put the specified
Instructions precision floating-point result in the destination register.
Add the contents of src1 (or the destination) to the contents of src2 and
put the result in the destination register. When the sum of two operands
Add Fp Single Instructions with opposite signs is exactly zero, the sum has a positive sign for all
rounding modes except round toward –1. For that rounding mode, the
sum has a negative sign.
Convert Source to Another Convert the contents of src1 to the specified precision, round according
Precision Fp Instructions to the rounding mode, and put the result in the destination register.
The trunc instructions truncate the value in the source floating-point
register and put the resulting integer in the destination floating-point
Truncate and Round register, using the third (general-purpose) register to hold a temporary
instructions value. (This is a macro-instruction.) The round instructions work like
trunc, but round the floating-point value to an integer instead of
truncating it.
Compute the quotient of two values. These instructions treat src1 as the
dividend and src2 as the divisor. Divide Fp instructions divide the
Divide Fp Instructions contents of src1 by the contents of src2 and put the result in the
destination register. If the divisor is a zero, the machine signals a error if
the divide-by-zero exception is enabled.
Multiplies the contents of src1 (or the destination) with the contents of
Multiply Fp Instructions
src2 and puts the result in the destination register.
Compute the negative value of the contents of src1 and put the specified
Negate FP Instructions
precision floating-point result in the destination register.
Subtract the contents of src2 from the contents of src1 (or the
destination). These instructions put the result in the destination register.
Subtract Fp Instructions When the difference of two operands with the same signs is exactly zero,
the difference has a positive sign for all rounding modes except round
toward –1. For that rounding mode, the sum has a negative sign.

Assembly Language Programmer’s Guide 6-7

 Chapter 6

Floating-Point Relational Operations

Table 6-6 summarizes the floating-point relational instructions. The first
column under Condition gives a mnemonic for the condition tested. As the
Chapter 6

“branch on true/false” condition can be used to logically negate any

condition, the second column supplies a mnemonic for the logical negation of
the condition in the first column. This provides a total of 32 possible
conditions. The four columns under Relations give the result of the
comparison based on each condition. The final column states if an invalid
operation is signaled for each condition.
For example, with an equal condition (EQ mnemonic in the True column), the
logical negation of the condition is not equal (NEQ), and a comparison that is
equal is True for equal and False for greater than, less than, and unordered,
and no Invalid Operation Exception is given if the relation is unordered.

Table 6-6: Floating-Point Relational Operators

Condition Relations
Invalid Operation
Mnemonic Greater Less Exception if
Equal Unordered Unordered
True False Than Than
F T F F F F no
UN OR F F F T no
EQ NEQ F F T F no
UEQ OLG F F T T no
OLT UGE F T F F no
ULT OGE F T F T no
OLE UGT F T T F no
ULE OGT F T T T no
SF ST F F F F yes
NGLE GLE F F F T yes
SEQ SNE F F T F yes
NGL GL F F T T yes
LT NLT F T F F yes
NGE GE F T F T yes
LE NLE F T T F yes
NGT GT F T T T yes

6-8 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

The mnemonics in have the following meanings:

Mnemonic Meaning Mnemonic Meaning
F False T True

Chapter 6
UN Unordered OR Ordered
EQ Equal NEQ Not Equal
UEQ Unordered or Equal OLG Ordered or Less than or Greater than
OLT Ordered Less Than UGE Unordered or Greater than or Equal
ULT Unordered or Less Than OGE Ordered Greater than or Equal
OLE Ordered Less than or Equal UGT Unordered or Greater Than
ULE Unorderd or Less than or Equal OGT Ordered Greater Than
SF Signaling False ST Signaling True
NGLE Not Greater than or Less than or Equal GLE Greater than, or Less than or Equal
SEQ Signaling Equal SNE Signaling Not Equal
NGL Not Greater than or Less than GL Greater Than or Less Less Than
LT Less Than NLT Not Less Than
NGE Not Greater Than GE Greater Than or Equal or Equal
LE Less Than or Equal NLE Not Less Than or Equal
NGT Not Greater Than GT Greater Than

To branch on the result of a relational:

/* branching on a compare result */

c.eq.s $f1,$f2 /* compare the single-precision values */

bc1t true /* if $f1 equals $f2, branch to true */
bc1f false /* if $f1 does not equal $f2, branch to */
/* false */

Assembly Language Programmer’s Guide 6-9

 Chapter 6

Floating-Point Relational Instruction Formats

In Table 6-7, src1 and src2 refer to the source registers.
NOTE: These are the most common Compare instructions. The machine
Chapter 6

provides other Compare instructions for IEEE compatibility

Table 6-7: Floating-Point Relational Instruction Formats

Description Op-code Operand
Compare F
Double c.f.d src1,src2
Single c.f.s
Compare UN
Double c.un.d
Single c.un.s
*Compare EQ
Double c.eq.d
Single c.eq.s
Compare UEQ
Double c.ueq.d
Single c.ueq.s
Compare OLT
Double c.olt.d
Single c.olt.s
Compare ULT
Double c.ult.d
Single c.ult.s
Compare OLE
Double c.ole.d
Single c.ole.s
Compare ULE
Double c.ule.d
Single c.ule.s
Compare SF
Double c.sf.d
Single c.sf.s
Compare NGLE
Double c.ngle.d src1, src2
Single c.ngle.s
Compare SEQ
Double c.seq.d
Single c.seq.s
Compare NGL
Double c.ngl.d
Single c.ngl.s

6-10 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Table 6-7: Floating-Point Relational Instruction Formats

Description Op-code Operand
*Compare LT

Chapter 6
Double c.lt.d
Single c.lt.s
Compare NGE
Double c.nge.d
Single c.nge.s
*Compare LE
Double c.le.d
Single c.le.s
Compare NGT
Double c.ngt.d
Single c.ngt.s

Floating-Point Relational Instruction Descriptions

This part of Chapter 6 describes the relational instruction descriptions by
function. Refer to Chapter 1 for information regarding registers. Please
consult Table 6-7 for the op-code names.

Table 6-8: Floating-Point Relational Instruction Descriptions

Instruction Description
Compare the contents of src1 with the contents of src2. If src1 equals
Compare EQ Instructions src2 a true condition results; otherwise, a false condition results. The
machine does not signal an exception for unordered values.
Compare the contents of src1 with the contents of src2. These
Compare F Instructions instructions always produce a false condition. The machine does not
signal an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
Compare LE than or equal to src2, a true condition results; otherwise, a false condition
results. The machine signals an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
Compare LT than src2, a true condition results; otherwise, a false condition results.
The machine signals an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
than src2 (or the contents are unordered), a true condition results;
Compare NGE
otherwise, a false condition results. The machine signals an exception for
unordered values.
Compare the contents of src1 with the contents of src2. If src1 equals
src2 or the contents are unordered, a true condition results; otherwise, a
Compare NGL
false condition results. The machine signals an exception for unordered
values.

Assembly Language Programmer’s Guide 6-11

 Chapter 6

Table 6-8: Floating-Point Relational Instruction Descriptions

Instruction Description
Compare the contents of src1 with the contents of src2. If src1 is
Chapter 6

Compare NGLE unordered, a true condition results; otherwise, a false condition results.
The machine signals an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
than or equal to src2 or the contents are unordered, a true condition
Compare NGT
results; otherwise, a false condition results. The machine signals an
exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
Compare OLE Instructions than or equal to src2, a true condition results; otherwise, a false condition
results. The machine does not signal an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
Compare OLT Instructions than src2, a true condition results; otherwise, a false condition results.
The machine does not signal an exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 equals
Compare SEQ Instructions src2, a true condition results; otherwise, a false condition results. The
machine signals an exception for unordered values.
Compare the contents of src1 with the contents of src2. This always
Compare SF Instructions produces a false condition. The machine signals an exception for
unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
than or equal to src2 (or src1 is unordered), a true condition results;
Compare ULE Instructions
otherwise, a false condition results. The machine does not signal an
exception for unordered values.
Compare the contents of src1 with the contents of src2. If src1 equals
src2 (or src1 and src2 are unordered), a true condition results; otherwise,
Compare UEQ Instructions
a false condition results. The machine does not signal an exception for
unordered values.
Compare the contents of src1 with the contents of src2. If src1 is less
than src2 (or the contents are unordered), a true condition results;
Compare ULT Instructions
otherwise, a false condition results. The machine does not signal an
exception for unordered values.
Compare the contents of src1 with the contents of src2. If either src1 or
Compare UN Instructions src2 is unordered, a true condition results; otherwise, a false condition
results. The machine does not signal an exception for unordered values.

6-12 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Floating-Point Move Formats

The floating-point move instructions move data from source to destination
registers (only floating-point registers are allowed).

Chapter 6
Table 6-9: Floating-Point Move Instruction Formats
Description Op-code Operand
Move FP
Single mov.s destination,src1
Double mov.d

Floating-Point Move Instruction Descriptions

This part of Chapter 6 describes the floating-point move instructions. Please
consult Table 6-9 for the op-code names.

Table 6-10: Floating-Point Move Instruction Descriptions

Instruction Description
Move the double or single-precision contents of src1 to the destination
Move FP Instructions
register, maintaining the specified precision.

System Control Coprocessor Instructions

The system control coprocessor (cp0) handles all functions and special and
privileged registers for the virtual memory and exception handling
subsystems. The system control coprocessor translates addresses from a large
virtual address space into the machine’s physical memory space. The
coprocessor uses a translation lookaside buffer (TLB) to translate virtual
addresses to physical addresses.

System Control Coprocessor Instruction Formats

These coprocessor system control instructions do not have operands.
Table 6-11: System Control Instruction Formats
Description Op-code
Cache** cache
Translation Lookaside Buffer Probe tlbp
Translation Lookaside Buffer Read tlbr
Translation Lookaside Buffer Write Random tlbwr
Translation Lookaside Write Index tlbwi
Synchronize* sync
* Not valid in mips1 architectures.
** Not valid in mips1 and mips2 architectures.

Assembly Language Programmer’s Guide 6-13

 Chapter 6

System Control Coprocessor Instruction Descriptions

This part of Chapter 6 describes the system control coprocessor instructions.
Chapter 6

Table 6-12: System Control Coprocessor Instruction Descriptions

Instruction Description
Cache is the R4000 instruction to perform cache operations. The 16-bit
offset is sign-extended and added to the contents of general register
base to form a virtual address. The virtual address is translated to a
physical address using the TLB. The 5-bit sub-opcode (“op”) specifies
Cache (cache) ** the cache operation for that address. Part of the virtual address is used to
specify the cache block for the operation. Possible operations include
invalidating a cache block, writeback to a secondary cache or memory,
etc.
** This instruction is not valid in mips1 or mips2 architectures.
Probes the translation lookaside buffer (TLB) to see if the TLB has an
entry that matches the contents of the EntryHi register. If a match occurs,
Translation Lookaside
the machine loads the Index register with the number of the entry that
Buffer Probe (tlbp)
matches the EntryHi register. If no TLB entry matches, the machine sets
the high-order bit of the Index register.
Loads the EntryHi and EntryLo registers with the contents of the
Translation Lookaside
translation lookaside buffer (TLB) entry specified in the TLB Index
Buffer Read (tlbr)
register.
Loads the specified translation lookaside buffer (TLB) entry with the
Translation Lookaside
contents of the EntryHi and EntryLo registers. The contents of the TLB
BufferWrite Random (tlbwr)
Random register specify the TLB entry to be loaded.
Loads the specified translation lookaside buffer (TLB) entry with the
Translation Lookaside
contents of the EntryHi and EntryLo registers. The contents of the TLB
Buffer Write Index (tlbwi)
Index register specify the TLB entry to be loaded.
Ensures that all loads and stores fetched before the sync are completed,
before allowing any following loads or stores. Use of sync to serialize
Synchronize (sync) *
certain memory references may be required in multiprocessor
environments. * This instruction is not valid in the mips1 architecture.

6-14 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Control and Status Register

Floating-point coprocessor control register 31 contains status and control
information. It controls the arithmetic rounding mode and the enabling of

Chapter 6
user-level traps, and indicates exceptions that occurred in the most recently
executed instruction, and any exceptions that may have occurred without
being trapped:

31 24 23 22 18 17 12 11 7 6 2 1 0
0 c 0 exceptions enables sticky- RM
bits
bits: 8 1 5 6 5 5 2
Control and Status Register
(c = compare bit)

11 10 9 8 7 17 16 15 14 13 12 6 5 4 3 2
V Z O U I E V Z O U I V Z O U I

Enable Bits Exception Bits Sticky Bits

Figure 6-2: Floating Control and Status Register 31

The exception bits are set for instructions that cause an IEEE standard
exception or an optional exception used to emulate some of the more
hardware-intensive features of the IEEE standard.
The exception field is loaded as a side-effect of each floating-point operation
(excluding loads, stores, and unformatted moves). The exceptions which
were caused by the immediately previous floating-point operation can be
determined by reading the exception field.
The meaning of each bit in the exception field is given below. If two
exceptions occur together on one instruction, the field will contain the
inclusive-OR of the bits for each exception:

Exception
Description
Field Bit
E Unimplemented Operation
I Inexact Exception
O Overflow Exception
U Underflow Exception
V Invalid Operation
Z Division-by-Zero

Assembly Language Programmer’s Guide 6-15

 Chapter 6

The unimplemented operation exception is normally invisible to user-level

code. It is provided to maintain IEEE compatibility for non-standard
implementations.
Chapter 6

The five IEEE standard exceptions are listed below:

Field Description
I Inexact Exception
O Overflow Exception
U Underflow Exception
V Invalid Operationz
Z Division-by-Zero

Each of the five exceptions is associated with a trap under user control, which
is enabled by setting one of the five bits of the enable field, shown above.
When an exception occurs, both the corresponding exception and status bits
are set. If the corresponding enable flag bit is set, a trap is taken. In some cases
the result of an operation is different if a trap is enabled.
The status flags are never cleared as a side effect of floating-point operations,
but may be set or cleared by writing a new value into the status register, using
a “move to coprocessor control” instruction.
The floating-point compare instruction places the condition which was
detected into the `c’ bit of the control and status register, so that the state of
the condition line may be saved and restored. The `c’ bit is set if the condition
is true, and cleared if the condition is false, and is affected only by compare
and move to control register instructions.

Exception Trap Processing

For each IEEE standard exception, a status flag is provided that is set on any
occurrence of the corresponding exception condition with no corresponding
exception trap signaled. It may be reset by writing a new value into the status
register. The flags may be saved and restored individually, or as a group, by
software. When no exception trap is signaled, a default action is taken by the
floating-point coprocessor, which provides a substitute value for the original,
exceptional, result of the floating-point operation. The default action taken
depends on the type of exception, and in the case of the Overflow exception,
the current rounding mode.

6-16 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Invalid Operation Exception

The invalid operation exception is signaled if one or both of the operands are
invalid for an implemented operation. The result, when the exception occurs

Chapter 6
without a trap, is a quiet NaN when the destination has a floating-point
format, and is indeterminate if the result has a fixed-point format. The invalid
operations are:
• Addition or subtraction: magnitude subtraction of infinities, such as (
+ 1 ) – ( – 1 ).
• Multiplication: 0 times 1, with any signs.
• Division: 0 over 0 or 1 over 1, with any signs.
• Square root: x , where x is less than zero.
• Conversion of a floating-point number to a fixed-point format when
an overflow, or operand value of infinity or NaN, precludes a faithful
representation in that format.
• Comparison of predicates involving < or > without ?, when the
operands are “unordered”.
• Any operation on a signaling NaN.
Software may simulate this exception for other operations that are invalid for
the given source operands. Examples of these operations include IEEE-
specified functions implemented in software, such as Remainder: x REM y,
where y is zero or x is infinite; conversion of a floating-point number to a
decimal format whose value causes and overflow or is infinity of NaN; and
transcendental functions, such as ln (–5) or cos -1(3).

Division-by-zero Exception
The division by zero exception is signaled on an implemented divide
operation if the divisor is zero and the dividend is a finite nonzero number.
The result, when no trap occurs, is a correctly signed infinity.
If division by zero traps are enabled, the result register is not modified, and
the source registers are preserved.
Software may simulate this exception for other operations that produce a
signed infinity, such as ln(0), sec(p/2), csc(0) or 0-1.

Assembly Language Programmer’s Guide 6-17

 Chapter 6

Overflow Exception
The overflow exception is signaled when what would have been the
magnitude of the rounded floating-point result, were the exponent range
Chapter 6

unbounded, is larger than the destination format’s largest finite number. The
result, when no trap occurs, is determined by the rounding mode and the sign
of the intermediate result.
If overflow traps are enabled, the result register is not modified, and the
source registers are preserved.

Underflow Exception
Two related events contribute to underflow. One is the creation of a tiny non-
zero result between ±2Emin (minimum expressible exponent) which, because
it is tiny, may cause some other exception later. The other is extraordinary
loss of accuracy during the approximation of such tiny numbers by
denormalized numbers.
The IEEE standard permits a choice in how these events are detected, but
requires that they must be detected the same way for all operations.
The IEEE standard specifies that “tininess” may be detected either: “after
rounding” (when a nonzero result computed as though the exponent range
were unbounded would lie strictly between ±2Emin, or “before rounding”
(when a nonzero result computed as though the exponent range and the
precision were unbounded would lie strictly between ±2Emin. The
architecture requires that tininess be detected after rounding.
Loss of accuracy may be detected as either “denormalization loss” (when the
delivered result differs from what would have been computed if the exponent
range were unbounded), or “inexact result” (when the delivered result differs
from what would have been computed if the exponent range and precision
were both unbounded). The architecture requires that loss of accuracy be
detected as inexact result.
When an underflow trap is not enabled, underflow is signaled (via the
underflow flag) only when both tininess and loss of accuracy have been
detected. The delivered result might be zero, denormalized, or ± 2Emin. When
an underflow trap is enabled, underflow is signaled when tininess is detected
regardless of loss of accuracy.
If underflow traps are enabled, the result register is not modified, and the
source registers are preserved.

6-18 Assembly Language Programmer’s Guide

 Coprocessor Instruction Set

Inexact Exception
If the rounded result of an operation is not exact or if it overflows without an
overflow trap, then the inexact exception is signaled. The rounded or

Chapter 6
overflowed result is delivered to the destination register, when no inexact trap
occurs. If inexact exception traps are enabled, the result register is not
modified, and the source registers are preserved.

Unimplemented Operation Exception

If an operation is specified that the hardware may not perform, due to an
implementation restriction on the supported operations or supported formats,
an unimplemented operation exception may be signaled, which always causes
a trap, for which there are no corresponding enable or flag bits. The trap
cannot be disabled.
This exception is raised at the execution of the unimplemented instruction.
The instruction may be emulated in software, possibly using implemented
floating-point unit instructions to accomplish the emulation. Normal
instruction execution may then be restarted.
This exception is also raised when an attempt is made to execute an
instruction with an operation code or format code which has been reserved for
future architectural definition. The unimplemented instruction trap is not
optional, since the current definition contains codes of this kind.
This exception may be signaled when unusual operands or result conditions
are detected, for which the implemented hardware cannot properly handle the
condition. These may include (but are not limited to), denormalized operands
or results, NaN operands, trapped overflow or underflow conditions. The use
of this exception for such conditions is optional.

Assembly Language Programmer’s Guide 6-19

 Chapter 6

Floating-Point Rounding
Bits 0 and 1 of the coprocessor control register 31 sets the rounding mode for
floating-point. The machine allows four rounding modes:
Chapter 6

• Round to nearest rounds the result to the nearest representable

value. When the two nearest representable values are equally near,
this mode rounds to the value with the least significant bit zero. To
select this mode, set bits 1..0 of control register 31 to 0.
• Round toward zero rounds toward zero. It rounds to the value that is
closest to and not greater in magnitude than the infinitely precise
result. To select this mode, set bits 1..0 of control register 31 to 1.
• Round toward positive infinity rounds to the value that is closest to
and not less than the infinitely precise result. To select this mode, set
bits 1..0 of control register 31 to 2.
• Round toward negative infinity rounds toward negative infinity. It
rounds to the value that is closest to and not greater than the
infinitely precise result. To select this mode, set bits 1..0 of control
register 31 to 3.
To set the rounding mode:
/* setting the rounding mode */
RoundNearest = Ox0
RoundZero = Ox1
RoundPosInf = Ox2
RoundNegInf = Ox3
cfc1 rt2, $31# move from coprocessor 1
and rt, Oxfffffffc# zero the round mode bits
or rt, RoundZero# set mask as round to zero
ctc1 rt, $f31# move to coprocessor 1

6-20 Assembly Language Programmer’s Guide

Linkage Conventions

Chapter 7
This chapter gives rules and examples to follow when designing an assembly
language program. The chapter concludes with a “learn by doing” technique
that you can use if you still have any doubts about how a particular calling
sequence should work. This involves writing a skeleton version of your
prospective assembly routine using a high level language, and then compiling
it with the –S option to generate a human-readable assembly language file.
The assembly language file can then be used as the starting point for coding
your routine.

Introduction
When you write assembly language routines, you should follow the same
calling conventions that the compilers observe, for two reasons:
• Often your code must interact with compiler-generated code,
accepting and returning arguments or accessing shared global data.
• The symbolic debugger gives better assistance in debugging
programs using standard calling conventions.
The conventions for the compiler system are a bit more complicated than
some, mostly to enhance the speed of each procedure call. Specifically:
• The compilers use the full, general calling sequence only when
necessary; where possible, they omit unneeded portions of it. For
example, the compilers don’t use a register as a frame pointer
whenever possible.
• The compilers and debugger observe certain implicit rules rather than
communicating via instructions or data at execution time. For
example, the debugger looks at information placed in the symbol

Assembly Language Programmer’s Guide 7-1

 Chapter 7

table by a “.frame” directive at compilation time, so that it can

tolerate the lack of a register containing a frame pointer at execution
time.

Program Design
This section describes three general areas of concern to the assembly
language programmer:
• Usable and restricted registers.
• Stack frame requirements on entering and exiting a routine.
• The “shape” of data (scalars, arrays, records, sets) laid out by the
Chapter 7

various high level languages.

Register Use and Linkage

The main processor has 32 integer registers. They are each 32-bit wide in
mips1 and mips2 architectures. In mips3 architecture, each register is 64-bit
wide. The uses and restrictions of these registers are described in Table 1-1 in
Chapter 1.
The floating-point coprocessor has 16 floating-point registers. Each register
can hold either a single precision (32 bit) or a double precision (64 bit) value.
All references to these registers uses an even register number (e.g., $f4).
Refer to Table 7-1 for details.

Table 7-1: Floating-Point Registers

Register Name Use and Linkage

Used to hold floating-point type function results ($f0) and complex type function
$f0..f3
results ($f0 has the real part, $f2 has the imaginary part).
Temporary registers, used for expression evaluation, whose values are not
$f4..f10
preserved across procedure calls.
Used to pass the first 2 single or double precision actual arguments, whose values
$f12..$f14
are not preserved across procedure calls.
Temporary registers, used for expression evaluations, whose values are not
$f16..$f18
preserved across procedure calls.
$f20..$f30 Saved registers, whose values must be preserved across procedure calls.

The Stack Frame

The compilers classify each routine into one of the following categories:
• Non-leaf routines, that is, routines that call other procedures.

7-2 Assembly Language Programmer’s Guide

 Linkage Conventions

• Leaf routines, that is, routines that do not themselves execute any
procedure calls. Leaf routines are of two types:
• Leaf routines that require stack storage for local variables
• Leaf routines that do not require stack storage for local variables.
You must decide the routine category before determining the calling
sequence.
To write a program with proper stack frame usage and debugging capabilities,
use the following procedure:
1. Regardless of the type of routine, you should include a .ent pseudo-op
and an entry label for the procedure. The .ent pseudo-op is for use by the

Chapter 7
debugger, and the entry label is the procedure name. The syntax is:
.ent procedure_name
procedure_name:
2. If you are writing a leaf procedure that does not use the stack, skip to step
3. For leaf procedure that uses the stack or non-leaf procedures, you must
allocate all the stack space that the routine requires. The syntax to adjust
the stack size is:
subu $sp,framesize
where framesize is the size of frame required; framesize must be a
multiple of 8. Space must be allocated for:
• Local variables.
• Saved general registers. Space should be allocated only for those
registers saved. For non-leaf procedures, you must save $31, which is
used in the calls to other procedures from this routine. If you use
registers $16–$23, you must also save them.
• Saved floating-point registers. Space should be allocated only for
those registers saved. If you use registers $f20–$f30 you must also
save them.
• Procedure call argument area. You must allocate the maximum
number of bytes for arguments of any procedure that you call from
this routine.
NOTE: Once you have modified $sp, you should not modify it again for the
rest of the routine.
3. Now include a .frame pseudo-op:
.frame framereg,framesize,returnreg

Assembly Language Programmer’s Guide 7-3

 Chapter 7

The virtual frame pointer is a frame pointer as used in other compiler

systems but has no register allocated for it. It consists of the framereg
($sp, in most cases) added to the framesize (see step 2 above). Figure 7-1
illustrates the stack components.

high memory
argument n
•
•
•
virtual argument 1
framepointer ($fp)
Chapter 7

frame offset local & temporaries

saved registers framesize

(including returnreg)

argument build
stack pointer($sp)
(framereg) •
•
•

low memory

Figure 7-1: Stack Organization

The returnreg specifies the register containing the return address (usually
$31). These usual values may change if you use a varying stack pointer or
are specifying a kernel trap routine.
4. If the procedure is a leaf procedure that does not use the stack, skip to
step 7. Otherwise you must save the registers you allocated space for in
step 2.

To save the general registers, use the following operations:

.maskbitmask,frameoffset
sw reg,framesize+frameoffset–N($sp)

7-4 Assembly Language Programmer’s Guide

 Linkage Conventions

The .mask directive specifies the registers to be stored and where they are
stored.A bit should be on in bitmask for each register saved (for example,
if register $31 is saved, bit 31 should be ‘1’ in bitmask.Bits are set in
bitmask in little-endian order, even if the machine configuration is big-
endian).The frameoffset is the offset from the virtual frame pointer (this
number is usually negative).N should be 0 for the highest numbered
register saved and then incremented by four for each subsequently lower
numbered register saved.For example:
sw $31,framesize+frameoffset($sp)
sw $17,framesize+frameoffset–4($sp)
sw $16,framesize+frameoffset–8($sp)
Figure 7-2 illustrates this example.

Chapter 7
high memory

virtual frame pointer ($fp)

frame
offset
saved $31
saved $17
saved $16 framesize

•
•
•
stack pointer ($sp)

low memory

Figure 7-2: Stack Example

Now save any floating-point registers that you allocated space for in step
2 as follows:
.fmask bitmask,frameoffset
s.[sd] reg,framesize+frameoffset–N($sp)

Assembly Language Programmer’s Guide 7-5

 Chapter 7

Notice that saving floating-point registers is identical to saving general

registers except we use the .fmask pseudo-op instead of .mask, and the
stores are of floating-point singles or doubles.The discussion regarding
saving general registers applies here as well, but remember that N should
be incremented by 8 for doubles.The stack framesize must be a multiple
of 8.
5. This step describes parameter passing: how to access arguments passed
into your routine and passing arguments correctly to other procedures. For
information on high-level language specific constructs (call-by-name, call-
by-value, string or structure passing), refer to Chapter 4 of the
RISCompiler and C Programmer’s Guide.
As specified in step 2, space must be allocated on the stack for all
Chapter 7

arguments even though they may be passed in registers. This provides a

saving area if their registers are needed for other variables.
General registers $4–$7 and float registers $f12, $f14 must be used for
passing the first four arguments (if possible). You must allocate a pair of
registers (even if it’s a single precision argument) that start with an even
register for floating-point arguments appearing in registers.
In the table below, the “fN” arguments are considered single- and double-
precision floating-point arguments, and “nN” arguments are everything
else. The ellipses (...) mean that the rest of the arguments do not go in
registers regardless of their type. The “stack” assignment means that you
do not put this argument in a register. The register assignments occur in
the order shown in order to satisfy optimizing compiler protocols:

Arguments Register Assignments

(f1, f2, ...) f1 -> $f12, f2 -> $f14
(f1, n1, f2, ...) f1 -> $f12, n1 -> $6, f2 ->stack
(f1, n1, n2, ...) f1 -> $f12, n1 -> $6, n2 -> $7
(n1, n2, n3, n4, ...) n1 -> $4, n2 -> $5, n3 -> $6, n4 -> $7
(n1, n2, n3, f1, ...) n1 -> $4, n2 -> $5, n3 -> $6, f1 -> stack
(n1, n2, f1, ...) n1 -> $4, n2 -> $5, f1 -> ($6, $6)
(n1, f1, ...) n1 -> $4, f1 -> ($6, $7)

6. Next, you must restore registers that were saved in step 4. To restore
general purpose registers:
lw reg,framesize+frameoffset–N($sp)
To restore the floating-point registers:
l.[sd] reg,framesize+frameoffset–N($sp)

7-6 Assembly Language Programmer’s Guide

 Linkage Conventions

Refer to step 4 for a discussion of the value of N.)

7. Get the return address:
lw $31,framesize+frameoffset($sp)
8. Clean up the stack:
addu $sp, framesize
9. Return:
j $31
10. To end the procedure:
.end procedurename

Chapter 7
The Shape of Data
In most cases, high-level language routine and assembly routines
communicate via simple variables: pointers, integers, booleans, and single-
and double-precision real numbers. Describing the details of the various high-
level data structures (arrays, records, sets, and so on) is beyond our scope
here. If you need to access such a structure as an argument or as a shared
global variable, refer to Chapter 4 of the RISCompiler and C Programmer’s
Guide, and the “Learn by Doing” technique described at the end of this
section.

Examples
This section contains the examples that illustrate program design rules. Each
example shows a procedure written and C and its equivalent written in
assembly language.
Figure 7-3 shows a non-leaf procedure. Notice that it creates a stackframe,
and also saves its return address since it must put a new return address into
register $31 when it invokes its callee:

Assembly Language Programmer’s Guide 7-7

 Chapter 7

float
nonleaf(i, j)
int i, *j;
{
double atof();
int temp;

temp = i - *j;
if (i < *j) temp = -temp;
return atof(temp);
}

.globl nonleaf
# 1 float
Chapter 7

# 2 nonleaf(i, j)
# 3 int i, *j;
# 4 {
.ent nonleaf 2
nonleaf:
subu $sp, 24 ## Create stackframe
sw $31, 20($sp) ## Save the return address
.mask 0x80000000, -4
.frame $sp, 24, $31
# 5 double atof();
# 6 int temp;
# 7
# 8 temp = i - *j;
lw $2, 0($5) ## Arguments are in $4 and $5
subu $3, $4, $2
# 9 if (i < *j) temp = -temp;
bge $4, $2, $32 ## Note: $32 is a label, not a reg
negu $3, $3
$32:
# 10 return atof(temp);
move $4, $3
jal atof
cvt.s.d $f0, $f0 ## Returnvalue goes in $f0
lw $31, 20($sp) ## Restore return address
addu $sp, 24 ## Delete stackframe
j $31 ## Return to caller
.end nonleaf

Figure 7-3: Non-Leaf Procedure

Figure 7-4 shows a leaf procedure that does not require stack space for local
variables. Notice that it creates no stackframe, and saves no return address.

7-8 Assembly Language Programmer’s Guide

 Linkage Conventions

int
leaf(p1, p2)
int p1, p2;
{
return (p1 > p2) ? p1 :p2;
}
.globl leaf
# 1 int
# 2 leaf(p1, p2)
# 3 int p1, p2;
# 4 {
.ent leaf 2
leaf:

Chapter 7
.frame $sp, 0, $31
# 5 return (p1 > p2) ? p1 : p2;
ble $4, $5, $32 ## Arguments in $4 and $5
move $3, $4
b $33
$32:
move $3, $5
$33:
move $2, $3 ## Return value goes in $2
j $31 ## Return to caller
# 6 }
.end leaf

Figure 7-4: Leaf Procedure Without Stack Space for Local Variables

Figure 7-5 shows a leaf procedure that requires stack space for local variables.
Notice that it creates a stack frame, but does not save a return address.

Assembly Language Programmer’s Guide 7-9

 Chapter 7

char
leaf_storage(i)
int i;
{
char a[16];
int j;
for (j = 0; j < 10; j++)
a[j] = ’0’ + j;
for (j = 10; j < 16; j++)
a[j] = ’a’ + j;
return a[i];
}
.globl leaf_storage
# 1 char
Chapter 7

# 2 leaf_storage(i)
# 3 int i;
# 4 {
.ent leaf_storage 2## "2" is the lexical level of
## the procedure. You may omit it
leaf_storage:
subu $sp, 24 ## Create stackframe.
.frame $sp, 24, $31
# 5 char a[16];
# 6 int j;
# 7
# 8 for (j = 0; j < 10; j++)
sw $0, 4($sp)
addu $3, $sp, 24
$32:
# 9 a[j] = ’0’ + j;
lw $14, 4($sp)
addu $15, $14, 48
addu $24, $3, $14
sb $15, -16($24)
lw $25, 4($sp)
addu $8, $25, 1
sw $8, 4($sp)
blt $8, 10, $32
# 10 for (j = 10; j < 16; j++)
li $9, 10
sw $9, 4($sp)
$33:
# 11 a[j] = ’a’ + j;
lw $10, 4($sp)
addu $11, $10, 97
addu $12, $3, $10
sb $11, -16($12)
lw $13, 4($sp)
addu $14, $13, 1
sw $14, 4($sp)
blt $14, 16, $33
# 12 return a[i];
addu $15, $3, $4 ## Argument is in $4.
lbu $2, -16($15) ## Return value goes in $2
addu $sp, 24 ## Delete stackframe.
j $31 ## Return to caller.
.end leaf_storage

Figure 7-5: Leaf Procedure With Stack Space for Local Variables

7-10 Assembly Language Programmer’s Guide

 Linkage Conventions

Learning by Doing
The rules and parameter requirements required between assembly language
and other languages are varied and complex. The simplest approach to coding
an interface between an assembly routine and a routine written in a high-level
language is to do the following:
• Use the high-level language to write a skeletal version of the routine
that you plan to code in assembly language.
• Compile the program using the –S option, which creates an assembly
language (.s) version of the compiled source file.
• Study the assembly-language listing and then, imitating the rules and

Chapter 7
conventions used by the compiler, write your assembly language
code.
The next two sections illustrate techniques to use in creating an interface
between assembly language and high-level language routines. The examples
shown are merely to illustrate what to look for in creating your interface.
Details such as register numbers will vary according to the number, order, and
data types of the arguments. You should write and compile realistic examples
of your own code in writing your particular interface.

Calling a High-Level Language Routine

The following steps show a technique to follow in writing an assembly
language routine that calls atof, a routine written in C that converts ASCII
characters to numbers; for more information, see the atof(3) in the RISC/os
Programmer’s Reference Manual.
1. Write a C program that calls atof. Pass global rather than local variables;
this makes them to recognize in the assembly language version of the C
program (and ensures that optimization doesn’t remove any of the code
on the grounds that it has no effect).
Below is an example of a C program that calls atof:

char c[] = "3.1415";

double d, atof();
float f; c is declared as a
caller() global variable.
{
d = atof(c);
f = (float) atof(c);
}

2. Compile the program using the compiler options shown below:

Assembly Language Programmer’s Guide 7-11

 Chapter 7

cc –S –O caller.c
The –S option causes the compiler to produce the assembly-language
listing; the –O option, though not required, reduces the amount of code
generated, making the listing easier to read.
After compilation, look at the file caller.s (shown below). The highlighted
section of the listing shows how the parameters are passed, the execution of
the call, and how the returned values are retrieved:

.globl c
.align 2
c:
Chapter 7

.word 875638323 : 1
.word 13617 : 1
.comm d 8
.comm f 4
.globl caller
.text
.ent caller 2
caller:
subu $sp, 24
sw $31, 20($sp)
.mask 0x80000000, -4
.frame $sp, 24, $31
# 1 char c[] = "3.1415";
# 2 double d, atof();
# 3 float f;
# 4 caller()
# 5 {
# 6 d = atof(c);

la $4, c ## load address of c

jal atof ## call atof
s.d $f0, d ## store result in d
# 7
la $4, c ## load address of c
jal atof ## call atof
cvt.s.d $f4, $f0 ## convert double result to float
s.s $f4, f ## store float result in f
lw $31, 20($sp)
addu $sp, 24
j $31
.end caller

Calling an Assembly Language Routine

This section shows a technique to follow in writing an assembly language
routine that calls a routine written in a high-level language (Pascal is used in
this example).

7-12 Assembly Language Programmer’s Guide

 Linkage Conventions

1. Write a facsimile of the assembly language routine you wish to call. In

the body of the routine, write statements that use the same arguments you
intend to use in the final assembly language routine. Copy the arguments
to global variables rather than local variables to make it easy for you to
read the resulting assembly language listing.
Below is the Pascal facsimile of the assembly language program:

type
str = packed array [1 .. 10] of char;
subr = 2 .. 5;
var
global_r: real;

Chapter 7
global_c: subr;
global_s: str;
global_b: boolean;

function callee(var r: real; c: subr; s: str): boolea

begin
global_r := r;
global_c := c;
global_s := s;
callee := c = 3;
end;

2. Compile the program using the compiler options shown below:

cc –S –O caller.c
The –S option causes the compiler to produce the assembly-language
listing; the –O option, though not required, reduces the amount of code
generated, making the listing easier to read.
3. After compilation, look at the file caller.s (shown below). The
highlighted section of the listing shows how the parameters are passed,
the execution of the call, and how the returned values are retrieved.

Assembly Language Programmer’s Guide 7-13

 Chapter 7

.lcomm $dat 0
.comm global_r 4
.comm global_c 1
.comm global_s 10
.comm global_b 1
.text
.globl callee
# 10 function callee(var r: real; c: subr; s: str): boolean;
.ent callee 2
callee:
.frame $sp, 0, $31
sw $5, 4($sp)
sw $6, 8($sp)
Chapter 7

lbu $3, 4($sp) ## Get subrange c, masking it to 8 bits

and $3, $3, 255

# 11 begin
# 12 global_r := r;
l.s $f4, 0($4) ## The pointer to "r" is in 0($4)
s.s $f4, global_r
# 13 global_c :=
$3,c;global_c
sb
# 14 global_s := s;
la $14, global_s ## For array "s", the caller gives you a
addu $15, $sp, 8 ## pointer at 8($sp). If youwant to use
.set noat ## it as a call-by-value argument just as
addu $24, $15, 10 ## Pascal does (that is, if you want to
$32: ## be able to modify a local copy without
lbu $1, 0($15) ## affecting the global copy) then you
addu $15, $15, 2 ## must copy it into your stack frame as
sb $1, 0($14) ## shown here (the code enclosed in ".set
lbu 31 $1, -1($15) ## noat" is a tight byte-copying loop).
addu $14, $14, 2 ## Otherwise, you may simply use the
sb $1, -1($14) ## pointer provided to you.
bne $15, $24, $32
.set at
# 15 callee := c = 3;
seq $5, $3, 3
and $5, $5, 255
# 16 end;
and $2, $5, 255 ## Return the boolean by leaving it in $2
j $31
.end

7-14 Assembly Language Programmer’s Guide

 Linkage Conventions

Memory Allocation
The machine’s default memory allocation scheme gives every process two
storage areas, that can grow without bound. A process exceeds virtual storage
only when the sum of the two areas exceeds virtual storage space. The link
editor and assembler use the scheme shown in Figure 7-6. An explanation of
each area in the allocation scheme follows the figure.

Chapter 7

Assembly Language Programmer’s Guide 7-15

 Chapter 7

0xffffffff
Reserved for Kernel
(accessible from Kernel Mode) 1
0x8fffffff (2GB)
0x7fffffff
Not Accessible 2
(by convention, not a hardware
implementation)
0x7ffff000 (4KB)
0x7fffefff Activation Stack 3
(grows toward zero)
$sp
Protected
Chapter 7

4
(grows from either edge)
Heap
5
(grows up)
.bss
.sbss
$gp .sdata
.lit4 6

.lit8
.data

0x10000000 Reserved for

0xfffffff Shared Libraries 7
Not Used
Program .text 8
(including header)
0x400000
0x3fffff Reserved 9
(4MB)
0x0

Figure 7-6: Layout of memory (User Program View)

1. Reserved for kernel operations.

2. Reserved for operating system use.
3. Used for local data in C programs.

7-16 Assembly Language Programmer’s Guide

 Linkage Conventions

4. Not allocated until a user requests it, as in System V shared memory

regions.
5. The heap is reserved for sbrk and break system calls, and it not always
present.
6. The machine divides all data into one of five sections:
a. bss - Uninitialized data with a size greater than the value specified
by the –G command line option.
b. sbss - Data less than or equal to the –G command line option. (512
is the default value for the –G option.)
c. sdata (small data) - Data initialized and specified for the sdata

Chapter 7
section.
d. data (data) - Data initialized and specified for the data section.
7. Reserved for any shared libraries.
8. Contains the .text section, .rdata section and all dynamic tables.
9. Reserved.

Assembly Language Programmer’s Guide 7-17

 Chapter 7
Chapter 7

7-18 Assembly Language Programmer’s Guide

Pseudo Op-Codes

8
This chapter describes pseudo op-codes (directives). These pseudo op-codes
influence the assembler’s later behavior. In the text, boldface type specifies a
keyword and italics represents an operand that you define.
The assembler has the pseudo op-codes listed in Table 8-1.

Table 8-1: Pseudo Op-Codes

Chapter 8
Pseudo-Op Description
Sets an alternate entry point for the current procedure. Use this
.aent name, symno information when you want to generate information for the
debugger. It must appear inside an .ent/.end pair.
Indicates that memory reference through the two registers ( reg1,
.alias reg1, reg2 reg2) will overlap. The compiler uses this form to improve
instruction scheduling.
Advance the location counter to make the expression low order
bits of the counter zero. Normally, the .half, .word, .float, and
.double directives automatically align their data appropriately. For
example, .word does an implicit .align 2 (.double does an .align 3).
You disable the automatic alignment feature with .align 0. The
.align expression assembler reinstates automatic alignment at the next .text, .data,
.rdata, or .sdata directive.

Labels immediately preceding an automatic or explicit alignment

are also realigned. For example, foo: .align 3; .word 0 is the same
as .align 3; foo: .word0.

Assembly Language Programmer’s Guide 8-1

 Chapter 8

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Assembles each string from the list into successive locations. The
.ascii directive does not null pad the string. You MUST put
.ascii string [, string]... quotation marks (”) around each string. You can use the
backslash escape characters. For a list of the backslash
characters, see Chapter 4.
Assembles each string in the list into successive locations and
.asciiz string [, string]... adds a null. You can use the backslash escape characters. For a
list of the backslash characters, see Chapter 4.
Tells the assembler’s second pass that this assembly came from
.asm0
the first pass. (For use by compilers.)
(For use by compilers.) Sets the beginning of a language block.
The .bgnb and .endb directives delimit the scope of a variable set.
The scope can be an entire procedure, or it can be a nested scope
.bgnb symno (for example a “{}” block in the C language). The symbol number
symno refers to a dense number in a .T file. For an explanation of
.T files, see the RISCompiler and C Programmer’s Guide. To set
the end of a language block, see .endb.
Truncates the expressions from the comma-separated list to 8-bit
values, and assembles the values in successive locations. The
Chapter 8

.byte expression1 [, expression2 ]

expressions must be absolute. The operands can optionally have
...[, expressionN]
the form: expression1 [ : expression2 ]. The expression2
replicates expression1’s value expression2 times.
Unless defined elsewhere, name becomes a global common
symbol at the head of a block of expression bytes of storage. The
.comm name, expression
linker overlays like-named common blocks, using the maximum of
the expressions.
.cpadd reg Emits code that adds the value of “_gp” to reg.
Expands into the three instructions function prologue that sets up
.cpload reg the $gp register. This directive is used by position-independent
code.
Causes the assembler to emit the following at the point where it
occurs:
sw $gp, offset ($sp)
.cprestore offset Also, causes the assembler to generate:
lw $gp, offset ($sp)
after every JAL or BAL operation. Offset should point to the saved
register area as described in Chapter 7.
.data Tells the assembler to add all subsequent data to the data section.

8-2 Assembly Language Programmer’s Guide

 Pseudo Op-Codes

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Initializes memory to 64-bit floating point numbers. The operands
can optionally have the form: expression1 [ : expression2 ]. The
expression1 is the floating point value. The optional expression2
.double expression is a non-negative expression that specifies a repetition count. The
[ , expression2] ...[, expressionN] expression2 replicates expression1’s value expression2 times.
This directive automatically aligns its data and any preceding
labels to a double-word boundary. You can disable this feature by
using .align 0.
Truncates the expressions in the comma-separated list to 64-bits
and assembles the values in successive locations. The
expressions must be absolute. The operands can optionally have
.dword expression
the form: expression1 [:expression2]. The expresssion2 replicates
[ , expression2 ] ...[, expressionN]
expression1’s value expression2 number of times. The directive
automatically aligns its data and preceding labels to a doubleword
boundary. You can disable this feature by using .align 0.
Sets the end of a procedure. Use this directive when you want to
.end [proc_name] generate information for the debugger. To set the beginning of a
procedure, see .ent.
Sets the end of a language block. To set the beginning of a
.endb symno

Chapter 8
language block, see .bgnb.
Signals the end of a repeat block. To start a repeat block, see
.endr
.repeat.
Sets the beginning of the procedure proc_name. Use this directive
.ent proc_name when you want to generate information for the debugger. To set
the end of a procedure, see .end.
name is a global undefined symbol whose size is assumed to be
expression bytes. The advantage of using this directive, instead of
permitting an undefined symbol to become global by default, is
that the assembler can decide whether to use the economical
.extern name expression
$gp-relative addressing mode, depending on the value of the –G
option. As a special case, if expression is zero, the assembler
refrains from using $gp to address this symbol regardless of the
size specified by –G.
Signals an error. Any compiler front-end that detects an error
condition puts this directive in the input stream. When the
.err assembler encounters a .err, it quietly ceases to assemble the
source file. This prevents the assembler from continuing to
process a program that is incorrect. (For use by compilers.)

Assembly Language Programmer’s Guide 8-3

 Chapter 8

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Specifies the source file corresponding to the assembly
instructions that follow. For use only by compilers, not by
.file file_number file_name_string programmers; when the assembler sees this, it refrains from
generating line numbers for dbx to use unless it also sees .loc
directives.
Initializes memory to single precision 32-bit floating point
numbers. The operands can optionally have the form: expression1
[ : expression2 ]. The optional expression2 is a non-negative
.float expression1
expression that specifies a repetition count. This optional form
[ , expression2 ] ... [, expressionN]
replicates expression1’s value expression2 times. This directive
automatically aligns its data and preceding labels to a word
boundary. You can disable this feature by using .align 0.
Sets a mask with a bit turned on for each floating point register
that the current routine saved. The least-significant bit
corresponds to register $f0. The offset is the distance in bytes
from the virtual frame pointer at which the floating point registers
.fmask mask offset
are saved. The assembler saves higher register numbers closer to
the virtual frame pointer. You must use .ent before .fmask and
only one .fmask may be used per .ent. Space should be allocated
Chapter 8

for those registers specified in the .fmask.

Describes a stack frame. The first register is the frame-register,
the offset is the distance from the frame register to the virtual
frame pointer, and the second register is the return program
.frame frame-register offset counter (or, if the first register is $0, this directive shows that the
return_pc_register return program counter is saved four bytes from the virtual frame
pointer). You must use .ent before .frame and only one .frame
may be used per .ent. No stack traces can be done in the
debugger without .frame.
Makes the name external. If the name is otherwise defined (by its
appearance as a label), the assembler will export the symbol;
.globl name otherwise it will import the symbol. In general, the assembler
imports undefined symbols (that is, it gives them the UNIX storage
class “global undefined” and requires the linker to resolve them).
For use by compilers. Sets the masks defining the registers whose
.gjaldef int_bitmask fp_bitmask value is preserved during a procedure call. See Table 1-1 and
Table 7-1 for the default for integer saved registers.
For use by compilers. Sets the default masks for live registers
.gjallive int_bitmask fp_bitmask
before a procedure call (A JAL instruction).
For use by compilers. Sets the default masks for live registers
.gjrlive int_bitmask fp_bitmask
before a procedure’s return (A JR instruction).

8-4 Assembly Language Programmer’s Guide

 Pseudo Op-Codes

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
This directive is similar to .word except that the relocation entry
for local-sym has the R_MIPS_GPREL32 type. After linkage, this
.gpword local-sym results in a 32-bit value that is the distance between local-sym and
“gp”. local-sym must be local. This directive is used by the code
generator for PIC switch tables.
Truncates the expressions in the comma-separated list to 16-bit
values and assembles the values in successive locations. The
expressions must be absolute. This directive can optionally have
.half expression1 [ , expression2 ]
the form: expression1 [ : expression2 ]. The expression2
... [ , expressionN]
replicates expression1’s value expression2 times. This directive
automatically aligns its data appropriately. You can disable this
feature by using .align 0.
Associates a named label with the current location in the program
.lab label_name
text. (For use by compilers).
Makes the name’s data type bss. The assembler allocates the
named symbol to the bss area, and the expression defines the
named symbol’s length. If a .globl directive also specifies the
name, the assembler allocates the named symbol to external bss.
.lcomm name, expression The assembler puts bss symbols in one of two bss areas. If the

Chapter 8
defined size is smaller than (or equal to) the size specified by the
assembler or compiler’s –G command line option, the assembler
puts the symbols in the sbss area and uses $gp to address the
data.
For use by compilers. Affects the next jump instruction even if it is
not the successive instruction. The .livereg directive may come
before any of the following instructions: JAL, JR, and SYSCALL.
By default, external J instructions and JR instructions through a
register other than $ra, are treated as external calls; that is; all
registers are assumed live. The directive .livereg cannot appear
before an external J (it will affect the next JR, JAL, or SYSCALL
instead of the J instruction). .livereg may appear before a JR
instruction through a register other than $ra. The directive can’t be
used before a BREAK instruction. For BREAK instructions, the
.livereg int_bitmask fp_bitmask
assembler also assumes all registers are live.

.livereg notes to the assembler which registers are live before a

jump, in order to avoid unsafe optimizations by the reorganizer.
The directive .livereg takes two arguments, int_bitmask, and
fp_bitmask, which are 32 bit bitmasks with a bit turned on for each
register that is live before a jump. The most significant bit
corresponds to register $0 (which is opposite to that used in other
assembly directives, .mask, .fmask). The first bitmap indicates live
integer registers and the second indicates live FPs.

Assembly Language Programmer’s Guide 8-5

 Chapter 8

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Specifies the source file and the line within that file that
corresponds to the assembly instructions that follow. The
assembler ignores the file number when this directive appears in
the assembly source file. Then, the assembler assumes that the
directive refers to the most recent .file directive. When a .loc
.loc file_number line_number
directive appears in the binary assembly language .G file, the file
number is a dense number pointing at a file symbol in the symbol
table .T file. For more information about .G and .T files, see the
RISCompilers and C Programmer’s Guide. (For use by
compilers).
Sets a mask with a bit turned on for each general purpose register
that the current routine saved. Bit one corresponds to register $1.
The offset is the distance in bytes from the virtual frame pointer
where the registers are saved. The assembler saves higher
.mask mask, offset register numbers closer to the virtual frame pointer. Space should
be allocated for those registers appearing in the mask. If bit zero is
set it is assumed that space is allocated for all 31 registers
regardless of whether they appear in the mask. (For use by
compilers).
Chapter 8

Register1 and register2, when used as indexed registers to

memory will never point to the same memory. The assembler will
.noalias reg1, reg2 use this as a hint to make more liberal assumptions about
resource dependency in the program. To disable this assumption,
see .alias.
Tells the assembler to put in an instruction that has no effect on
the machine state. While several instructions cause no-operation,
the assembler only considers the ones generated by the nop
nop directive to be wait instructions. This directive puts an explicit
delay in the instruction stream. Note: Unless you use “.set
noreorder”, the reorganizer may eliminate unnecessary “nop”
instructions.
Tells the assembler that certain options were in effect during
compilation. (These options can, for example, limit the
.option options assembler’s freedom to perform branch optimizations.) This option
is intended for compiler-generated .s files rather than for hand-
coded ones.
Repeats all instructions or data between the .repeat directive and
the .endr directive. The expression defines how many times the
.repeat expression data repeats. With the .repeat directive, you cannot use labels,
branch instructions, or values that require relocation in the block.
To end a .repeat, see .endr.
.rdata Tells the assembler to add subsequent data into the rdata section.
.sdata Tells the assembler to add subsequent data to the sdata section.

8-6 Assembly Language Programmer’s Guide

 Pseudo Op-Codes

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Instructs the assembler to enable or to disable certain options.
Use .set options only for hand-crafted assembly routines. The
assembler has these default options: reorder, macro, and at. You
can specify only one option for each .set directive. You can
specify these .set options:\

The reorder option lets the assembler reorder machine language

instructions to improve performance. The noreorder option
prevents the assembler from reordering machine language
instructions. If a machine language instruction violates the
hardware pipeline constraints, the assembler issues a warning
message.
.set option
The bopt/nobopt option lets the assembler perform branch
optimization. This involves moving an instruction that is the target
of a branch or jump instruction into the delay slot; this is performed
only if no unpredictable side effects can occur.

The macro option lets the assembler generate multiple machine

instructions from a single assembler instruction.

Chapter 8
The nomacro option causes the assembler to print a warning
whenever an assembler operation generates more than one
machine language instruction. You must select the noreorder
option before using the nomacro option; otherwise, an error
results.
The at option lets the assembler use the $at register for macros,
but generates warnings if the source program uses $at. When you
use the noat option and an assembler operation requires the $at
register, the assembler issues a warning message; however, the
noat option does let source programs use $at without issuing
warnings.

The nomove option tells the assembler to mark each subsequent

instruction so that it cannot be moved during reorganization.
.set option (continued)
Because the assembler can still insert nop instructions where
necessary for pipeline constraints, this option is less stringent than
noreorder. The assembler can still move instructions from below
the nomove region to fill delay slots above the region or vice
versa. The nomove option has part of the effect of the “volatile” C
declaration; it prevents otherwise independent loads or stores
from occurring in a different order than intended.

The move option cancels the effect of nomove.

Assembly Language Programmer’s Guide 8-7

 Chapter 8

Table 8-1: Pseudo Op-Codes

Pseudo-Op Description
Advances the location counter by the value of the specified
.space expression
expression bytes. The assembler fills the space with zeros.
This permits you to lay out a structure using labels plus directives
like .word, .byte, and so forth. It ends at the next segment directive
.struct expression (.data, .text, etc.). It does not emit any code or data, but defines
the labels within it to have values which are the sum of expression
plus their offsets from the .struct itself.
Takes one of these forms: name = expression or name = register.
You must define the name only once in the assembly, and you
CANNOT redefine the name. The expression must be computable
(symbolic equate)
when you assemble the program, and the expression must involve
operators, constants, and equated symbols. You can use the
name as a constant in any later statement.
Tells the assembler to add subsequent code to the text section.
.text
(This is the default.)
Specifies the major and minor version numbers (for example,
.verstamp major minor
version 0.15 would be .verstamp 0 15).
(For use by compilers). Describes a register variable by giving the
Chapter 8

.vreg register offset symno offset from the virtual frame pointer and the symbol number
symno (the dense number) of the surrounding procedure.
Truncates the expressions in the comma-separated list to 32-bits
and assembles the values in successive locations. The
expressions must be absolute. The operands can optionally have
.word expression1 [, expression2
the form: expression1 [ : expression2 ]. The expression2
] ... [ , expressionN]
replicates expression1’s value expression2 times. This directive
automatically aligns its data and preceding labels to a word
boundary. You can disable this feature by using .align 0.

8-8 Assembly Language Programmer’s Guide

MIPS Object File Format

9
This chapter provides information on the object file format and has the
following major topics:
• An overview of the components that make up the object file, and the
differences between the MIPS object-file format and the UNIX
System V common object file format (COFF).
• A description of the headers and sections of the object file. Detailed
information is given on the logic followed by the assembler and link
editor in handling relocation entries.
• The format of object files (OMAGIC, NMAGIC, ZMAGIC, and
LIBMAGIC), and information used by the system loader in loading
object files at run-time.
• Archive files and link editor defined symbols.

Chapter 9

Assembly Language Programmer’s Guide 9-1

 Chapter 9

Overview
The assembler and the link editor generate object files that have sections
ordered as shown in Figure 9-1. Any areas empty of data are omitted, except
that the File Header, Optional Header, and Section Header are always present.
The sections of the Symbol table portion (indicated in Figure 9-1) that appear
in the final object file format vary, as follows:
• The Line Numbers, Optimization Symbols, and Auxiliary Symbols
tables appear only when debugging is on (when the user specifies one
of the compiler –g1, –g2 or –g3 options).
• When the user specifies the –x option (strip non-globals) for the link
edit phase, the link editor strips the Line Number, Local Symbols,
Optimization Symbols, Auxiliary Symbols, Local Strings, and
Relative File Descriptor tables from the object file, and updates the
Procedure Descriptor table.
• The link editor strips the entire Symbol table from the object file
when the user specifies the –s option (strip) for the link edit phase.
Any new assembler or link editor designed to work with the compiler system
should lay out the object file sections in the order shown in Figure 9-1. The
link editor can process object files that are ordered differently, but
performance may be degraded.
Chapter 9

9-2 Assembly Language Programmer’s Guide

 MIPS Object File Format

File Header

Optional Headers

Section Headers

Section Data
text small data
initialization text small bss (0 size)
read-only data large bss (0 size)
large data shared library info.
8-byte literal pool ucode (ucode ob-
4-byte literal pool jects only)
*
Section Relocation Information
text large data
read-only data small data
Symbolic Header

Comments

Line Numbers*
Dense Numbers
(ucode objects only)
Procedure Descriptor Table

Local Symbols
Optimization Symbols*

Chapter 9
Auxiliary Symbols*

Local Strings

External Strings

File Descriptor * Created only if debugging is ON.

Relative File Descriptor Symbol Table. Missing

if fully stripped.
External Symbols Missing if stripped of
non-globals.

Figure 9-1: Object File Format

Assembly Language Programmer’s Guide 9-3

 Chapter 9

Readers already familiar with standard UNIX System V COFF (common

object file format) may be interested in the differences between it and the
MIPS compiler system format, as described next.
The compiler system File Header definition is based on UNIX System V
header file filehdr.h with the following modifications.
• The symbol table file pointer and the number of symbol table entries
now specify the file pointer and the size of the Symbolic Header
respectively (described in Chapter 10).
• All tables that specify symbolic information have their file pointers
and number of entries in this Symbolic Header.
The Optional Header definition has the same format as specified in the UNIX
System V header file aouthdr.h, except the following fields have been added:
bss_start, gprmask, cprmask, and gp_value. See Table 9-4.
The Section Header definition has the same format as the UNIX System V’s
header file scnhdr.h. except the line number fields are used for global
pointers. See Table 9-5.
The relocation information definition is similar to UNIX 4.3 BSD, which has
local relocation types; however, you should read the Section Relocation
Information section in this chapter for information on differences.

The File Header

The format of the File Header, defined in filehdr.h, is shown in
Table 9-1.
Table 9-1: File Header Format

Declaration Field Description

Chapter 9

unsigned short f_magic; Magic number

unsigned short f_nscns; Number of sections
long f_timdat; Time and date stamp
long f_symptr; File pointer to symbolic header
long f_nsyms; Size of symbolic header
unsigned short f_opthdr; Size of optional header
unsigned short f_flags; Flags

f_symptr points to the Symbolic Header of the Symbol table, and f_nsyms
gives the size of the header. For a description of the Symbolic Header, see
Chapter 10.

9-4 Assembly Language Programmer’s Guide

 MIPS Object File Format

File Header Magic Field (f_magic)

The magic number in the f_magic entry in the File Header specifies the target
machine on which an object file can execute. Table 9-2 shows the values and
mnemonics for the magic numbers; the header file filehdr.h contains the
macro definitions.

Table 9-2: File Header Magic Numbers

Symbol Value Description

MIPSEBMAGIC 0x0160 Big-endian target (headers and tables
MIPSEBMAGIC_2 0x0163 have same byte order as host machine).
MIPSEBMAGIC_3 0x0140
MIPSELMAGIC 0x0162 Little-endian target (headers and
MIPSELMAGIC_2 0x0166 tables have same byte order as host
MIPSELMAGIC_3 0x0142 machine).
SMIPSEBMAGIC 0x6001 Big-endian target (headers and tables
SMIPSEBMAGIC_2 0x6301 have opposite byte order as host
SMIPSEBMAGIC_3 0x4001 machine).
SMIPSELMAGIC 0x6201 Little-endian target (headers and
SMIPSELMAGIC_2 0x6601 tables have opposite byte order as host
SMIPSELMAGIC_3 0x4201 machine).
MIPSEBUMAGIC 0x0180 MIPS big-endian ucode object file.
MIPSELUMAGIC 0x0182 MIPS little-endian ucode object file.

NOTE: The “_2” magic numbers are defined for mips2 object files. They
cannot be used on a mips1 implementation.
NOTE: The “_3” magic numbers are defined for mips3 object files. They
cannot be used on a mips1 or mips2 implementation.

Chapter 9
Flags (f_flags)
The f_flags field describes the object file characteristics. Table 9-3 describes
the flags and gives their hexadecimal values. The table notes those flags that
do not apply to compiler system object files.

Assembly Language Programmer’s Guide 9-5

 Chapter 9

Table 9-3: File Header Flags

Symbol Value Description

F_RELF bLG 0x0001 Relocation information stripped from file
F_EXEC 0x0002 File is executable (i.e. no unresolved external references).
F_LNNO 0x0004 Line numbers stripped from file.
F_LSYMS 0x0008 Local symbols stripped from file.
F_MINMAL 0x0010 !Minimal object file (”.m”) output of fextract.
F_UPDATE 0x0020 !Fully bound update file, output of ogen.
F_SWABD 0x0040 !File whose bytes were swabbed (in names).
F_AR16WR 0x0080 !File has the byte ordering of an AR16WR. (e.g.11/70) machine
(it was created there, or was produced by conv).
F_AR32WR 0x0100 !File has the byte ordering of an AR32WR machine (e.g. vax).
F_AR32W 0x0200 !File has the byte ordering of an AR32W machine (e.g.
3b,maxi,MC68000).
F_PATCH 0x0400 !File contains “patch” list in Optional Header.
F_NODF 0x0400 !(Minimal file only) no decision functions for replaced functions.
F_MIPS_NO_SHARED 010000 Cannot be dynamically shared.
F_MIPS_SHARABLE 020000 A dynamically shared object.
F_MIPS_CALL_SHARED 030000 Dynamic executable.
F_MIPS_NO_REORG 040000 Do not reorder sections.
F_MIPS_UGEN_ALLOC 0100000 File whose procedures contain pre-allocated space on the stack
that makes conversion to PIC easier.
!Not used by compiler system object modules.
Chapter 9

9-6 Assembly Language Programmer’s Guide

 MIPS Object File Format

Optional Header
The link editor and the assembler fill in the Optional Header, and the system
(kernel) loader (or other program that loads the object module at run-time)
uses the information it contains, as described in the section Loading Object
Files in this chapter.
Table 9-4 shows the format of the Optional Header (defined in the header file
aouthdr.h).

Table 9-4: Optional Header Definition

Declaration Field Description

short magic; See Table 9-5.
short vstamp; Version stamp.
long tsize; Text size in bytes, padded to 16-byte
boundary.
long dsize; Initialized data in bytes, padded to 16-byte
boundary.
long bsize; Uninitialized data in bytes, padded to 16-
byte boundary.
long entry; Entry point.
long text_start; Base of text used for this file.
long data_start; Base of data used for this file.
long bss_start; Base of bss used for this file.
long gprmask; General purpose register mask.
long cprmask[4]; Co-processor register masks.
long gp_value; The gp value used for this object.

Chapter 9

Assembly Language Programmer’s Guide 9-7

 Chapter 9

Optional Header Magic Field (magic)

Table 9-5 shows the values of the magic field for the Optional Header; the
header file aouthdr.h contains the macro definitions.

Table 9-5: RISC/os Magic Numbers

Symbol Value Description

OMAGIC 0407 Impure Format. The text is not write-protected or
sharable; the data segment is contiguous with the
text segment.
NMAGIC 0410 Shared Text. The data segment starts at the next
page following the text segment and the text
segment is write-protected.
ZMAGIC 0413 The object file is to be demand loaded and has a
special format; the text and data segments are
separated. Text segment is also write protected.
(The MIPS default). The object may be either
dynamic or static.
LIBMAGIC 0443 The object file is a target shared library to be
demand loaded and file has a special format like that
of a ZMAGIC file.

See the Object Files section in this chapter for information on the format of
OMAGIC, NMAGIC, ZMAGIC, and LIBMAGIC files.

Section Headers
Table 9-6 shows the format of the Section Header (defined in the header file
scnhdr.h).

Table 9-6: Section Header Format

Chapter 9

Declaration Field Description

char s_name[8]; Section name.
long s_paddr; Physical address.
long s_vaddr; Virtual address.
long s_size; Section size.
long s_scnptr; File pointer to raw data for section.
long s_relptr; File pointer to relocation.
long s_lnnoptr; File pointer to gp (global pointer) tables.
unsigned short s_nreloc; Number of relocation entries.
unsigned short s_nlnno; Number of gp tables.
long s_flags; Flags.

9-8 Assembly Language Programmer’s Guide

 MIPS Object File Format

Section Name (s_name)

Table 9-7 shows the defined section names for the s_name field of the Section
Header; the header file scnhdr.h contains the macro definitions.

Table 9-7: Section Header Constants for Section Names

Declaration Field Description

_TEXT ”.text” Text section.
_INIT ”.init” Initialization text section for shared libraries.
_FINI ”.fini” Cleanup text section.
_RDATA ”.rdata” Read only data section.
_DATA ”.data” Large data section.
_LIT8 ”.lit8” 8 byte literal pool section.
_LIT4 ”.lit4” 4 byte literal pool section.
_SDATA ”.sdata” Small data section.
_BSS ”.bss” Large bss section.
_SBSS ”.sbss” Small bss section.
_LIB ”.lib” Shared library information section
_UCODE ”.ucode” ucode section.
_GOT ”.got” * Global offset table.
_DYNAMIC ”.dynami * Dynamic linking information.
c”
_DYNSYM ”.dynsym * Dynamic linking symbol table.
”
_REL_DYN ”.rel.dyn” * Relocation information.
_DYNSTR ”.dynstr” * Dynamic linking strings.
_HASH ”.hash” * Symbol hash table.

Chapter 9
_DSOLIST ”.dsolist” * Dynamic shared object list table.
_CONFLICT ”.conflict” * Additional dynamic linking information.
_REGINFO ”.reginfo” * Register usage information.
* these sections exist only in ZMAGIC type files and are used during
dynamic linking

Assembly Language Programmer’s Guide 9-9

 Chapter 9

Flags (s_flags)
Table 9-8 shows the defined values for s_flags; the header file scnhdr.h
contains the definitions (those flags that are not used by compiler system
object files are noted).

Table 9-8: Format of s_flags Section Header Entry

Symbol Value Description

STYP_REG 0x00 Regular section; allocated, relocated, loaded.
STYP_DSECT 0x01 !Dummy; not allocated, relocated, not loaded.
STYP_NOLOAD 0x02 !Noload; allocated, relocated, not loaded.
STYP_GROUP 0x04 !Grouped; formed of input sections.
STYP_PAD 0x08 !Padding; not allocated, not relocated, loaded.
STYP_COPY 0x10 !Copy; for decision function used by field update; not allocated,
not relocated, loaded; relocated, and line number entries
processed normally.
STYP_TEXT 0x20 Text only.
STYP_DATA 0x40 Data only.
STYP_BSS 0x80 Contains bss only.
STYP_RDATA 0x100 Read only data only.
STYP_SDATA 0x200 Small data only.
STYP_SBSS 0x400 Contains small bss only.
STYP_UCODE 0x800 Section contains ucode only.
STYP_LIT4 0x10000000 Section 4 byte literals only.
S_NRELOC_OVFL 0x20000000 s_nreloc overflowed, the value is in r_vaddr of the first entry.
STYP_LIB 0x40000000 Section contains shared library information only.
STYP_INIT 0x80000000 Section initialization text only.
STYP_FINI 0x01000000 .fini section text.
STYP_COMMENT 0x02100000 Comment section.
Chapter 9

STYP_LIT8 0x08000000 Section 8 byte literals only.

STYP_CONFLICT 0x00100000 Additional linking information.
STYP_DSOLIST 0x00040000 Dynamic shared object list table.
STYP_HASH 0x00020000 Symbol has table.
STYP_DYNSTR 0x00010000 Dynamic linking strings.
STYP_GOT 0x00001000 Global offset table.
STYP_DYNAMIC 0x00002000 Dynamic linking information section.
STYP_DYNSYM 0x00004000 Dynamic linking symbol table.
STYP_REL_DYN 0x0008000 Relocation information for runtime linker.
!Not used by compiler system object modules.

S_NRELOC_OVFL is used when the number of relocation entries in a

section overflows the s_nreloc field of the section header. In this case,

9-10 Assembly Language Programmer’s Guide

 MIPS Object File Format

s_nreloc contains the value 0xffff and the s_flags field has the
S_NRELOC_OVFL flag set; the value true is in the r_vaddr field of the first
relocation entry for that section. That relocation entry has a type of R_ABS
and all other fields are zero, causing it to be ignored under normal
circumstances.
NOTE: For performance reasons, the link editor uses the s_flags entry
instead of s_name to determine the type of section. However, the link editor
does correctly fill in the s_name entry.

Global Pointer Tables

The gp (global pointer) tables are part of the object file that is produced by
the assembler. These are used by the link editor in calculating the best –G num
to compile the objects are specified as recompilable by the –count option.
There is a gp table for the .sdata and .bss sections only.
The gp table gives the section size corresponding to each applicable value
specified by the –G num option (always including 0), sorted by smallest size
first. The s_lnnoptr field in the section header points to this value and the
s_nlnno field contains the number of entries (including the header). If there is
no small section, the related gp table is attached to the corresponding large
section to provide the link editor with this information.
When an object does not contain a data and bss section, the –G num option
specified for the object at compilation is unknown. Because the size of the
literal pools cannot be known, this complicates the calculation of a best –G
num. However, a reliable calculation can be made when there is an 8-byte
literal pool, which ensures that the object was compiled with a –G of at least
eight.
The global pointer table has the following format:

Chapter 9
union gp_table {
struct {
long current_g_value; /* actual value */
long unused;
} header;
struct {
long g_value; /* hypothetical value */
long bytes; /* section size corresponding */
/* to hypothetical value */
} entry;
};

Assembly Language Programmer’s Guide 9-11

 Chapter 9

Shared Library Information

The .lib section contains the shared libraries used by executable objects. The
absence of a .lib section header indicates that no shared libraries are used.
Shared libraries are a feature of System V Release 3; thus, only objects
compiled with –systype sysv should contain .lib sections. The field s_nlib in
the section header is defined to be the same as s_paddr and contains the
number of shared library entries in the .lib section. The shared library
information definition shown below defines a compiler system .lib section
entry. Note the size and offset are in sizeof(long)’s not bytes. The size (in
bytes) of each entry must be a multiple of SCNROUND. The name the offset
field refers to is a C null-terminated string.
struct libscn {
long size; /* size of this entry (including */
/* target name)*/
long offset; /* offset from start of entry */
/* to target name*/
long tsize; /* text size in bytes*/
long dsize; /* initialized data size in bytes */
long bsize; /* uninitialized data size in bytes
*/
long text_start; /* base of text used for */
/* this library*/
long data_start; /* base of data used for */
/* this library */
long bss_start; /* base of bss used for */
/* this library */
/* pathname of target shared library */
};

Section Data
Chapter 9

RISCompiler system files are represented by the following sections:

.dynamic, .liblist, .rel.dyn, .conflict, .dynstr, .dynsym, .hash, .rdata (read-
only data), .text, .init (shared library initialization text), .fini (process
termination text), .data (data), .lit8 (8-byte literal pool), .lit4 (4-byte literal
pool), .sdata (small data), .sbss (small block started by storage), .bss (block
started by storage), .lib (shared library information), and .ucode (intermediate
code). Table 9-2 shows the layout of the sections.
The .dynamic, .liblist, .rel.dyn, .conflict, .dynstr, .dynsym, and .hash sections
exist only in ZMAGIC files and are used during dynamic linking. These
sections are described in more detail in Chapter 11. Dynamic linking is
discussed in Chapter 12.

9-12 Assembly Language Programmer’s Guide

 MIPS Object File Format

The .text section contains the machine instructions that are to be executed; the
.rdata, .data, .lit8, .lit4, and .sdata contain initialized data, and the .sbss and
.bss sections reserve space for uninitialized data that is created by the kernel
loader for the program before execution and filled with zeros.

.dynamic
.liblist
.rel.dyn
.conflict
.dynstr
.dynsym text segment
.has

.rdata
.text
.init
.fini
.data
.lit8
.lit4 data segment

.sdata
.got

Chapter 9
.sbss
bss segment
.bss

Figure 9-2: Organization of Section Data

As noted in Figure 9-2, the sections are grouped into the text segment
(containing the .text, .init, and .fini sections), the data segment (.rdata, .data,
.lit8, .lit4, and .sdata), and the bss segment (.sbss and .bss). A section is
described by and referenced through the Section Header; the Optional Header
provides the same information for segments.

Assembly Language Programmer’s Guide 9-13

 Chapter 9

The link editor references the data shown in Figure 9-2 both as sections and
segments, through the Section Header and Optional Header respectively.
However, the system (kernel) loader, when loading the object file at run-time,
references the same data only by segment, through the Optional Header.
Chapter 9

9-14 Assembly Language Programmer’s Guide

 MIPS Object File Format

Section Relocation Information

Relocation Table Entry
Table 9-9 shows the format of an entry in the Relocation Table (defined in the
header file reloc.h).

Table 9-9: Format of a Relocation Table Entry

Declaration Field Description

long r_vaddr; (Virtual) address of an item to be relocated.
unsigned r_symndx:24, Index into external symbols or section numbers; see r_extern below.
r_reserved:3,
r_type:4, Relocation type.
r_extern:1; = 1 for an external relocation entry; r_symndx is an index into External
Symbols. = 0 for a local, relocation entry; r_symndx is the number of the
section containing the symbol.

Symbol Index (r_symndx) and Extern Field (r_extern)

For external relocation entries, r_extern is set to 1 and r_symnndx is the index
into External Symbols for this entry. In this case, the value of the symbol is
used as the value for relocation.
For local relocation entries, r_extern is set to 0, and r_symndx contains a
constant that refers to a section. In this case, the starting address of the section
to which the constant refers is used as the value for relocation.
Table 9-10 gives the section numbers for r_symndx; the reloc.h file contains
the macro definitions.

Table 9-10: Section Numbers for Local Relocation Entries

Chapter 9
Symbol Value Description
R_SN_TEXT 1 .text section.
R_SN_INIT 7 .init section.
R_SN_RDATA 2 .rdata section.
R_SN_DATA 3 .data section.
R_SN_SDATA 4 .sdata section.
R_SN_SBSS 5 .sbss section.
R_SN_BSS 6 .bss section.
R_SN_LIT 8 .lit8 section.
R_SN_LIT4 9 .lit4 section.
R_SN_FINI 12 .fini section.

Assembly Language Programmer’s Guide 9-15

 Chapter 9

Relocation Type (r_type)

Table 9-11 shows valid symbolic entries for the relocation type (r_type) field
(defined in the header file reloc.h)

Table 9-11: Relocation Types

Symbol Value Description

R_ABS 0x0 Relocation already performed.
R_REFHALF 0x1 16-bit reference to the symbol’s virtual address.
R_REFWORD 0x2 32-bit reference to the symbol’s virtual address.
R_JMPADDR 0x3 26-bit jump reference to the symbol’s virtual address.
R_REFHI 0x4 Reference to the high 16-bits of symbol’s virtual address.
R_REFLO 0x5 Reference to the low 16-bits of symbol’s virtual address.
R_GPREL 0x6 Reference to the offset from the global pointer to the symbol’s virtual
address.
R_LITERAL 0x7 Reference to a literal in a literal pool as an offset from the global
pointer.
R_REL32 0x8 Reference to the distance from the offset to the global pointer.
0x9-0x15 Reserved
0x1D-0x1F Reserved

Assembler and Link Editor Processing

Compiler system executable object modules with all external references
defined have the same format as relocatable modules and are executable
without re-link editing.
Local relocation entries must be used for symbols that are defined. Therefore,
external relocations are used only for undefined symbols. Figure 9-3 gives an
Chapter 9

overview of the Relocation Table entry for an undefined external symbol.

9-16 Assembly Language Programmer’s Guide

 MIPS Object File Format

Relocation Table Entry

r_vaddr External Symbols
r_symndx

value=0
r_extern=1

Section Data Sign-extended to 32 bits.

constant

Figure 9-3: Relocation Table Entry for Undefined External Symbols

The assembler creates this entry as follows:

Sets r_vaddr to point to the item to be relocated.
Places a constant to be added to the value for relocation at the address
for the item to be relocated (r_vaddr).
Sets r_symndx to the index of the External Symbols entry that
contains the symbol value.
Sets r_type to the constant for the type of relocation types. Table 9-11
shows the valid constants for the relocation type.

Chapter 9
Sets r_extern to 1.
NOTE: The assembler always sets the value of the undefined entry in
External Symbols to 0. It may assign a constant value to be added to the
relocated value at the address where the location is to be done. If the width of
the constant is less than a full word, and an overflow occurs after relocation,
the link editor flags this as an error.
When the link editor determines that an external symbol is defined, it changes
the Relocation Table entry for the symbol to a local relocation entry. Figure
9-4 gives an overview of the new entry.

Assembly Language Programmer’s Guide 9-17

 Chapter 9

Relocation Table Entry Section n Optional

r_vaddr Header
r_symndx

r_type s_vaddr
r_extern=0

Section Data Section n Data

constant symbol location

Sign-extended to 32 bits.

Figure 9-4: Relocation Table Entry for a Local Relocation Entry

To change this entry from an external relocation entry to a local relocation

entry, the link editor:
• Picks up the constant from the address to be relocated (r_vaddr).
• If the width of the constant is less than 32 bits, sign-extends the
constant to 32 bits.
• Adds the value for relocation (the value of the symbol) to the
constant and places it back in the address to be relocated.
Chapter 9

• Sets r_symndx to the section number that contains the external

symbol.
• Sets r_extern to 0.

9-18 Assembly Language Programmer’s Guide

 MIPS Object File Format

Examples
The examples that follow use external relocation entries.
Example 1: 32-Bit Reference – R_REFWORD. This example shows
assembly statements that set the value at location b to the global data value y.
.globl y
.data
b: .word y #R_REFWORD relocation type at address b for
symbol y

In processing this statement, the assembler generates a relocation entry of

type R_REFWORD for the address b and the symbol y. After determining the
address for the symbol y, the loader adds the 32-bit address of y to the 32-bit
value at location b, and places the sum in location b. The loader handles 16-
bit addresses (R_REFHALF) in the same manner, except it checks for
overflow after determining the relocation value.
Example 2: 26-Bit Jump – R_JMPADDR. This example shows assembly
statements that call routine x from location c.
.text
x: #routine x
...
c: jal x #R_JMPADDR relocation type at address c for symbol
x

In processing these statements, the assembler generates a relocation entry of

type R_JMPADDR for the address and the symbol x. After determining the
address for the routine, the loader shifts the address right two bits, adds the
low 26 bits of the result to the low 26 bits of the instruction at address c (after

Chapter 9
sign-extending it), and places the results back into the low 26 bits at address c.
R_JMPADDR relocation entries are produced for the assembler’s j (Jump)
and jal (Jump and Link) instructions. These instructions take the high four
bits of the target address from the address of the delay slot of their instruction.
The link editor makes sure that the same four bits are in the target address
after relocation; if not, it generates an error message.
If the entry is a local relocation type, the target of the Jump instruction is
assembled in the instruction at the address to be relocated. The high four bits
of the jump target are taken from the high 4 bits of the address of the delay
slot of the instruction to be relocated.
Example 3: High/Low Reference - R_REFHI/R_REFLO. This example
shows an assembler macro that loads the absolute address y, plus a constant,
into Register 6:

Assembly Language Programmer’s Guide 9-19

 Chapter 9

In processing this statement, the assembler generates a 0 as the value y, and

the following machine language statements:
f: lui $at,constant>>16 #R_REFHI relocation type at
address f for symbol y
g: addiu $r6,constant&0xffff($at) #R_REFLO relocation
type at address for symbol y

In this example, the assembler produces two relocation entries.

NOTE: When a R_REFHI relocation entry appears, the next relocation entry
must always be the corresponding R_REFLO entry. This is required in order
to reconstruct the constant that is to be added to the value for relocation.
In determining the final constant values for the two instructions, the link
editor must take into account that the addiu instruction of the R_REFLO
relocation entry sign-extends the immediate value of the constant.
In determining the sum of the address for the symbol y and the constant, the
link editor does the following:
• It uses the low 16 bits of this sum for the immediate value of the
R_REFLO relocation address.
• Because all instructions that are marked with a R_REFLO perform a
signed operation, the assembler adjusts the high portion of the sum if
Bit 15 is set. Then it uses the high 16 bits of the sum for the
immediate value of the R_REFHI instruction at the relocation
address. For example:

lw $r6,0x10008000
Chapter 9

lui $at,0x1001 at = 0x10010000

+ 0xFFFF8000
lw $r6,0x8000($at)
0x10008000

Example 4: Offset Reference – R_GPREL. This example shows an

assembly macro that loads a global pointer relative value y into Register 6:
lw $r6,y
In processing this statement, the assembler generates a 0 as the value y and
the following machine language statement:
h: lw $r6,0($gp) #R_GPREL relocation type at

9-20 Assembly Language Programmer’s Guide

 MIPS Object File Format

address h for symbol y

and a R_GPREL relocation entry would be produced. The assembler then
uses the difference between the address for the symbol y and the address of
the global pointer, as the immediate value for the instruction. The link editor
gets the value of the global pointer used by the assembler from gp_value in
the Optional Header (Table 9-4).
Example 4: Example of the R_LITERAL. This example shows of an
R_LITERAL uses a floating-point literal. The assembler macro:
li.s $f0,1.234
is translated into the following machine instruction:
h: lwc1 $f0,–32752(gp)# R_LITERAL relocation
type at address h for the
literal 1.234
and a R_LITERAL relocation entry is produced; the value of the literal is put
into the .lit4 section. The link editor places only one of all like literal
constants in the literal pool. The difference between the virtual address of the
literal and the address of the global pointer is used as the immediate value for
the instruction. The link editor handles 8-byte literal constants similarly,
except it places each unique constant in the .lit8 section. The value of the –G
num option used when compiling determines if the literal pools are used.

Object Files
This section describes the object-file formats created by the link editor,
namely the Impure (OMAGIC), Shared Text (NMAGIC), Demand Paged
(ZMAGIC), and target-shared libraries (LIBMAGIC) formats. Before
reading this section, you should be familiar in the format and contents of the
text, data, and bss segments as described in the Section Data section of this

Chapter 9
chapter.
NOTE: This chapter discusses the creation of LIBMAGIC files (shared
libraries). These are not to be confused with dynamic shared objects that have
type ZMAGIC. Dynamic shared objects are discussed in Chapters 11 and 12.
The following constraints are imposed on the address at which an object can
be loaded and the boundaries of its segments. The operating system can
dictate additional constraints.
• Segments must not overlap and all addresses must be less than
0x80000000.
• Space should be reserved for the stack, which starts below
0x80000000 and grows through lower addresses; that is, the value of
each subsequent address is less than that of the previous address.

Assembly Language Programmer’s Guide 9-21

 Chapter 9

• The default text segment address for ZMAGIC and NMAGIC files is
0x0n0400000 and the default data segment address is 0x10000000.
• The default text segment address for OMAGIC files is 0x10000000
with the data segment following the text segment.
• The –B num option (specifying a bss segment origin) cannot be
specified for OMAGIC files; the default, which specifies that the bss
segment follow the data segment, must be used.
• RISC/os requires a 2-megabyte boundary for segments.

Impure Format (OMAGIC) Files

An OMAGIC file has the format shown in Figure 9-5.

.bss
bss segment
.sbss
.sdata

.lit4
.lit8 data segment
.data

.rdata
.init
text segment
.text
Chapter 9

aligned on a 16-byte boundary

Figure 9-5: Layout of OMAGIC Files in Virtual Memory

The OMAGIC format has the following characteristics:

• Each section follows the other in virtual address space aligned on an
16-byte boundary.
• No blocking of sections.
• Text, data and bss segments can be placed anywhere in the virtual
address space using the link editor’s -T, -D and -B options.

9-22 Assembly Language Programmer’s Guide

 MIPS Object File Format

• The addresses specified for the segments must be rounded to 16-byte

boundaries.
• The text segment contains the .text, and .init sections.
• The sections in the data segment are ordered as follows: .rdata, .data,
.lit8, .lit4, and .sdata.
• The sections in the bss segment are ordered as follows: .sbss and .bss.

Shared Text (NMAGIC) Files

An NMAGIC file has the format shown in Figure 9-6.

.bss
bss segment
.sbss
.sdata

.lit4
data segment
.lit8
.data
.init
text segment
.fini
.text

Chapter 9
.rdata

aligned on a 16-byte boundary

aligned on a page-size boundary

Figure 9-6: Layout of NMAGIC Files in Virtual Memory

An NMAGIC file has the following characteristics:

• The virtual address of the data segment is on a pagesize boundary.
• No blocking of sections.
• Each section follows the other in virtual address space aligned on an
16-byte boundary.

Assembly Language Programmer’s Guide 9-23

 Chapter 9

• Only the start of the text and data segments, using the link editor’s –
T and –D options, can be specified for a shared text format file; the
start of the text and data segments must be a multiple of the pagesize.

Demand Paged (ZMAGIC) Files

A ZMAGIC file is a demand paged file in the format shown in Figure 9-7.
A ZMAGIC file has the following characteristics:
• The text segment and the data segment are blocked, with pagesize as
the blocking factor. Blocking reduces the complexity of paging in the
files.
• The size of the sum of the of the File, Optional, and Sections Headers
(Table 9-1, Table 9-4, and Table 9-6) rounded to 16 bytes is included
in blocking of the text segment.
• The text segment starts by default at 0x400000 (4 Mbyte), plus the
size of the sum of the headers again rounded to 16 bytes. With the
standard software, the text segment starts at 0x400000 + header size.
NOTE: This is required because the first 32K bytes of memory are reserved
for future use by the compiler system to allow data access relative to the
constant register 0.
• Only the start of the text and data segments, using the link editor’s –
T and –D options can be specified for a demand paged format file
and must be a multiple of the pagesize.
Chapter 9

9-24 Assembly Language Programmer’s Guide

 MIPS Object File Format

2G 32K (not accessible

upage by user).
2G-32K
Stack Area
bottom of stack
sbrk arena increases automatically
as required.
.bss
bss segment
0 fill area .sbss
0 area fill

.sdata
data segment
.lit4 (blocked by pagesize)

.lit8

256M .data
empty
fill area

.init
.text
.rdata

text segment
(blocked by pagesize)

Chapter 9

4 Mbyte + header
header
4 Mbyte
empty
0

Figure 9-7: Layout of ZMAGIC Files in Virtual Memory

Assembly Language Programmer’s Guide 9-25

 Chapter 9

Figure 9-8 shows a ZMAGIC file as it appears in a disk file.

Symbol Table
0 Fill Area
.got

.sdata data segment

(blocked by pagesize)
.lit4
.lit8
.data
fill area
.fini

.init
.text

.rdata
.hash

.dynsym text segment

.dynstr (blocked by pagesize)
.msym
.conflict
Chapter 9

.rel.dyn
.liblist
.dynamic

headers

Figure 9-8: Layout of a ZMAGIC File on Disk

9-26 Assembly Language Programmer’s Guide

 MIPS Object File Format

Target Shared Library (LIBMAGIC) Files

Typically, mkshlib(1) creates target shared libraries. The link editor creates
such libraries when its –c option is specified (each shared library file name is
displayed during the link if the –v option is supplied).
LIBMAGIC files are demand paged and have the same format as ZMAGIC
file except as follows:
• Headers are put on their own page
• The text section starts on the next page from the value of the
–T num option. This prevents the number and size of headers from
affecting the start of the first real text. The first real text is the branch
table and must stay at the same address.
Both the –T and –D options should be specified, because the defaults would
cause the target shared library to overlay the ZMAGIC files and cause an
execution failure. The link editor –c option requires that the files to be linked
are compiled with the –G 0 option (which sets the link editor –G 0 option).
NOTE: Shared library refers to System V Release 3 type shared libraries.
Elsewhere, we use “shared objects” or “dynamically linked executable” for
shared libraries in the sense similar to System V Release 4 type shared
libraries.

Objects Using Shared Libraries

Object files that use shared libraries contain a .lib section following the data
segment (including the zero fill area created by blocking it to a pagesize). All
object file contain an .init section used by shared library initialization code.
Shared library initialization instructions are generated by mkshlib(1) from the
#init directive in the library specification file. This following code from the

Chapter 9
shared library specification
#initbar.o
_libfoo_extext
generates these instructions generated in the .init section:
la $2,ext
sw $2,_libfoo_ext
Initialization instructions are not bounded by any procedure; the initialization
instructions from each .init section are concatenated and the runtime startup
(crt1.o) branches to its label in its .init section. Then the execution falls
through all the concatenated .init sections until reaching crtn.o (the last object
with a .init section) which contains the RETURN instruction.

Assembly Language Programmer’s Guide 9-27

 Chapter 9

Object files without shared libraries contain a small .init section that executes,
producing no significant results.

Ucode objects
Ucode objects contain only a file header, the ucode section header, the ucode
section and all of the symbolic information. A ucode section never appears in
a machine code object file.

Loading Object Files

The link editor produces object files with their sections in a fixed order
similar to UNIX system object files that existed before COFF. See Figure 9-
1 for the a description of the sections and how they are formatted.
The sections are grouped into segments, which are described in the Optional
Header. In loading the object module at run-time, the system (kernel) loader
needs only the magic number in the File Header and the Optional Header to
load an object file for execution.
The starting addresses and sizes of the segments for all types of object files
are specified similarly, and they are loaded in the same manner.
After reading in the File Header and the Optional Header, the system (kernel)
loader must examine the file magic number to determine if the program can
be loaded. Then, the system (kernel) loader loads the text and data segments.
The starting offset in the file for the text segment is given by the macro
N_TXTOFF(f,a)
in the header file a.out.h, where f is the File Header structure and a is the
option header structure for the object file to be loaded. The tsize field in the
Chapter 9

Optional Header (Table 9-4) contains the size of the text segment and
text_start contains the address at which it is to be loaded.
The starting offset of the data segment follows the text segment. The dsize
field in the Section Header (Table 9-6) contains the size of the data segment;
data_start contains the address at which it is to be loaded.
The system (kernel) loader must fill the .bss segment with zeros. The
bss_start field in the Optional Header specifies the starting address; bsize
specifies the number of bytes to be filled with zeros. In ZMAGIC files, the
link editor adjusts bsize to account for the zero filled area it created in the data
segment that is part of the .sbss or .bss sections.
If the object file itself does not load the global pointer register it must be set
to the gp_value field in the Optional Header (Table 9-4).

9-28 Assembly Language Programmer’s Guide

 MIPS Object File Format

The other fields in the Optional Header are gprmask and cprmask[4], whose
bits show the registers used in the .text, .init , and .fini sections. They can be
used by the operating system, if desired, to avoid save register relocations on
context-switch.

Archive files
The link editor can link object files in archives created by the archiver. The
archiver and the format of the archives are based on the UNIX System V
portable archive format. To improve performance, the format of the archives
symbol table was changed so that it is a hash table, not a linear list.
The archive hash table is accessed through the ranhashinit() and ranlookup()
library routines in libmld.a, which are documented in the manual page
ranhash(3x). The archive format definition is in the header file ar.h.

Chapter 9

Assembly Language Programmer’s Guide 9-29

 Chapter 9

Link Editor Defined Symbols

Certain symbols are reserved and their values are defined by the link editor.
A user program can reference these symbols, but can not define one; an error
is generated if a user program attempts to define one of these symbols. Table
9-12 lists the names and values of these symbols; the header file syms.h
contains their preprocessor macro definitions.
Table 9-12: Link Editor Defined Symbols

Symbol Value Description

_ETEXT ”etext” 1st location after .text
_EDATA ”edata” 1st location after .sdata (all
initialized data)
_END ”end” 1st location after .bss (all data)
_FTEXT ”_ftext” !1st location of .text
_FDATA ”_fdata” !1st location of .data
_FBSS ”_fbss” !1st location of the .bss
_PROCEDURE_TABLE ”_procedure_table” runtime procedure table
_PROCEDURE_TABLE_SIZE ”_procedure_table_size” runtime procedure table size
_PROCEDURE_STRING_TABLE ”_procedure_string_table” string table for runtime procedure
_COBOL_MAIN ”_cobol_main” 1st cobol main symbol
_GP ”_gp” !the value of the global pointer
__UNWIND “__unwind” Unwinds the stack.
__PC_NLC_GOTO “__pc_nlc_goto” Handles Pascal non- local goto
commands.
__FIND_RDP “__find_rdp” Finds runtime procedure tables.
__GP_DISPLACEMENT “__gp_disp” ! Distance between the global
pointer and the instructions that
Chapter 9

references this symbol.

!compiler system only

The dynamic linker also reserves and defines certain symbols; see Chapters
11 and 12 for more information.
The first three symbols come from the standard UNIX system link editors and
the rest are compiler system specific. The last symbol is used by the start up
routine to set the value of the global pointer, as shown in the following
assembly language statements:
globl _GP
la $gp,_GP

9-30 Assembly Language Programmer’s Guide

 MIPS Object File Format

The assembler generates the following machine instructions for these

statements:
a: lui gp,0# R_REFHI relocation type at address a for symbol _GP
b: add gp,0 # R_REFLO relocation type at address b for symbol
_GP

which would cause the correct value of the global pointer to be loaded.
The link editor symbol _COBOL_MAIN is set to the symbol value of the first
external symbol with the cobol_main bit set. COBOL objects uses this
symbol to determine the main routine.

Runtime Procedure Table Symbols

The five link editor defined symbols, _PROCEDURE_TABLE,
__FIND_RDP __PC_NLC_GOTO, __UNWIND,
_PROCEDURE_TABLE_SIZE and _PROCEDURE_STRING_TABLE,
relate to the runtime procedure table. The Runtime Procedure Table is used
by the exception systems in ADA, PL/I and COBOL. Its description is found
in the header file sym.h. The table is a subset of the Procedure Descriptor
Table portion of the Symbol Table with one additional field, exception_info.
When the procedure table entry is for an external procedure, and an External
Symbol Table exists, the link editor fills in exception_info with the address of
the external table. Otherwise, its fill in exception_info with zeros.
The name of the External Symbol Table is the procedure name concatenated
with the string _exception_info (actually, the preprocessor macro
EXCEPTION_SUFFIX as defined in the header file exception.h).
The Runtime Procedure Table provides enough information to allow a
program to unwind its stack. It is typically used by the routines in libexc.a.

Chapter 9
The comments in the header file exception.h describes the routines in that
library.
The Runtime Procedure Table is sorted by procedure address and always has
a dummy entry with a zero address and a 0xffffffff address. When required,
the table is padded with an extra zero entry to ensure that the total number of
entries is an uneven (odd) number.
The Runtime Procedure Table and String Table for the runtime procedure
table are placed at then end of the .data section in the object file.

Assembly Language Programmer’s Guide 9-31

 Chapter 9
Chapter 9

9-32 Assembly Language Programmer’s Guide

Symbol Table

10
This chapter describes the symbol table and symbol table routines used to
create and make entries in the table. The chapter contains the following major
sections:
• Overview, which gives the purpose of the Symbol table, a summary
of its components, and their relationship to each other.
• Format of Symbol Table Entries, which shows the structures of
Symbol table entries and the values you assign them through the
Symbol Table routines.
• Symbol Table Routine Reference, which lists the symbol table
routines supplied with the compiler and summarizes the function of
each.
NOTE: Third Eye Software, Inc. owns the copyright (dated 1984) to the
format and nomenclature of the Symbol Table used by the compiler system
as documented in this chapter.
Third Eye Software, Inc. grants reproduction and use rights to all parties,
PROVIDED that this comment is maintained in the copy.
Third Eye makes no claims about the applicability of this symbol table to a
particular use.
Chapter 10

Assembly Language Programmer’s Guide 10-1

 Chapter 10

Overview
The symbol table in created by the compiler front-end as a stand-alone file.
The purpose of the table is to provide information to the link editor and the
debugger in performing their respective functions. At the option of the user,
the link editor includes information from the Symbol table in the final object
file for use by the debugger. See Figure 9-1 in Chapter 9 for details.

Symbolic Header

Comment Section*

Dense Numbers

Procedure Descriptor Table

Local Symbols

Optimization Symbols

Auxiliary Symbols

Local Strings

External Strings

File Descriptor

Relative File Descriptor

Created only if debugging
is ON. External Symbols

1 table per compilation. 1 table per source file and

per include file.

* This section holds compacted form of relocation entries for Pixie.

Chapter 10

Currently only dynamically linked executables and shared objects

have this section.

Figure 10-1: The Symbol Table - Overview

10-2 Assembly Language Programmer’s Guide

 Symbol Table

The elements that make up the Symbol table are shown in Figure 10-1. The
front-end creates one group of tables (the shaded areas in Figure 10-1) that
contain global information relative to the entire compilation. It also creates a
unique group of tables (the unshaded areas in the figure) for the source file
and each of its include files.
Compiler front-ends, the assembler, and the link editor interact with the
symbol table as summarized below:
• The front-end, using calls to routines supplied with the compiler
system, enters symbols and their descriptions in the table.
• The assembler fills in line numbers, optimization symbols, updates
Local Symbols and External Symbols, and updates the Procedure
Descriptor table.
• The link editor eliminates duplicate information in the External
Symbols and the External Strings tables, removes tables with
duplicate information, updates Local Symbols with relocation
information, and creates the Relative File Descriptor table.
The major elements of the table are summarized in the paragraphs that follow.
Some of these elements are explored in more detail later in the chapter.
Symbolic Header. The Symbolic Header (HDRR for HeadDeR Record)
contains the sizes and locations (as an offset from the beginning of the file)
of the subtables that make up the Symbol Table. Figure 10-2 shows the
symbolic relationship of the header to the other tables.

Chapter 10

Assembly Language Programmer’s Guide 10-3

 Chapter 10

Symbolic Header
Line Numbers
Dense Numbers
Procedure Descriptor Table
Local Symbols
Optimization Symbols
Auxiliary Symbols
Local Strings
External Strings
File Descriptor Table

Figure 10-2: Functional Overview of the Symbolic Header

Line Numbers. The assembler creates the Line Number table. It creates an
entry for every instruction. Internally, the information is stored in an encoded
form. The debugger uses the entries to map instruction to the source lines and
vice versa.
Dense Numbers. The Dense Number table is an array of pairs. An index into
this table is called a dense number. Each pair consists of a file table index (ifd)
and an index (isym) into Local Symbols. The table facilitates symbol look-up
for the assembler, optimizer, and code generator by allowing direct table
access rather than hashing.
Procedure Descriptor Table. The Procedure Descriptor table contains
register and frame information, and offsets into other tables that provide
detailed information on the procedure. The front-end creates the table and
links it to the Local Symbols table. The assembler enters information on
registers and frames. The debugger uses the entries in determining the line
numbers for procedures and frame information for stack traces.
Local Symbols. The Local Symbols table contains descriptions of program
variables, types, and structures, which the debugger uses to locate and
Chapter 10

interpret runtime values. The table gives the symbol type, storage class, and
offsets into other tables that further define the symbol.

10-4 Assembly Language Programmer’s Guide

 Symbol Table

A unique Local Symbols table exists for every source and include file; the
compiler locates the table through an offset from the file descriptor entry that
exists for every file. The entries in Local Symbols can reference related
information in the Local Strings and Auxiliary Symbols subtables. This
relationship is shown in Figure 10-3.
.

File Descriptor Table

Entry for File 0
Entry for File 1 Local Symbols

Entry for File 2

Local Strings
Entry for File n
Auxiliaries

Local Strings

Auxiliaries

Local Strings

Auxiliaries

Figure 10-3: Logical Relationship between the File Descriptor Table and
Local Symbols

Optimization Symbols. To be defined at a future date.

Auxiliary Symbols. The Auxiliary Symbols tables contain data type
information specific to one language. Each entry is linked to an entry in Local
Symbols. The entry in Local Symbols can have multiple, contiguous entries.
Chapter 10

The format of an auxiliary entry depends on the symbol type and storage
class. Table entries are required only when the compiler debugging option is
ON.

Assembly Language Programmer’s Guide 10-5

 Chapter 10

Local Strings. The Local Strings subtables contain the names of local
symbols.
External Strings. The External Strings table contains the names of external
symbols.
File Descriptor. The File Descriptor table contains one entry each for each
source file and each of its include files. (The structure of an entry is given in
Table 10-14 later in this chapter.) The entry is composed of pointers to a
group of subtables related to the file. The physical layout of the subtables is
shown in Figure 10-4.

File Descriptor Table

File Descriptor Entry
Line Numbers
Procedure Descriptor Table

Local Symbols
Optimization Symbols
Auxiliary Symbols
Local Strings
Relative File Descriptor

Figure 10-4: Physical Relationship of a File Descriptor Entry to Other

Tables

The file descriptor entry allows the compiler to access a group of subtables
unique to one file. The logical relationship between entries in this table and
in its subtables is shown in Figure 10-5.
Chapter 10

10-6 Assembly Language Programmer’s Guide

 Symbol Table

File Descriptor Table

Entry for File 0
Entry for File 1 Line Numbers
Entry for File 2
Procedure Descriptor Table
-
- Local Symbols
Optimization Symbols
Auxiliary Symbols
Local Strings
Relative File Descriptor

Line Numbers
Procedure Descriptor Table

Local Symbols
Optimization Symbols
Auxiliary Symbols
Local Strings
Relative File Descriptor

Figure 10-5: Logical Relationship between the File Descriptor Table and
Other Tables

Relative File Descriptor. See the section Link Editor Processing later in this
chapter.
Chapter 10

External Symbols. The External Symbols contains global symbols entered

by the front-end. The symbols are defined in one module and referenced in
one or more other modules. The assembler updates the entries, and the link
editor merges the symbols and resolves their addresses.

Assembly Language Programmer’s Guide 10-7

 Chapter 10

Format of Symbol Table Entries

Symbolic Header
The structure of the Symbolic Header is shown below in Table 10-1; the
sym.h header file contains the header declaration.

Table 10-1: Format of the Symbolic Header

Declaration Name Description

short magic To verify validity of the table
short vstamp Version stamp
long ilineMax Number of line number entries
long cbLine Number of bytes for line number entries
long cbLineOffset Index to start of line numbers
long idnMax Max index into dense numbers
long cbDnOffset Index to start dense numbers
long ipdMax Number of procedures
long cbPdOffset Index to procedure descriptors
long isymMax Number of local symbols
long cbSymOffset Index to start of local symbols
long ioptMax Maximum index into optimization entries
long cbOptOffset Index to start of optimization entries
long iauxMax Number of auxiliary symbols
long cbAuxOffset Index to the start of auxiliary symbols
long issMax Max index into local strings
long cbSsOffset Index to start of local strings
long issExtMax Max index into external strings
long cbSsExtOffset Index to the start of external strings
long ifdMax Number of file descriptors
long cbFdOffset Index to file descriptor
long crfd Number of relative file descriptors
long cbRfdOffset Index to relative file descriptors
Chapter 10

long iextMax Maximum index into external symbols

long cbExtOffset Index to the start of external symbols.

10-8 Assembly Language Programmer’s Guide

 Symbol Table

The lower byte of the vstamp field contains LS_STAMP and the upper byte
MS_STAMP (see the stamp.h header file). These values are defined in the
stamp.h file. The iMax fields and the cbOffset field must be set to 0 if one of
the tables shown in Table 10-1 isn’t present. The magic field must contain the
constant magicSym, also defined in longsymconst.h.

Line Numbers
Table 10-2 shows the format of an entry in the Line Numbers table; the sym.h
header file contains its declaration.

Table 10-2: Format of a Line Number Entry

Declaration Name
typedef long LINER, *pLINER

The line number section in the Symbol table is rounded to the nearest four-
byte boundary.
Line numbers map executable instructions to source lines; one line number is
stored for each instruction associated with a source line. It is stored as a long
integer in memory and in packed format on disk.
The layout on disk is as follows:

Bit 7 4 0

Delta Count

The compiler assigns a line number to only those lines of source code that
generate one or more executable instructions.
Delta is a four-bit value in the range –7...7, defining the number of source
lines between the current source line, and the previous line generating
executable instructions. The Delta of the first line number entry is the
displacement from the lnLow field in the Procedure Descriptor Table.
Chapter 10

Count is a four-bit field with a value in the range 0...15 indicating the number
(1...16) of executable instructions associated with a source line. If more than
16 instructions (15+1) are associated with a source line, new line number
entries are generated with Delta = 0.

Assembly Language Programmer’s Guide 10-9

 Chapter 10

An extended format of the line number entry is used when Delta is outside the
range of –7...7.
The layout of the extended field on disk is as follows:

Bit 7 4 0
1 0 0 0

Constant 78 Count
Bit 7 4 0

Upper eight bits of Delta

Bit 7 4 0

Lower eight bits of Delta

NOTE: The compiler allows a maximum of 32,767 comment lines, blank

lines, continuation lines and other lines not producing executable
instructions, between two source lines that do produce executable
instructions.
Line number example. This section gives an example of how the compiler
assigns line numbers. For the source listing shown below, the compiler
generates line numbers only for the highlighted lines (6, 7, 17, 18, and 19);
the other lines are either blank or contain comments.
Chapter 10

10-10 Assembly Language Programmer’s Guide

 Symbol Table

1 #include <stdio.h>
2 main ()
3 {
4 char c;
5
6 printf ("this program just prints its input\n");
7 while ((c = getc(stdin)) != EOF) {
8 /* this is a greater than a seven line comment
9 * 1
10 * 2
11 * 3
12 * 4
13 * 5
14 * 6
15 * 7
16 */
17 printf("%c", c);
18 } /* end while */
19 } /* end main */

Figure 10-6: Source Listing for Line Number Example

Figure 10-7 shows the instructions generated for lines 3, 7, 17, 18, and 19.
Table 10-3 shows the compiler-generated liner entries for each source line.

Table 10-3: Source Listing for Line Number Example

Liner
Source Line
Contents Meaning
3 02 delta 0, count 2
6 31 delta 3, count 1
7 1f delta 1, count 15
7 03 delta 0, count 3
1
17 82 00 0a -81, count 2, delta 10
18 1f delta 1, count 15
Chapter 10

182 03 delta 02, count 3

19 15 delta 1, count 5
1
Extended format (count is greater than seven lines).
2
Continuation.

Assembly Language Programmer’s Guide 10-11

 Chapter 10

[main:3, 0x4001a0] addiu sp,sp,-32

[main:3, 0x4001a4] sw r31,20(sp)
[main:3, 0x4001a8] sw r16,16(sp) 3 ,6
[main:6, 0x4001ac] jal printf
[main:6, 0x4001b0] addiu r4,gp,-32752
[main:7, 0x4001b4] lw r14,-32552(gp)
[main:7, 0x4001b8] nop
[main:7, 0x4001bc] addiu r15,r14,-1
[main:7, 0x4001c0] bltz r15,0x4001e4
[main:7, 0x4001c4] sw r15,-32552(gp)
[main:7, 0x4001c8] lw r24,-32548(gp)
[main:7, 0x4001cc] nop
[main:7, 0x4001d0] lbu r25,0(r24)
[main:7, 0x4001d4] addiu r8,r24,1 7
[main:7, 0x4001d8] sb r25,31(sp)
[main:7, 0x4001dc] b 0x4001f4
[main:7, 0x4001e0] sw r8,-32548(gp)
[main:7, 0x4001e4] jal _filbuf
[main:7, 0x4001e8] addiu r4,gp,-32552
[main:7, 0x4001ec] move r16,r2
[main:7, 0x4001f0] sb r16,31(sp)
[main:7, 0x4001f4] lbu r9,31(sp)
[main:7, 0x4001f8] li r1,-1
[main:7, 0x4001fc] beq r9,r1,0x400260
[main:7, 0x400200] nop
[main:17, 0x400204] lbu r5,31(sp)
[main:17, 0x400208] jal printf 17
[main:17, 0x40020c] addiu r4,gp,-32716
[main:18, 0x400210] lw r10,-32552(gp)
[main:18, 0x400214] nop
[main:18, 0x400218] addiu r11,r10,-1
[main:18, 0x40021c] bltz r11,0x400240
[main:18, 0x400220] sw r11,-32552(gp)
[main:18, 0x400224] lw r12,-32548(gp)
[main:18, 0x400228] nop
[main:18, 0x40022c] lbu r13,0(r12)
[main:18, 0x400230] addiu r14,r12,1 18
[main:18, 0x400234] sb r13,31(sp)
[main:18, 0x400238] b 0x400250
[main:18, 0x40023c] sw r14,-32548(gp)
[main:18, 0x400240] jal _filbuf
[main:18, 0x400244] addiu r4,gp,-32552
[main:18, 0x400248] move r16,r2
[main:18, 0x40024c] sb r16,31(sp)
[main:18, 0x400250] lbu r15,31(sp)
[main:18, 0x400254] li r1,-1
[main:18, 0x400258] bne r15,r1,0x400204
[main:18, 0x40025c] nop
[main:19, 0x400260] b 0x400268
[main:19, 0x400264] nop
Chapter 10

[main:19, 0x400268] lw r31,20(sp)

[main:19, 0x40026c] lw r16,16(sp) 19
[main:19, 0x400270] jr r31
[main:19, 0x400274] addiu sp,sp,32

Figure 10-7: Source Listing for Line Number Example

10-12 Assembly Language Programmer’s Guide

 Symbol Table

Procedure Descriptor Table

Table 10-4 shows the format of an entry in the Procedure Descriptor table; the
sym.h header file contains its declaration.

Table 10-4: Format of a Procedure Descriptor Table Entry

Declaration Name Description

unsigned, long adr Memory address of start of procedure
long isym Start of local symbols
long iline Procedure’s line numbers *
long regmask Saved register mask
long regoffset Saved register offset
long iopt Procedure’s optimization symbol entries
long fregmask Save floating point register mask
long fregoffset Save floating point register offset
long frameoffset Frame size
long framereg Frame pointer register
long pcreg Index or reg of return program counter
long lnLow Lowest line in the procedure
long lnHigh Highest line in the procedure
Byte offset for this procedure from the
long cbLineOffset
base of the file descriptor entry.

*If NULL, and cycm field in file descriptor table = 0, then this field is indexed
to the actual table.

Local Symbols
Table 10-5 shows the format of an entry in the Local Symbols table; the sym.h
header file contains its declaration.

Table 10-5: Format of a Local Symbols Entry

Declaration Name Description

long iss Index into local strings of symbol name
long value Value of symbol. See Table 10-7
Chapter 10

unsigned st : 6 Symbol type. See Table 10-8

unsigned sc : 5 Storage class. See Table 10-9
unsigned reserved : 1
Index into local or auxiliary symbols See
unsigned index : 20
Table 3,5.

Assembly Language Programmer’s Guide 10-13

 Chapter 10

The meanings of the fields in a local symbol entry are explained in the
following paragraphs.
iss. The iss (for index into string space) is an offset from the issBase field of
an entry in the file descriptor table, to the name of the symbol.
value. An integer representing an address, size, offset from a frame pointer.
The value is determined by the symbol type, as illustrated in Table 10-6.
st and sc. The symbol type (st) defines the symbol; the storage class (sc),
where applicable explains how to access the symbol type in memory. The
valid st and sc constants are given in Table 10-8 and Table 10-9. These
constants are defined in symconst.h.
index. The index is an offset into either Local Symbols or Auxiliary Symbols,
depending of the storage type (st) as shown in Table 10-6. The compiler uses
isymBase in the file descriptor entry as the base for a Local Symbol entry and
iauxBase for an Auxiliary Symbols entry.

Table 10-6: Index and Value as a Function of Symbol Type and Storage
Class

Symbol Type Storage Class Index Value

stFile scText isymMac address
stLabel scText indexNil address
1
stGlobal scD/B iaux address
stStatic scD/B1 iaux address
stParam scAbs iaux frame offset2
scRegister iaux register number
scVar iaux frame offset2
scVarRegister iaux register number
stLocal scAbs iaux frame offset2
scRegister iaux register number
stProc scText iaux address
scNil iaux address
scUndefined iaux address
stStaticProc scText iaux address
stMember
Chapter 10

enumeration scInfo indexNil ordinal

structure scInfo iaux bit offset3
union scInfo iaux bit offset
stBlock

10-14 Assembly Language Programmer’s Guide

 Symbol Table

Table 10-6: Index and Value as a Function of Symbol Type and Storage
Class

Symbol Type Storage Class Index Value

enumeration scInfo symMac4 max enumeration
structure scInfo symMac size
text bock scText symMac relative address5
common block scCommon symMac size
variant scVariant symMac isymTag6
variant arm scInfo symMac iauxRanges7
union scInfo symMac size
stEnd
enumeration scInfo isymStart8 0
file scText isymStart relative address5
procedure scText isymStart relative address5
structure scInfo isymStart 0
textblock scText isymStart relative address5
union scInfo isymStart 0
common block scCommon isymStart 0
variant scVariant isymStart 0
variant arm scInfo isymStart 0
stTypedef scInfo iaux 0
1 scD/B is the storage class determined by the assembler, either large/small or data/bss.
2 frame offset is the offset from the virtual frame pointer.
3 bit offset is computed from the beginning of the procedure.
4 isymMac is the isym of the corresponding stEnd symbol plus 1.
5 relative address is the relative displacement from the beginning of the procedure.
6 isymTab is the isym to the symbol that is the tag for the variant.
7 iauxRanges is the iaux to ranges for the variant arm.
8 isymStart is the isym of the corresponding begin block (stBlock, stFile, stProc, etc.).

The link editor ignores all symbols except the types stProc, stStatic, stLabel,
stStaticProc, which it will relocate. Other symbols are used only by the
debugger, and need be entered in the table only when the compiler debugger
option is ON.
Chapter 10

Assembly Language Programmer’s Guide 10-15

 Chapter 10

Symbol Type (st). Table 10-7 gives the allowable constants that can be
specified in the st field of Local Symbols entries; the symconst.h header file
contains the declaration for the constants.

Table 10-7: Symbol Type (st) Constants Supported by the Compiler

Constant Value Description

stNil 0 Dummy entry
stGlobal 1 External symbol
stStatic 2 Static
stParam 3 Procedure argument
stLocal 4 Local variable
stLabel 5 Label
stProc 6 Procedure
stBlock 7 Start of block
stEnd 8 End block, file, or procedures
stMember 9 Member of structure, union, or enumeration
stTypedef 10 Type definition
stFile 11 File name
stStaticProc 14 Load time only static procs
stConstant 15 Const.

Storage Class (st) Constants. Table 10-8 gives the allowable constants that
can be specified in the sc field of Local Symbols entries; the symconst.h
header file contains the declaration for the constants.

Table 10-8: Storage Class Constants Supported by the Compiler

Constant Value Description

scNil 0 Dummy entry
scText 1 Text symbol
scData 2 Initialized data symbol
scBss 3 Un-initialized data symbol
scRegister 4 Value of symbol is register number
scAbs 5 Symbol value is absolute; not to be relocated
scUndefined 6 Used but undefined in the current module
Chapter 10

reserved 7
scBits 8 This is a bit field
scDbx 9 Dbx internal use
scRegImage 10 Register value saved on stack
scInfo 11 Symbol contains debugger information

10-16 Assembly Language Programmer’s Guide

 Symbol Table

Table 10-8: Storage Class Constants Supported by the Compiler

Constant Value Description

scUserStruct 12 Address in struct user for current process
scSData 13 (Load time only) small data
scSBss 14 (Load time only) small common
scRData 15 (Load time only) read only data
scVar 16 Var parameter (Fortran or Pascal)
scCommon 17 Common variable
scSCommon 18 Small common
scVarRegister 19 Var parameter in a register
scVariant 20 Variant records
scUndefined 21 Small undefined
scInit 22 Init section symbol.

Optimization Symbols
Reserved for future use.

Auxiliary Symbols
Table 10-9 shows the format of an entry, which is a union, in Auxiliary
Symbols; the sym.h file contains its declaration.

Table 10-9: Storage Class Constants Supported by the Compiler

Declaration Name Description

TIR ti Type information record
RNDXR rndx Relative index into local symbols
long dnLow Low dimension
long dnHigh High dimension
long isym Index into local symbols for stEnd
long iss Index into local strings (not used)
Width of a structure field not declared with the
long width
default value for size
long count Count of ranges for variant arm
Chapter 10

All of the fields except the ti field are explained in the order they appear in
the above layout. The ti field is explained last.

Assembly Language Programmer’s Guide 10-17

 Chapter 10

rndx. Relative File Index. The front-end fills this field in describing
structures, enumerations, and other complex types. The relative file index is
a pair of indexes. One index is an offset from the start of the File Descriptor
table to one of its entries. The second is an offset from the file descriptor entry
to an entry in the Local Symbols or Auxiliary Symbols table.
dnLow. Low Dimension of Array.
dnHigh. High Dimension of Array.
isym. Index into Local Symbols. This index is always an offset to an stEnd
entry denoting the end of a procedure.
width. Width of Structured Fields.
count. Range Count. Used in describing case variants. Gives how many
elements are separated by commas in a case variant.
ti. Type Information Record. Table 10-10 shows the format of a ti entry; the
sym.h file contains its declaration.

Table 10-10: Format of an Type Information Record Entry

Declaration Name Description

unsigned fBitfield : 1 Setif bit width is specified
unsigned continued : 1 Next auxiliary entry has tq info
unsigned bt : 6 Basic type
unsigned tq4 : 4 Type qualifier
unsigned tq5 : 4
unsigned tq0 : 4
unsigned tq1 : 4
unsigned tq2 : 4
unsigned tq3 : 4

All groups of auxiliary entries have a type information record with the
following entries:
• fbitfield – Set if the basic type (bt) is of non-standard width.
• bt (for basic type) specifies if the symbol is integer, real complex,
numbers, a structure, etc. The valid entries for this field are shown in
Chapter 10

Table 10-11; the sym.h file contains its declaration.

• tq (for type qualifier) defines whether the basic type (bt) has an array
of, function returning, or pointer to qualifier. The valid entries for
this field are shown in Table 10-12; the sym.h file contains its
declaration.

10-18 Assembly Language Programmer’s Guide

 Symbol Table

Table 10-11: Basic Type (bt) Constants

Constant Value Default Size* Description

btNil 0 0 Undefined, void
btAdr 1 32 Address – same size as pointer
btChar 2 8 Symbol character
btUChar 3 8 Unsigned character
btShort 4 16 Short (16 bits)
btUShort 5 16 Unsigned short
btInt 6 32 Integer
btUInt 7 32 Unsigned integer
btLong 8 32 Long (32 bits)
btULong 9 32 Unsigned long
btFloat 10 32 floating-point (real)
btDouble 11 64 Double-precision floating-point real
btStruct 12 n/a Structure (Record)
btUnion 13 N/A union (variant)
btEnum 14 32 Enumerated
btTypedef 15 n/a Defined via a typedef; rndx points at a stTypedef symbol
btRange 16 32 Sub-range of integer
btSet 17 32 Pascal sets
btComplex 18 64 FORTRAN complex
btDComplex 19 128 FORTRAN double complex
btIndirect 20 Indirect definition; rndx points to TIR aux
btMax 64
*Size in bits.

Table 10-12: Type Qualifier (tq) Constants

Constant Value Description

tqNil 0 Place holder. No qualifier
tqPtr 1 Pointer to
tqProc 2 Function returning
tqArray 3 Array of
Chapter 10

tqVol 5 Volatile
tqMax 8

Assembly Language Programmer’s Guide 10-19

 Chapter 10

File Descriptor Table

Table 10-13 shows the format of an entry in the File Descriptor table; the
sym.h file contains its declaration.

Table 10-13: Format of File Descriptor Entry

Declaration Name Description

unsigned, long adr Memory address of start of file
long rss Source file name
long issBase Start of local strings
long cbSs Number of bytes in local strings
long isymBase Start of local symbol entries
long csym Count of local symbol entries
long ilineBase Start of line number entries
long cline Count of line number entries
long ioptBase Start of optimization symbol entries
long copt Count of optimization symbol entries
short ipdFirst Start of procedure descriptor table
short cpd Count of procedures descriptors
long iauxBase Start of auxiliary symbol entries
long caux Count of auxiliary symbol entries
long rfdBase Index into relative file descriptors
long crfd Relative file descriptor count
unsigned lang : 5 Language for this file
unsigned fMerge : 1 Whether this file can be merged
unsigned fReadin : 1 True if it was read in (not just created)
If set, was compiled on big endian machine aux’s is in compile
unsigned fBigendian : 1
host’s sex
unsigned reserved : 22 Reserved for future use
long cbLineOffset Byte offset from header or file ln’s
long cbLine
Chapter 10

10-20 Assembly Language Programmer’s Guide

 Symbol Table

External Symbols
The External Symbols table has the same format as Local Symbols, except an
offset (ifd) field into the File Descriptor table has been added. This field is
used to locate information associated with the symbol in an Auxiliary
Symbols table. Table 10-14 shows the format of an entry in External
Symbols; the sym.h file contains its declaration.

Table 10-14: Format an Entry in External Symbols

Declaration Name Description

unsigned, long adr Memory address of start of file
long rss Source file name
long issBase Start of local strings
long cbSs Number of bytes in local strings
long isymBase Start of local symbol entries
long csym Count of local symbol entries
long ilineBase Start of line number entries
long cline Count of line number entries
long ioptBase Start of optimization symbol entries
long copt Count of optimization symbol entries
short ipdFirst Start of procedure descriptor table
short cpd Count of procedures descriptors
long iauxBase Start of auxiliary symbol entries
long caux Count of auxiliary symbol entries
long rfdBase Index into relative file descriptors
long crfd Relative file descriptor count
unsigned lang : 5 Language for this file
unsigned fMerge : 1 Whether this file can be merged
unsigned fReadin : 1 True if it was read in (not just created)
If set, was compiled on big endian machine aux’s is in compile
unsigned fBigendian : 1
host’s sex
unsigned reserved : 22 Reserved for future use
long cbLineOffset Byte offset from header or file ln’s
Chapter 10

long cbLine
short reserved Reserved for future use
short ifd Pointer to file descriptor entry
SYMR asym Same as Local Symbols.

Assembly Language Programmer’s Guide 10-21

 Chapter 10
Chapter 10

10-22 Assembly Language Programmer’s Guide

Execution and Linking Format

Chapter 11
11
This chapter describes the Execution and Linking Format (ELF) for object
files. The following topics are covered:
• The Components of an elf object file
• Symbol Table Format
• Global Data Area
• Register Information
• Relocation
Program loading and dynamic linking are discussed in Chapter 12.
There are three types of object files:
• Relocatable files contain code and data and are linked with other
object files to create an executable file or shared object file.
• Executable files contain a program that can be executed.
• Shared object files contain code and data that can be linked. These
files may be linked with relocatable or shared object files to create
other object files. They may also be linked with an executable file
and other shared objects to create a process image.

Assembly Language Programmer’s Guide 11-1

 Chapter 11

Object File Format

An object file is organized as follows:
Chapter 11

ELF header

optional Program Header Table

Section 1

.
.
.

Section n

Section Header Table

Each object file begins with an ELF header that describes the file. Sections
contain information that is used when the file is linked with other objects (e.g.
code, data, relocation information). The Section Header Table contains
information describing the sections of the file and has an entry for each
section. Files that will be linked with other objects must contain a Section
Header Table.
If the Program Header Table is present, it contains information that is used to
create a process image. Files used to build an executable program must have
a Program Header Table; relocatable files do not need one.

11-2 Assembly Language Programmer’s Guide

 Execution and Linking Format

ELF Header
The ELF header has the following format:

Chapter 11
Declaration Field
unsigned char e_ident[EI_NIDENT];
Elf32_Half e_type;
Elf32_Half e_machine;
Elf32_Word e_version;
Elf32_Addr e_entry;
Elf32_Off e_phoff;
Elf32_Off e_shoff;
Elf32_Word e_flags;
Elf32_Half e_ehsize;
Elf32_Half e_phentsize;
Elf32_Half e_phnum;
Elf32_Half e_shentsize;
Elf32_Half e_shnum;
Elf32_Half e_shstrndx;
#define EI_NIDENT 16

e_ident
contains machine-independent data concerning the file contents.
The index values for the e_ident member are:
EI_MAG0 0 File identification.
EI_MAG1 1 File identification.
EI_MAG2 2 File identification.
EI_MAG3 3 File identification.
EI_CLASS 4 File class.
EI_DATA 5 Byte order.
EI_VERSION 6 File version.
EI_PAD 7 Start of padding bytes.
EI_NIDENT 16 Size of e_ident[].

e_ident[EI_MAGO ... EI_MAG3]

contain a magic number identifying the file as an ELF object file:
Position Value Name
e_ident[EI_MAG0] 0x7f ELFMAG0
e_ident[EI_MAG1] ’E’ ELFMAG1

Assembly Language Programmer’s Guide 11-3

 Chapter 11

Position Value Name

e_ident[EI_MAG2] ’L’ ELFMAG2
e_ident[EI_MAG3] ’F’ ELFMAG3
Chapter 11

e_ident[EI_CLASS]
indicates the file class or capacity and must have the value
ELFCLASS32.
e_ident[EI_DATA]
indicates the byte ordering of processor specific data in the object file
and must be either ELFDATA2LSB (little-endian byte order) or
ELFDATA2MSB (big-endianendian byte order).
e_ident[EI_VERSION]
indicates the version number of the ELF header and must be
EV_CURRENT.
e_ident[PAD]
marks the beginning of unused bytes in the ELF header. These bytes
are reserved and set to zero.
e_type
identifies the type of the object file and can have the following values:
ET_NONE 0 No file type
ET_REL 1 Relocatable
ET_EXEC 2 Executable
ET_DYN 3 Shared object
ET_CORE 4 Core file
ET_LOPROC 0xff00 Processor specific
ET_HIPROC 0xffff Processor specific

e_machine
indicates the required architecture and must have the value
EM_MIPS.
e_version
indicates the object file version and must have the value
EV_CURRENT. The value of EV_CURRENT is 1; in the future, this
value will increase as extensions are added to the ELF header.
e_entry
contains the virtual address to which the system transfers control
when the process is started. If the file has no entry point, this value is
zero.

11-4 Assembly Language Programmer’s Guide

 Execution and Linking Format

e_phoff
contains the offset in bytes of the Program Header Table and may be
zero if the table is not present.

Chapter 11
e_shoff
contains the offset in bytes of the Section Header Table. If the file
has no Section Header Table, its value is zero.
e_flags
contains bit flags associated with the file. The following flags are
defined:
EF_MIPS_NOREORDER 0x00000001
EF_MIPS_PIC 0x00000002
EF_MIPS_CPIC 0x00000004
EF_MIPS_ARCH 0xf0000000
EF_MIPS_ARCH_2 0x10000000
EF_MIPS_ARCH_3 0x20000000
This bit is asserted when at least one .noreorder directive in an
assembly source program contributes to the object module.
If EF_MIPS_PIC is set, the file contains position-independent code
that is relocatable.
If EF_MIPS_CPIC is set, the file contains code that conforms to the
standard calling sequence rules for calling position-independent code.
The code in this file is not necessarily position-independent.
The bits indicated by EF_MIPS_ARCH identify extensions to the
MIPS1 architecture. AN ABI compliant file must have zero in these
four bits.
e_ehsize
contains the size in bytes of the ELF header.
e_phentsize
contains the size in bytes of an entry in the file’s Program Header
Table.
e_phnum
indicates the number of entries in the Program Header Table. If a file
has no Program Header Table, this value is zero. The product of
e_phnum and e_phentsize gives the size in bytes of the Program
Header Table.
e_shentsize
contains the size in bytes of an entry in the Section Header Table
(also referred to as a Section Header).

Assembly Language Programmer’s Guide 11-5

 Chapter 11

e_shnum
indicates the number of entries in the Section Header Table. If a file
has no Section Header Table, this value is zero. The product of
Chapter 11

e_shnum and e_shentsize gives the size in bytes of the Section

Header Table.
e_shstrndx
contains the Section Header Table index of the entry associated with
the Section Name String Table. If the table does not exist, this value
is SHN_UNDEF.

11-6 Assembly Language Programmer’s Guide

 Execution and Linking Format

Sections
Each section has a section header (an entry in the Section Header Table).

Chapter 11
There may be entries in the Section Header Table that have no associated
section. Each section occupies a contiguous, possibly empty, sequence of
bytes and may not overlap any other section.

Section Header Table

The Section Header Table is an array of structures that describe the sections
of the object file. A Section Header Table Index is a subscript into this array
of structures. Some of these indexes are reserved. An object file can not have
a section that corresponds to a reserved index.
The following Special Section Indexes are defined:
NAME VALUE
SHN_UNDEF 0
SHN_LORESERVE 0xff00
SHN_LOPROC 0xff00
SHN_HIPROC 0xff1f
SHN_ABS 0xfff1
SHN_COMMON 0xfff2
SHN_HIRESERVE 0xffff
SHN_MIPS_ACCOMMON 0xff00
SHN_MIPS_SCOMMON 0xff03
SHN_MIPS_SUNDEFINED 0xff04

The special section indexes have the following meanings:

SHN_UNDEF
marks an undefined, missing, or meaningless section reference. A
symbol defined relative to section number SHN_UNDEF is an
undefined symbol.
SHN_LORESERVE
specifies the lower bound of the reserved indexes.
SHN_LOPROC through SHN_HIPROC
are reserved for processor specific semantics.
SHN_ABS
specifies absolute values for the corresponding references. Symbols
defined relative to this section number have absolute values and are
not affected by relocation.

Assembly Language Programmer’s Guide 11-7

 Chapter 11

SHN_COMMON
indicates that the corresponding references are common symbols,
such as FORTRAN COMMON or unallocated C external variables.
Chapter 11

SHN_HIRESERVE
specifies the upper bound of the reserved indexes. The Section
Header Table does not contain entries for the reserved indexes.
SHN_MIPS_ACOMMON
indicates that the corresponding references are common symbols. The
st_value member for a common symbol contains its virtual address. If
the section is relocated, the alignment indicated by the virtual address
is preserved, up to modulo 65536.
SHN_MIPS_SCOMMON
indicates that the corresponding references are common symbols
which can be placed in the global data area (are gp-addressable). This
section only occurs in reloctable object files.
SHN_MIPS_SUNDEFINED
Undefined symbols with this special index in the st_shndx field can
be placed in the global data area (are gp-addressable). This section
only occurs in reloctable object files.

Section Header
A section header (an entry in the Section Header Table) has the following
structure:
Declaration Field
Elf32_Word sh_name;
Elf32_Word sh_type;
Elf32_Word sh_flags;
Elf32_Addr sh_addr;
Elf32_Off sh_offset;
Elf32_Word sh_size;
Elf32_Word sh_link;
Elf32_Word sh_info;
Elf32_Word sh_addralign;
Elf32_Word sh_entsize;

sh_name
specifies the name of the section. Its value is an index into the section
header string table section and gives the location of a null terminated
string that is the name of the section.

11-8 Assembly Language Programmer’s Guide

 Execution and Linking Format

sh_type
indicates the type of the section and may have the following values:
Name Value

Chapter 11
SHT_NULL 0
SHT_PROGBITS 1
SHT_SYMTAB 2
SHT_STRTAB 3
SHT_RELA 4
SHT_HASH 5
SHT_DYNAMIC 6
SHT_NOTE 7
SHT_NOBITS 8
SHT_REL 9
SHT_SHLIB 10
SHT_DYNSYM 11
SHT_LOPROC 0x70000000
SHT_HIPROC 0x7fffffff
SHT_LOUSER 0x80000000
SHT_HIUSER 0xffffffff
SHT_MIPS_LIBLIST 0x70000000
SHT_MIPS_CONFLICT 0x70000002
SHT_MIPS_GPTAB 0x70000003
SHT_MIPS_UCODE 0x70000004

SHT_NULL
marks the section header as inactive. There is no associated section.
Other members of the section header have undefined values.
SHT_PROGBITS
indicates that the section contains information defined by the
program. The format and meaning of the information are determined
by the program.
SHT_SYMTAB and SHT_DYNSYM
sections contain a symbol table. An object file may have only one
section of each type. SHT_SYMTAB contains symbols used in link
editing, but may also be used for dynamic linking. It may contain
many symbols unnecessary for dynamic linking. Consequently, an
object may also contain a SHT_DYNSYM section that contains a
minimal set of dynamic linking symbols.
SHT_STRTAB
indicates that the section holds a string table. An object file may have
multiple string table sections.

Assembly Language Programmer’s Guide 11-9

 Chapter 11

SHT_RELA
indicates that the section contains relocation entries with explicit
addends, such as type Elf32_Rela for the 32-bit class of object files.
Chapter 11

An object file may have multiple relocation sections.

SHT_HASH
marks a section that holds a symbol hash table. An object file may
have only one hash table.
SHT_DYNAMIC
indicates that the section contains information used in dynamic
linking. An object file may have only one dynamic section.
SHT_NOTE
indicates that the section holds information that marks the file in
some way.
SHT_NOBITS
indicates that the section occupies no space but otherwise resembles a
section of type SHT_PROGBITS. Although this section occupies no
space in a file, its sh_offset field contains the conceptual file offset.
SHT_REL
indicates that the section contains relocation entries without explicit
addends. An object file may have multiple relocation sections.
SHT_SHLIB
is a reserved type that has no semantics. A program that contains a
section of this type does not conform to the ABI.
SHT_LOPROC through SHT_HIPROC
are reserved for processor-specific semantics.
SHT_LOUSER
indicates the lower bound of the range of indexes reserved for
application programs.
SHT_HIUSER
indicates the upper bound of the range of indexes reserved for
application programs. Sections types between SHT_LOUSER and
SHT_HIUSER may be used by applications without conflicting with
current or future section types reserved for system use.
SHT_MIPS_LIBLIST
indicates that the section contains information about the set of
dynamic shared object libraries, such as library name and version,
used when statically linking a program. See the Quickstart section in
Chapter 12 of this manual for details.

11-10 Assembly Language Programmer’s Guide

 Execution and Linking Format

SHT_MIPS_CONFLICT
marks a section that contains a list of symbols in an executable object
whose definitions conflict with symbols defined in shared objects.

Chapter 11
SHT_MIPS_GPTAB
indicates that the section contains the global pointer table. The global
pointer table contains a list of possible global data area sizes which
allows the linker to provide the user with information on the optimal
size criteria to use for gp register relative addressing. See the Global
Data Area section of this chapter.
SHT_MIPS_UCODE
indicates that the section contains MIPS ucode instructions.
Other section type values are reserved. The section header for index 0
(SHN_UNDEF) marks undefined section references. This entry has the
following values:
Name Value Note
sh_name 0 No name
sh_type SHT_NULL Inactive
sh_flags 0 No flags
sh_addr 0 No address
sh_offset 0 No file offset
sh_size 0 No size
sh_link SHN_UNDEF No link information
sh_info 0 No auxiliary information
sh_addralign 0 No alignment
sh_entsize 0 No entries

sh_flag
contains bit flags describing attributes of the file. The following flags
are defined:
SHF_WRITE 0x1
SHF_ALLOC 0x2
SHF_EXECINSTR 0x4
SHF_MASKPROC 0xf0000000
SHF_MIPS_GPREL 0x10000000
SHF_WRITE
If this bit is set, the section contains data that must be writable during
process execution.

Assembly Language Programmer’s Guide 11-11

 Chapter 11

SHF_ALLOC
This bit indicates that the section occupies memory during process
execution.
Chapter 11

SHF_EXECINSTR
If this bit is set, the section contains executable machine instructions.
SHF_MASK_PROC
All the bits included in this mask are reserved for processor-specific
semantics.
SHF_MIPS_GPREL
This bit indicates that the section contains data that must be made
part of the global data area during program execution. Data in this
section is addressable with a gp relative address. The sh_link field of
a section with this attribute must be a Section Header Index of a
section of type SHT_MIPS_GPTAB.
sh_addr
If the section appears in the memory image of a process, this member
contains the address of the first byte of the section. Otherwise, its
value is zero.
sh_offset
Contains the byte offset from the beginning of the file of this section.
sh_size
Contains the size of the section in bytes.
sh_link
Contains a Section Header Table index link. The interpretation of this
value depends on the section type (see Table 11-1).
sh_info
Contains miscellaneous information. The interpretation of the value
depends on the section type (see Table 11-1).

Table 11-1: sh_link and sh_info values

sh_type sh_link sh_info

The section header index of the
SHT_DYNAMIC string table used by entries in the 0
section.
The section header index of the
SHT_HASH symbol table to which the hash table 0
applies.
SHT_REL The section header index of the The section header index of the
SHT_DYNSYM associated symbol table. section to which the relocation applies.

11-12 Assembly Language Programmer’s Guide

 Execution and Linking Format

Table 11-1: sh_link and sh_info values

sh_type sh_link sh_info

Chapter 11
One greater than the symbol table
SHT_SYMTAB The section header index of the
index of the last local symbol (bind
SHT_DYNSYM associated string table.
STB_LOCAL).
The section header index of the
SHT_MIPS_LIBLIST string table used by entries in this The number of entries in this section.
section.
The section header index of the
SHT_MIPS_GPTAB not used
SHF_ALLOC + SHF_WRITE section.
other SHN_UNDEF 0

sh_addralign
Indicates address alignment constraints for the section. For example,
if a section contains a doubleword value, the entire section must be
aligned on a doubleword boundary. The value of this member may be
0 or a positive integral power of 2; 0 or 1 indicates that the section
has no alignment constraints.
sh_entsize
If the section holds a table of fixed-size entries, such as a symbol
table, this member gives the size in bytes of each entry. A value of
zero indicates that the section does not contain a table of fixed-size
entries.

Assembly Language Programmer’s Guide 11-13

 Chapter 11

Special Sections
An object file has the following special sections:
Chapter 11

Table 11-2: Special Sections

Name Type Attributes
.bss SHT_NOBITS SHF_ALLOC+SHF_WRITE
.comment SHT_PROGBITS none
.data SHT_PROGBITS SHF_ALLOC+SHF_WRITE
.data1 SHT_PROGBITS SHF_ALLOC+SHF_WRITE
.debug SHT_PROGBITS none
.dynamic SHT_DYNAMIC SHF_ALLOC
.dynstr SHT_STRTAB SHF_ALLOC
.dynsym SHT_DYNSYM SHF_ALLOC
.fini SHT_PROGBITS SHF_ALLOC+ SHF_EXECINSTR
.got SHT_PROGBITS see below
.hash SHT_HASH SHF_ALLOC
.init SHT_PROGBITS SHF_ALLOC+ SHF_EXECINSTR
.interp SHT_PROGBITS see below
.line SHT_PROGBITS none
.note SHT_PROGBITS none
.plt SHT_PROGBITS see below
.relname SHT_REL see below
.relaname SHT_RELA see below
.rodata SHT_PROGBITS SHF_ALLOC
.rodata1 SHT_PROGBITS SHF_ALLOC
.shstrtab SHT_STRTAB none
.strtab SHT_STRTAB see below
.symtab SHT_SYMTAB see below
.text SHT_PROGBITS SHF_ALLOC+ SHF_EXECINSTR
.sdata SHT_PROGBITS SHF_ALLOC+SHF_WRITE+SHF_MIPS_GPREL
.sbss SHT_NOBITS SHF_ALLOC+SHF_WRITE+SHF_MIPS_GPREL
.lit4 SHT_PROGBITS SHF_ALLOC+SHF_MIPS_GPREL
.lit8 SHT_PROGBITS SHF_ALLOC+ SHF_MIPS_GPREL
.reginfo SHT_MIPS_REGINFO SHF_ALLOC
.liblist SHT_MIPS_LIBLIST SHF_ALLOC
.conflict SHT_CONFLICT SHF_ALLOC
.gptab SHT_MIPS_GPTAB none
.got SHT_PROGBITS SHF_ALLOC+ SHF_WRITE+ SHF_MIPS_GPREL
.ucode SHT_MIPS_UCODE none
.mdebug SHT_MIPS_DEBUG none

11-14 Assembly Language Programmer’s Guide

 Execution and Linking Format

NOTE: A MIPS ABI compliant system must support the .sdata, .sbss, .lit4,
.lit8, .reginfo, and .gptab sections. A MIPS ABI compliant system must
recognize, but may choose to ignore, the .liblist, .msym, and .conflict sections.

Chapter 11
However, if any one of these sections is supported, all three must be
supported. A MIPS ABI compliant system is not required to support the
.ucode section, but if this section is present, it must conform to the description
in this manual.
.bss
This section holds uninitialized data. The system initializes the data
to zeros when the program is started. This section occupies no file
space.
.comment
This section holds version information.
.data
This section contains initialized data.
.debug
This section hold information used for symbolic debugging.
.dynamic
This section contains information used for dynamic linking. See
Chapter 12 for more details on dynamic linking.
.dynstr
This section contains strings needed for dynamic linking, usually
strings representing the names associated with symbol table entries.
.fini
This section uholds executable instructions that are executed when
the program terminates normally.
.got
This section hold the Global Offset Table.
.hash
This section contains a hash table for symbols. See the Symbol Table
section of this chapter for a description of the symbol table.
.init
This section holds executable instructions that are executed before
the system calls the main entry point for the program.
.interp
This section holds the path name of a program interpreter. If the file
has a loadable segment that includes the section, the section attributes
include SHF_ALLOC.

Assembly Language Programmer’s Guide 11-15

 Chapter 11

.line
This section contains line number information describing the
correspondence between source code lines and machine code. This
Chapter 11

information is used for symbolic debugging.

.note
This section contains information that marks the file in some way.
See the Note Section section in Chapter 12.
.plt
This section holds the Procedure Linkage Table.
.relname
This section contains relocation information. name is supplied by the
section to which the relocation applies. If the file has a loadable
segment that includes the section, the section attributes include
SHF_ALLOC.
.relaname
This section contains relocation information. name is supplied by the
section to which the relocation applies. If the file has a loadable
segment that includes the section, the section attributes include
SHF_ALLOC.
.rodata
This section holds read-only data that is generally found in the non-
writable segment of the process image. See Chapter 12 for more
information on segments.
.rodata1
This section holds read-only data that is generally found in the non-
writable segment of the process image. See Chapter 12 for more
information on segments.
.shstrtab
This section contains strings representing section names.
.strtab
This section contains strings, including the strings representing the
names associated with symbol table entries. If the file has a loadable
segment that includes the symbol string table, the section attributes
include SHF_ALLOC.
.symtab
This section holds a symbol table. See the Symbol Table section of
this chapter for a description. If the file has a loadable segment that
includes the symbol table, the section attributes include
SHF_ALLOC.

11-16 Assembly Language Programmer’s Guide

 Execution and Linking Format

.text
This section contains executable instructions.
.sdata

Chapter 11
This section holds initialized short data.
.sbss
This section holds uninitialized short data. The system sets the data to
zeros when the program is started. Unlike the .bss section, this
section occupies file space.
.lit4
This section holds 4 byte read-only literals. The section is part of a
non-writable segment in the process image.
.lit8
This section holds 8 byte read-only literals. The section is part of a
non-writable segment in the process image.
.reginfo
This section contains information on the program’s register usage.
.liblist
This section contains information on the libraries used at static link
time.
.conflict
This section provides additional dynamic linking information for
symbols in an executable file that conflict with symbols defined in
the dynamic shared libraries.
.gptab
This section contains a Global Pointer Table. The section is named
.gptab.sbss, .gptab.sdata, .gptab.bss, or .gptab.data depending on the
data section to which the section refers.
.ucode
This section holds U-code instructions generated by the compiler.
.mdebug
This section contains MIPS specific symbol table information. The
contents of this section are described in Chapter 10. The information
in this section is dependent on the location of other sections in the
file. If an object is relocated, this section must be updated. This
section must be discarded if an object file is relocated and the ABI
compliant system chooses not to update the section.
.got
This section contains the Global Offset Table. The sh_info field holds
the Global Pointer value used for this Global Offset Table.

Assembly Language Programmer’s Guide 11-17

 Chapter 11

.dynamic
This section is a MIPS-specific dynamic section. It is the same as the
previously mentioned .dynamic section except that its attributes do
Chapter 11

not include SHF_WRITE.

String Tables
String table sections contain null-terminated character sequences (strings)
that represent symbol and section names. A string is referenced by an index
into the String Table Section.
The first byte of a string section, accessed by index zero, contains a null
character. The last byte also contains a null character, ensuring that all strings
are null terminated. A string whose index is zero specifies either no name or
a null name, depending on the context.
A String Table Section may be empty. In this case, the sh_size field for the
section would contain zero. Non-zero indexes are invalid for an empty string
table.
The following figure shows an example of a string table:

Index 0 1 2 3 4 5 6 7 8 9

\ a b c d \ v a r n

1 a m e \ f o o \ b a

2 r \

A string table index may refer to any byte in the section; references to
substrings are permitted. A single string may be referenced multiple times
and unreferenced strings may exist.

ELF Symbol Table

The ELF symbol table is found in the .symtab section of an object file. It is an
array of structures containing information needed to locate and relocate the
symbol definitions and references of a program.

11-18 Assembly Language Programmer’s Guide

 Execution and Linking Format

A symbol table index is a subscript into this array. Index zero is the first entry
in the table and is also the undefined symbol index. A symbol table entry has
the following format:

Chapter 11
Declaration Name
Elf32_Word st_name;
Elf32_Addr st_value;
Elf32_Word st_size;
unsigned char st_info;
unsigned char st_other;
Elf32_Half st_shndx;

st_name
Holds an index into the symbol string table. If its value is non-zero, it
indicates a string that is the symbol name. Otherwise, the symbol
table entry has no associated name.
st_value
Contains the value of the associated symbol.
st_size
Contains the size (the number of bytes comprising the data object) of
the associated symbol. This value is zero if the symbol has no size or
the size is unknown.
st_info
Specifies the type of the symbol and its binding attributes. The
following code fragment shows how to manipulate the binding and
type:
#define ELF32_ST_BIND(i) ((i)>>4)
#define ELF32_ST_TYPE(i) ((i)&0xf)
#define ELF32_ST_INFO(b,t) ((b)<<4+((t)&0xf))
A symbol’s binding determines the linkage visibility. The value of st_info
may be one of the following:
STB_LOCAL 0
STB_GLOBAL 1
STB_WEAK 2
STB_LOPROC 13
STB_HIPROC 15
STB_LOCAL indicates local symbols. These symbols are not visible
outside of the object file containing the definition. Local symbols
with the same name may exist in multiple object files without causing
conflicts.

Assembly Language Programmer’s Guide 11-19

 Chapter 11

STB_GLOBAL indicates global symbols. Global symbols are visible to

all the object files being combined. A global symbol defined in one
file satisfies a reference to an undefined global symbol in another
Chapter 11

file.
STB_WEAK indicates weak symbols. Weak symbols are similar to global
symbols, but have lower precedence.
STB_LOPROC through STB_HIPROC values are reserved for processor-
specific semantics.
In each symbol table, all local symbols precede the global and weak
symbols. As indicated in the Section Header section of this chapter,
the sh_info field of the section header contains the symbol table
index for the first non-local symbol. Global and weak symbols differ
in two ways:
• When the link editor combines several relocatable object files, it
does not allow multiple definitions of STB_GLOBAL symbols
with the same name. If a defined global symbol exists, the
appearance of a weak symbol with the same name does not cause
an error. The link editor ignores the weak symbol and uses the
global definition.
• When the link editor searches archive libraries, it extracts
members of the archive that contain definitions of undefined
global symbols. The definition in the extracted member may be
either a global or a weak symbol. The link editor does not extract
archive members to resolve undefined weak symbols; unresolved
weak symbols have a value of zero.
st_other
Contains zero and is currently unused.
st_shndx
Contains the Section Header Table index for the symbol table entry.

11-20 Assembly Language Programmer’s Guide

 Execution and Linking Format

Symbol Type
The following symbol types are defined:

Chapter 11
Name Value
STT_NOTYPE 0
STT_OBJECT 1
STT_FUNC 2
STT_SECTION 3
STT_FILE 4
STT_LOPROC 13
STT_HIPROC 15

STT_NOTYPE indicates that the symbol has no type.

STT_OBJECT indicates that the symbol is associated with a data object,
such as a variable.
STT_FUNC indicates that the symbol is associated with a function or
other executable code.
STT_SECTION indicates that the symbol is associated with a section.
Entries of this type are primarily for relocation and normally have
STB_LOCAL binding.
STT_FILE indicates that the symbol name is the name of the source file
associated with the object file. A file symbol has STB_LOCAL
binding, its section index is SHN_ABS, and, if present, it precedes
the other STB_LOCAL symbols.
STT_LOPROC through STT_HIPROC are reserved for processor-specific
semantics.
Function symbols (type STT_FUNC) have special significance. When
another object file references a function that is part of a shared object, the link
editor creates a Procedure Linkage Table entry for the referenced symbol.
Symbols in shared objects that have types other than STT_FUNC are not
referenced automatically through the Procedure Linkage Table.
If the value of a symbol refers to a location within a section, the st_shndx field
for the symbol contains an index into the Section Header table. When the
section is relocated, the symbol’s value is changed and references to the
symbol continue to point to the same location in the program. Some special
section index values have other semantics:
SHN_ABS indicates that the symbol has an absolute value that does not
change because of relocation.

Assembly Language Programmer’s Guide 11-21

 Chapter 11

SHN_COMMON indicates that the symbol is a label for a common block

that has not yet been allocated. The symbol’s value gives alignment
constraints, similar to a section’s sh_addralign field. The link editor
Chapter 11

allocates the storage for the symbol at an address that is a multiple of

st_value. The symbol’s size tells how many bytes are required.
SHN_UNDEF indicates that the symbol is undefined. When the link
editor combines the object file with another that contains the
definition for the symbol, this file’s references to the symbol are
linked to the actual definition.
The symbol table entry for index zero (STN_UNDEF) is reserved and holds
the following information:
Name Value Note
st_name 0 No name
st_value 0 Zero Value
st_size 0 No size
st_info 0 No type, local binding
st_other 0
st_shndx SHN_UNDEF No section

Symbol Values
Symbol table entries for different object file types have slightly different
interpretation for the st_value field:
• In relocatable files, st_value contains the alignment constraints for a
symbol whose section index is SHN_COMMON.
• In relocatable files, st_value contains a section offset for a defined
symbol; st_value is an offset from the beginning of the section that
st_shndx indicates.
• In executable and shared object files, st_value contains a virtual
address. The section offset gives way to a virtual address for which
the section number is irrelevant.
If an executable file contains a reference to a function defined in a shared object,
the symbol table section for the file contains an entry for that symbol. The
st_shndx field of the symbol table entry for the function contains
SHN_UNDEF. If there is a stub for the function in the executable file, and the
st_value field for the symbol table entry is non-zero, the field contains the
virtual address of the first instruction of the function’s stub. Otherwise, the
st_value field contains zero. This stub is used to call the dynamic linker at
runtime for lazy text evaluation.

11-22 Assembly Language Programmer’s Guide

 Execution and Linking Format

Global Data Area

The global data area is part of the data segment of an executable program. It

Chapter 11
contains short data items which can be addressed by the gp register relative
addressing mode. The global data area comprises all the sections with the
SHF_MIPS_GPREL attribute.
The compilers generate short-form (one machine instruction) gp relative
addressing for all data items in any of these sections with the
SHF_MIPS_GPREL attribute. The compilers must generate two machine
instructions to load or store data items outside the global data area. A program
executes faster if more data items are placed in the global data area.
The size of the global data area is limited by the addressing constraints on gp
relative addressing, namely plus or minus 32 kilobytes relative to gp. This
limits the size of the global data area to 64 kilobytes.
The compilers decide whether or not a data item is placed in the global data
area based on its size. All data items less than or equal to a given size are
placed in the global data area. Initialized data items are placed in a .sdata
section, uninitialized data items are placed in a .sbss section, and floating-
point literals are placed in .lit4 and .lit8 sections. The .got section is also
combined into the global data area.
In order to provide the user with information on the optimal size criteria for
placement of data items in the .sdata and .sbss sections, the linker maintains
tables of possible global data area sizes for each of these sections. These
tables are maintained in .gptab sections. Each .gptab section contains both the
actual value used as the size criteria for an object file and a sorted list of
possible short data and bss area sizes based on different data item size
selections. The size criteria value is also known as the –G num.
The .gptab section is an array of structures that have the following format:
typedef union {
struct {
Elf32_Word gt_current_g_value;
Elf32_Word gt_unused;
} gt_header;
struct {
Elf32_Word gt_g_value;
Elf32_Word gt_bytes;
} gt_entry;
} Elf32_gptab;
gt_header.gt_current_g_value
Is the –G num used for this object file. Data items of this size or
smaller are referenced with gp relative addressing and reside in a
SHF_MIPS_GPREL section.

Assembly Language Programmer’s Guide 11-23

 Chapter 11

gt_header.gt_unused
Is not used in the first entry of the array.
gt_entry.gt_g_value
Chapter 11

Is a hypothetical –G num value.

gt_entry.gt_bytes
Is the length of the global data area if the corresponding
gt_entry.gt_g_value were used.
Each of the gt_entry. gt_g_value fields is the size of a data item encountered
during compilation or assembly, including zero. Each separate size criteria
results in an overall size for the global data area. The various entries are sorted
and duplicates are removed. The resulting set of entries, including the –G num
used, yields the .gptab section.
There are always at least two .gptab sections, one that corresponds to
initialized data and one which corresponds to uninitialized data. The sh_info
field of the section specifies the section index of the data section to which this
.gptab section applies. Normally the two .gptab sections would apply to the
.sdata and .sbss sections, but if one or both of these sections do not exist, the
.gptab applies to the .data and .bss sections.
The section to which the .gptab section applies is derived from its name. The
four possible name for this type of section are .gptab.sbss, .gptab.sdata,
.gptab.bss, and .gptab.data.

11-24 Assembly Language Programmer’s Guide

 Execution and Linking Format

Register Information
The compilers and assembler collect information on the registers used by the

Chapter 11
code in the object file. This information is communicated to the operating
system kernel in the .reginfo section. The operating system kernel could use
this information to decide what registers it might not need to save or which
coprocessors the program uses. The section also contains a field which
specifies the initial value for the gp register, based on the final location of the
global data area in memory. The register information structure has the
following format:
typedef struct {
Elf32_Word ri_gprmask;;
Elf32_Word ri_cprmask[4];
Elf32_Word ri_gp_value;
} ELF_RegInfo;
ri_gprmask
contains a bit-mask of general registers used by the program. Each
set bit indicates a general integer register used by the program. Each
clear bit indicates a general integer register not used by the program.
For instance, bit 31 set indicates register $31 is used by the program;
bit 27 clear indicates register $27 is not used by the program.
ri_cprmask
contains the bit-mask of co-processor registers used by the program.
The MIPS RISC architecture can support up to four co-processors,
each with 32 registers. Each array element corresponds to one set of
co-processor registers. Each of the bits within the element
corresponds to individual registers in the co-processor register set.
The 32 bits of the words correspond to the 32 registers, with bit
number 31 corresponding to register 31, bit number 30 to register 30,
etc. Set bits indicate the corresponding register is used by the
program; clear bits indicate the program does not use the
corresponding register.
ri_gp_value
contains the gp register value. In relocatable object files it is used for
relocation of the R_MIPS_GPREL and R_MIPS_LITERAL
relocation types.
NOTE: Only co-processor 1 may be used by ABI compliant programs. This
means that only the ri_cprmask[1] array element may have a non-zero value.
ri_cprmask[0], ri_cprmask[2], and ri_cprmask[3] must all be zero in an ABI
compliant program.

Assembly Language Programmer’s Guide 11-25

 Chapter 11

Relocation
Relocation entries describe how to alter instruction and data fields for
relocation; bit numbers appear in the lower box corners.
Chapter 11

half16
31 15 0

word32
31 0

targ26
31 0

hi16
31 15 0

lo16
31 15 0

rel16
31 15 0

lit16
31 15 0

p
31 15 0

Calculations below assume the actions are transforming a relocatable file into
either an executable or a shared object file. Conceptually, the linker merges
one or more relocatable files to form the output. It first decides how to
combine and locate the input files, then updates the symbol values, and finally
performs the relocation.
Relocations applied to executable or shared object files are similar and
accomplish the same result. The descriptions in Table 11-3 use the following
notation:
A The addend used to compute the value of the relocatable field.
AHL Another type of addend used to compute the value of the relocatable
field. See the note below for more detail.
P The location (section offset or address) of the storage unit being
relocated (computed using r_offset).

11-26 Assembly Language Programmer’s Guide

 Execution and Linking Format

S The value of the symbol whose index resides in the relocation entry,
unless the symbol is STB_LOCAL and is of type STT_SECTION,
in which case it means the original sh_addr minus the final sh_addr.

Chapter 11
G The offset into the global offset table at which the address of the
relocation entry’s symbol resides during execution.
GP The final gp value that is used for the relocatable, executable, or
shared object file being produced.
GPO The gp value used to create the relocatable object.
EA The effective address of the symbol prior to relocation.
L The .lit4 or .lit8 literal table offset. Prior to relocation, the addend
field of a literal reference contains the offset into the global data
area. During relocation, each literal section from each contributing
file is merged into one and sorted, after which duplicate entries are
removed and the section compressed, leaving only unique entries.
The relocation factor L is the mapping from the old offset from the
original gp to the value of gp used in the final file.
A relocation entry’s r_offset value designates the offset or virtual address of
the first byte of the affected storage unit. The relocation type specifies which
bits to change and how to calculate their values. Because MIPS uses only
Elf32_Rel relocation entries, the field to be relocated holds the addend.
The AHL addend is a composite computed from the addends of two
consecutive relocation entries. Each relocation type of R_MIPS_HI16 must
have an associated R_MIPS_LO16 entry immediately following it in the list
of relocations. These relocation entries are always processed as a pair and
both addend fields contribute to the AHL addend. If AHI and ALO are the
addends from the paired R_MIPS_HI16 and R_MIPS_LO16 entries, then the
addend AHL is computed as ((AHI << 16) + (short)ALO). R_MIPS_LO16
entries without an immediately preceding R_MIPS_HI16 entry are orphaned
and the previously defined R_MIPS_HI16 is used for computing the addend.
NOTE: Field names in the following table tell whether the relocation type
checks for overflow. A calculated relocation value may be larger than the
intended field, and a relocation type may verify (V) that the value fits or
truncate(T) the result. As an example, V–half16 means that the computed
value may not have significant non-zero bits outside the half16 field.

Table 11-3: Relocation Calculations

Name Value Field Symbol Calculation

R_MIPS_NONE 0 none local none
R_MIPS_16 1 V–half16 external S + sign-extended(A)

Assembly Language Programmer’s Guide 11-27

 Chapter 11

Table 11-3: Relocation Calculations

Name Value Field Symbol Calculation

Chapter 11

1 V–half16 local S + sign-extended(A)

R_MIPS_32 2 T–word32 external S+A
2 T–word32 local S+A
R_MIPS_REL32 3 T–word32 external A – EA + S
3 T–word32 local A – EA + S
R_MIPS_26 4 T–arg26 local (((A<<2) | (P & 0xf0000000) + S)>>2)
4 T–arg26 external (sign–extended(A<<2)+ S) >> 2
R_MIPS_HI16 5 T–hi16 external ((AHL + S) – (short)(AHL+ S)) >> 16
5 T–hi16 local ((AHL + S) – (short) (AHL+ S)) >> 16
R_MIPS_LO16 6 T–lo16 external AHL + S
6 T–lo16 local AHL + S
R_MIPS_GPREL 7 V–rel16 external sign-extended (A) + S + GP
7 V–rel16 local sign-extended(A) + S + GPO – GP
R_MIPS_LITERAL 8 V–lit16 local signed-extended(A) + L
R_MIPS_GOT16 9 V–rel16 external G
9 V–rel16 local see below
R_MIPS_PC16 10 V–pc16 external sign-extended(A) + S – P
R_MIPS_CALL16 11 V-rel16 external G
R_MIPS_GPREL32 12 /t-word32 local A+S+GP0-GP

In the Symbol column in Table 11-3, if the symbol referenced by the symbol
table index in the relocation entry is STB_LOCAL/STT_SECTION, then it is
a local relocation. If it is not, the relocation is considered an external
relocation.
The R_MIPS_32 and R_MIPS_REL32 relocation types are the only
relocations performed by the dynamic linker.
If an R_MIPS_GOT16 refers to a locally defined symbol, the relocation is
done differently than if it refers to an external symbol. In the local case it must
be followed immediately by an R_MIPS_LO16 relocation. The AHL addend
is extracted and the section in which the referenced data item resides is
determined (this requires all sections in an object module to have unique
addresses and no overlap). From this address the final address of the data item
is calculated. If necessary, a global offset table entry is created to hold the
high 16 bits of this address (an existing entry is used when possible). The
rel16 field is replaced by the offset of this entry in the global offset table. The
lo16 field in the following R_MIPS_LO16 relocation is replaced by the low

11-28 Assembly Language Programmer’s Guide

 Execution and Linking Format

16 bits of the actual destination address. This is meant for local data
references in position-independent code so that only one global offset table
entry is necessary for every 64 kilobytes of local data.

Chapter 11
The first instance of R_MIPS_GOT16 causes the link editor to build a global
offset table if one has not already been built.

R_MIPS_CALL16 relocation entries load function addresses from the global

offset table and indicate the dynamic linker can perform lazy binding.

Assembly Language Programmer’s Guide 11-29

 Chapter 11
Chapter 11

11-30 Assembly Language Programmer’s Guide

Program Loading and Dynamic Linking

12

Chapter 12
Executable files and object files are used to create a process image when a
program is started by the system. This chapter describes the object file
structures that relate to program execution and also describes how the process
image is created from these files. Topics in this chapter include:
• Program Header
• Object File Segments
• Program Loading
• Dynamic Linking
• Quickstart

Assembly Language Programmer’s Guide 12-1

 Chapter 12

Program Header
The Program Header table is an array of structures, each of which describes
a segment or other data used to create a process image. A Program Header is
meaningful only for a shared object or executable file. A description of the
Program Header for MIPS COFF format is in Chapter 9. The structure of a
Program Header for ELF entry is as follows:

Declaration Field
Elf32_Word p_type;
Elf32_Off p_offset;
Elf32_Addr p_vaddr;
Chapter 12

Elf32_Addr p_paddr;
Elf32_Word p_filesz;
Elf32_Word p_memsz;
Elf32_Word p_flags;
Elf32_Word p_align;

The size of the Program Header is specified by the ELF Header e_phentsize
and e_phnum fields (see Chapter 11).
p_type indicates what kind of segment this entry describes or how to interpret
the array element’s information. p_type may have the following values:

Name Value
PT_NULL 0
PT_LOAD 1
PT_DYNAMIC 2
PT_INTERP 3
PT_NOTE 4
PT_SHLIB 5
PT_PHDR 6
PT_LOPROC 0x70000000
PT_HIPROC 0x7fffffff
PT_MIPS_REGINFO 0x70000000

PT_NULL indicates that the Program Header entry is unused; the

values of the other fields of the entry are undefined.
PT_LOAD indicates a loadable segment, described by p_filesz and
p_memsz. The file bytes are mapped into the beginning of the
memory segment. If the memory size is larger than the file size, the
extra bytes contain zeros and follows the segment’s initialized area.

12-2 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

The file size may not be larger than the memory size. Loadable
segments appear in the Program Header table in ascending order
based on the p_vaddr field.
PT_DYNAMIC indicates that the entry contains dynamic linking
information. See the Dynamic Section section of this chapter for
more details.
PT_INTERP indicates that the entry specifies the location and size of
a null-terminated path name to invoke as an interpreter. This type is
meaningful only for executable files (though it may occur for shared
objects) and may not occur more than once in a file. If a segment of
this type is present, it must precede any loadable segment entry.

Chapter 12
PT_NOTE indicates that the entry gives the location and size of
auxiliary information.
PT_SHLIB is reserved and has unspecified semantics. A program
which contains a Program Header entry of this type does not conform
to the ABI.
PT_PHDR indicates that the entry specifies the location and size of
the Program Header table, both in the file and in the memory image
of the program. This type may not occur more than once in a file and
it may only occur if the Program Header table is part of the memory
image of the program. If present, it must precede any loadable
segment entries.
PT_LOPROC through PT_HIPROC values are reserved for
processor-specific semantics.
PT_MIPS_REGINFO indicates that this entry specifies register usage
information. This type may not occur more than once in a file. Its
presence is mandatory and it must precede any loadable segment
entry. See Register Information in Chapter 11.
p_offset gives the offset from the beginning of the file to the first byte of the
segment.
p_vaddr gives the virtual address in memory of the first byte of the segment.
p_paddr is reserved for the segment’s physical address (on systems for which
physical addressing is relevant).
p_filesz contains the number of bytes in the file image of the segment; the
value may be zero.
p_memsz holds the number of bytes in the memory images of the segment;
the value may be zero.

Assembly Language Programmer’s Guide 12-3

 Chapter 12

p_flags contains the flags associated with the segment. The following flags
are defined:

Name Value Meaning

PF_X 0x1 Execute
PF_W 0x2 Write
PF_R 0x4 Read
PF_MASKPROC 0xf0000000 Unspecified

All bits in PF_MASKPROC are reserved for processor-specific semantics.

p_align indicates the alignment of segments in memory and in the file. Values
Chapter 12

0 and 1 mean no alignment is required. Otherwise, p_align should be a

positive, integral power of 2, and p_vaddr should equal p_offset modulo
p_align.

Base Address
Executable file and shared object files have a base address, which is the
lowest virtual address associated with the process image of the program. The
base address is used to relocate the process image of the program during
dynamic linking.
During execution, the base address is calculated from the memory load
address, the maximum page size, and the lowest virtual address of the
program’s loadable segment. The virtual addresses in the Program Header
might not represent actual virtual addresses (see the Program Loading section
of this chapter). The base address is computed by determining the memory
address associated with the lowest p_vaddr for a PT_LOAD segment, and
then truncating this memory address to the nearest multiple of the maximum
page size. The memory address may or may not match the p_addr values.

Segment Permissions
A program that is to be loaded by the system must have at least one loadable
segment, even though this is not required by the file format. When the process
image is created, it has access permissions as specified in the p_flags field.

12-4 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

If a permission bit is zero, that type of access is denied. All flag combinations
are valid but the system may allow more access than requested. However, a
segment does not have write permission unless it is specified explicitly. Table
12-1 shows the exact and allowable interpretations for p_flags.

Table 12-1: p_flags Values and Interpretation

Flags Value Exact Allowable
none 0 All access denied All access denied
PF_X 1 Execute only Read, execute
PF_W 2 Write only Read, write, execute
PF_W+PF_X 3 Write, execute Read, write, execute

Chapter 12
PF_R 4 Read only Read, execute
PF_R+PF_X 5 Read, execute Read, execute
PF_R+PF_W 6 Read, write Read, write, execute
PF_R+PF_W+PF_X 7 Read, write, execute Read, write, execute

Segment Contents
An object file segment may contain one or more sections. The number of
sections in a segment is not important for program loading, but specific
information must be present for linking and execution. The figures below
illustrate typical segment contents for a MIPS executable or shared object. The
order of sections within a segment may vary.
Text segments contain read-only instructions and data, typically including the
following sections:

.reginfo
.dynamic
.liblist
.rel.dyn
.conflict
.dynstr
.dynsym
.hash
.rodata
.text

Assembly Language Programmer’s Guide 12-5

 Chapter 12

Data segments contain writable data and instructions, typically including the
following sections:

.got
.lit4
.lit8

.sdata
.data
.sbss
Chapter 12

.bss

Program Loading
As the system creates or augments a process image, it logically copies a file’s
segment to a virtual memory segment. When, and if, the system physically
reads the file depends on the program’s execution behavior, system load, etc.
A process does not require a physical page unless it references the logical
page during execution, and processes commonly leave many pages
unreferenced. Therefore, delaying physical reads frequently obviates them,
improving system performance. To obtain this efficiency in practice,
executable and shared object files must have segment images whose virtual
addresses are zero, modulo the file system block size.
Virtual addresses for MIPS text segments must be aligned on 4 K (0x1000)
or larger powers of 2 boundaries. MIPS text segments include ELF headers
and program headers. MIPS data segments must be aligned on 64 K
(0x10000) or larger powers of 2 boundaries. File offsets for MIPS segments
must be aligned on 4 K (0x1000) or larger powers of 2 boundaries. Regardless
of the 4 K alignment, segments may not overlap in any given 256 K chunk of
virtual memory; this helps prevent alias problems in systems with virtual
caches. Page size on MIPS systems may vary, but does not exceed 64 K
(0x10000).
Figure 12-1 shows an example of an executable file and Table 12-2 shows the
Program Header entries for the example text and data segments.

12-6 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

File Offset File Virtual Address

0 Text Segment
ELF header
Program header table
Other information
... 0x42bfff
0x2bf00 bytes 0x440000
0x2c000 Data segment
...
0x5000 bytes 0x444fff
0x31000 Other information

Chapter 12
...

Figure 12-1: Example Executable File

Table 12-2: Text and Data Segments

Member Text Data
p_type PT_LOAD PT_LOAD
p_offset 0 0x2c000
p_vaddr 0x400000 0x440000
p_paddr unspecified unspecified
p_filesz 0x2c000 0x5000
p_memsz 0x2c000 0x7000
p_flags PF_R+PF_X PF_R+PF_W+PF_X
p_align 0x10000 0x10000

Because the page size can be larger than the alignment restriction of a
segment’s file offset, up to four file pages hold impure text or data (depending
on page size and file system block size).
• The first text page contains the ELF header, the Program Header
table, and other information.
• The last text page may hold a copy of the beginning of data.
• The first data page may have a copy of the end of text.
• The last data page should be zero or else it will conflict with sbrk
call.
Logically, the system enforces the memory permissions as if each segment
were complete and separate; segments’ addresses are adjusted to ensure that
each logical page in the address space has a single set of permissions. In the

Assembly Language Programmer’s Guide 12-7

 Chapter 12

example above, with 16KB pages, the region of the file holding the end of text
and the beginning of data is mapped twice, at one virtual address for text and
at another virtual address for data.
The end of the data segment requires special handling for uninitialized data,
which must be set to zeros. If a file’s last data page includes information not
in the logical memory page, the extraneous data must be set to zero, not the
unknown contents of the executable file.
One aspect of segment loading differs between executable files and shared
objects. Executable file segments typically contain absolute code. To let the
process execute correctly, these segments must reside at the virtual addresses
used to build the executable file. Thus the system uses the unchanged p_vaddr
Chapter 12

values as virtual addresses.

On the other hand, shared object segments typically contain position-
independent code. This lets a segment’s virtual address change from one
process to another, without invalidating execution behavior. Though the
system chooses virtual addresses for individual processes, it maintains the
segments’ relative positions. Because position-independent code uses relative
addressing between segments, the difference between virtual addresses in
memory must match the difference between virtual addresses in the file. The
following table shows possible shared object virtual address assignments for
several processes, illustrating constant relative positioning. The table also
illustrates base address computations.
Table 12-3: Example Shared Object Segment Addresses
Source Text Data Base Address
File 0x00000200 0x26000 0x00000000
Process1 0x50000200 x50026000 0x50000000
Process2 0x50010200 0x5003b000 0x50010000
Process3 0x60020200 0x6004b000 0x60020000
Process4 0x60030200 0x6005b000 0x60030000

12-8 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

Dynamic Linking

Program Interpreter
An executable file can have only one PT_INTERP Program Header entry.
When the system calls exec(2) to start the process, the path name of the
interpreter is retrieved from the PT_INTERP segment and the initial process
image is created from the interpreter file’s segments. It is then the
interpreter’s responsibility to receive control from the system and create the
application program’s environment.
The interpreter receives control in one of two ways. First, it may receive the

Chapter 12
file descriptor of the executable file, positioned at the beginning of the file.
The file descriptor can then be used to read or map the executable file’s
segments into memory. Second, depending on the executable file format, the
system may load the executable file into memory before giving control to the
interpreter. With the possible exception of the file descriptor, the interpreter’s
initial process state is the same as what the executable file would have
received. The interpreter cannot require a second interpreter and may be
either a shared object or an executable file.
A shared object is loaded as position-independent with addresses that may
vary from one process to another; the system creates the segments in the
dynamic segment area used by mmap(2) and related services. As a result, a
shared object interpreter typically does not conflict with the executable file’s
original segment addresses.
An executable file is loaded at fixed addresses; the system creates its
segments using the virtual addresses from the Program Header table.
Consequently, an executable file interpreter’s virtual addresses may conflict
with those of the executable file. The interpreter is responsible for resolving
any conflicts.

Dynamic Linker
When building an executable file that uses dynamic linking, the link editor
adds a Program Header entry of type PT_INTERP to the executable file. This
entry tells the system to invoke the dynamic linker as the program interpreter.
Typically, the dynamic linker requested is libsys, the system library. exec(2)
and the dynamic linker cooperate to create the process image, which involves
the following:
• Adding the file segments to the process image.
• Adding shared object segments to the process image.
• Performing relocations for the executable file and its shared objects.

Assembly Language Programmer’s Guide 12-9

 Chapter 12

• Closing the file descriptor for the executable file, if a file descriptor
was passed to the dynamic linker.
• Transferring control to the program, making it appear that the
program received control directly from exec(2).
The link editor also constructs various data for shared objects and executable
file that assist the dynamic linker. These data are located in loadable
segments, are available during execution, and consist of the following:
• A dynamic section of type SHT_DYNAMIC holds various data,
including a structure that resides at the beginning of the section and
hold the addresses of other dynamic linking information.
• The .hash section of type SHT_HASH contains a symbol hash table.
Chapter 12

• The .got and .plt sections, of type SHT_PROGBITS, contain the

Global Offset Table and the Procedure Linkage Table, respectively.
Shared objects may be located at virtual addresses that are different from the
addresses in the Program Header table. The dynamic linker relocates the
memory image and updates absolute addresses before control is given to the
program.
If the environment variable LD_BIND_NOW has a non-null value, the
dynamic linker processes all relocations before transferring control to the
program. The dynamic linker may evaluate procedure linkage table entries
lazily, avoiding symbol resolution and relocation for functions that are not
called.
The dynamic linker performs linking of objects at run-time and is invoked
either through the operating system kernel or by start-up code in the
executable. In either case, the initial entry point for the dynamic linker is in
entry zero of the Global Offset Table. Each entry should be considered a
subroutine:
void entry+0();
Normal entry point for the dynamic linker when invoked by the
operating system kernel. This entry takes no arguments and returns
no values.
void entry+8(Elf32_Addr base, char **envp);
This entry point is 8 bytes beyond the entry point given by location
zero in the GOT. This entry is called when the dynamic linker is
invoked by start-up code in the executable. The argument base should
be the value of the extern _BASE_ADDRESS. It is a pointer to the
first location in the text segment. The envp argument is a pointer to
the environment. This entry point returns a pointer to the dynamic
linker’s object list.

12-10 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

Dynamic Section
An object file that is used in dynamic linking has an entry in its Program
Header Table of type PT_DYNAMIC. This segment contains the .dynamic
section, which is labeled _DYNAMIC and is an array with entries of the
following type:
typedef struct {
Elf32_Sword d_tag;
union {
Elf32_Word d_val;
Elf32_Addr d_ptr;
} d_un;

Chapter 12
} Elf32_Dyn;
dtag indicates how the d_un field is to be interpreted.
d_val represents integer values.
d_ptr represents program virtual address. A file’s virtual addresses may not
match the memory virtual addresses during execution. The dynamic linker
computes actual addresses based on the virtual address from the file and the
memory base address. Object files do not contain relocation entries to correct
addresses in the dynamic structure.
The tag (d_tag) requirements for executable and shared object files are
summarized in the following table. If the executable entry indicates
mandatory, the dynamic linking array must contain an entry of that type.
Optional indicates that an entry for the tag may exist but is not required.

Assembly Language Programmer’s Guide 12-11

 Chapter 12

Name Value d_un Executable Shared Object

DT_NULL 0 ignored mandatory mandatory
DT_NEEDED 1 d_val optional optional
DT_PLTRELSZ 2 d_val optional optional
DT_PLTGOT 3 d_ptr optional optional
DT_HASH 4 d_ptr mandatory mandatory
DT_STRTAB 5 d_ptr mandatory mandatory
DT_SYMTAB 5 d_ptr mandatory mandatory
DT_RELA 7 d_ptr mandatory optional
Chapter 12

DT_RELASZ 8 d_val mandatory optional

DT_RELAENT 9 d_val mandatory optional
DT_STRSZ 10 d_val mandatory mandatory
DT_SYMENT 11 d_val mandatory mandatory
DT_INIT 12 d_ptr optional optional
DT_FINI 13 d_ptr optional optional
DT_SONAME 14 d_val ignored optional
DT_RPATH 15 d_val optional ignored
DT_SYMBOLIC 16 ignored ignored optional
DT_REL 17 d_ptr mandatory optional
DT_RELSZ 18 d_val mandatory optional
DT_RELENT 19 d_val mandatory optional
DT_PLTREL 20 d_val optional optional
DT_DEBUG 21 d_ptr optional ignored
DT_TEXTREL 22 ignored optional optional
DT_JMPREL 23 d_ptr optional optional
DT_LOPROC 0x70000000 unspecified unspecified unspecified
DT_HIPROC 0x7fffffff unspecified unspecified unspecified

DT_NULL
An entry of this type marks the end of the _DYNAMIC array.
DT_NEEDED
This element contains the string table offset of a null terminated
string that is the name of a library. The offset is an index into the
table indicated in the DT_STRTAB entry. The dynamic array may
contain multiple entries of this type. The order of this entries is
significant.

12-12 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

DT_PLTRELSZ
This element contains the total size in bytes of the relocation entries
associated with the Procedure Linkage Table. If an entry of type
DT_JMPREL is present, it must have an associated DT_PLTRELSZ
entry.
DT_PLTGOT
Procedure Linkage Table and/or the Global Offset Table.
DT_HASH
This element contains the address of the symbol hash table.
DT_STRTAB
This entry contains the address of the string table.

Chapter 12
DT_SYMTAB
This entry contains the address of the symbol table with Elf32_Sym
entries for the 32-bit class of files.
DT_RELA
This element contains the address of a relocation table. Entries in the
table have explicit addends, such as Elf32_Rela. An object file may
have multiple relocation sections. When the link editor builds the
relocation table for an executable or shared object, these sections are
concatenated to form a single table. While the sections are
independent in the object file, the dynamic linker sees a single table.
When the dynamic linker creates a process image or adds a shared
object to a process image, it reads the relocation table and performs
the associated actions. If this element is present, the dynamic
structure must also contains DT_RELASZ and DT_RELAENT
entries. When relocation is mandatory for a file, either DT_RELA or
DT_REL may be present.
DT_RELASZ
This entry contains the size in bytes of the DT_RELA relocation
table.
DT_RELAENT
This entry contains the size in bytes of the DT_RELA relocation
entry.
DT_STRSZ
This element contains the size in bytes of the string table.
DT_SYMENT
This entry contains the size in bytes of a symbol table entry.
DT_INIT
This element contains the address of the initialization function.

Assembly Language Programmer’s Guide 12-13

 Chapter 12

DT_FINI
This element contains the address of the termination function.
DT_SONAME
This entry contains the string table offset of a null-terminated string
that gives the name of the shared object. The offset is an index into
the table indicated in the DT_STRTAB entry.
DT_RPATH
This element contains the string table offset of a null-terminated
search library search path string. The offset is an index into the table
indicated in the DT_STRTAB entry.
DT_SYMBOLIC
Chapter 12

If this element is present, the dynamic linker uses a different symbol

resolution algorithm for references within a library. The symbol
search starts from the shared object instead of the executable file. If
the shared object does not supply the referenced symbol, the
executable file and other shared objects are searched.
DT_REL
This entry is similar to DT_RELA, except that its table has implicit
addends. If this element is present, the dynamic structure must also
contain DT_RELASZ and DT_RELENT elements.
DT_RELSZ
This entry contains the size in bytes of the DT_REL relocation table.
DT_RELENT
This entry contains the size in bytes of the DT_REL relocation entry.
DT_PLTREL
This element specifies the type of relocation entry referred to by the
Procedure Linkage Table. The d_val member holds DT_REL or
DT_RELA, as appropriate. All relocations in a Procedure Linkage
Table must use the same relocation.
DT_DEBUG
This entry is used for debugging. Its contents are not specified.
Programs that access this entry do not conform to the ABI.
DT_TEXTREL
If this element is not present, then no relocation entry should cause a
modification to a non-writable segment. If this element is present,
one or more relocation entries might request modifications to a non-
writable segment.

12-14 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

DT_JMPREL
If this element is present, its d_ptr field contains the address of
relocation entries associated only with the Procedure Linkage Table.
The dynamic linker may ignore these entries during process
initialization if lazy binding is enabled.
DT_LOPROC through DT_HIPROC
The values in this range are reserved for processor-specific
semantics.
Table 12-4 lists MIPS-specific tags

Table 12-4: Dynamic Arrays Tags d_tag

Chapter 12
Name Value d_un Executable Shared Object
DT_MIPS_RLD_VERSION 0x70000001 d_val mandatory mandatory
DT_MIPS_TIME_STAMP 0x70000002 d_val optional optional
DT_MIPS_ICHECKSUM 0x70000003 d_val optional optional
DT_MIPS_IVERSION 0x70000004 d_val optional optional
DT_MIPS_FLAGS 0x70000005 d_val mandatory mandatory
DT_MIPS_BASE_ADDRESS 0x70000006 d_ptr mandatory mandatory
DT_MIPS_CONFLICT 0x70000008 d_ptr optional optional
DT_MIPS_LIBLIST 0x70000009 d_ptr optional optional
DT_MIPS_LOCAL_GOTNO 0x7000000a d_val mandatory mandatory
DT_MIPS_CONFLICTNO 0x7000000b d_val optional optional
DT_MIPS_LIBLISTNO 0x70000010 d_val optional optional
DT_MIPS_SYMTABNO 0x70000011 d_val optional optional
DT_MIPS_UNREFEXTNO 0x70000012 d_val optional optional
DT_MIPS_GOTSYM 0x70000013 d_val mandatory mandatory
DT_MIPS_HIPAGENO 0x70000014 d_val mandatory mandatory
DT_MIPS_RLD_MAP 0x70000016 d_val optional optional

DT_MIPS_RLD_VERSION
This element holds an index into the object file’s string table, which
holds the version of the Runtime Linker Interface. The version is
currently 1.
DT_MIPS_TIME_STAMP
This entry contains a 32-bit time stamp.
DT_MIPS_CHECKSUM
This elements’s value is the sum of all external strings and common
sizes.

Assembly Language Programmer’s Guide 12-15

 Chapter 12

DT_MIPS_IVERSION
This element holds an index into the object file’s string table. The
version string is a series of colon (:) separated version strings. An
index value of zero means no version string was specified.
DT_MIPS_FLAGS
This entry contains a set of 1-bit flags. Flag definitions appear below.
DT_MIPS_BASE_ADDRESS This element contains the base address as
defined in the generic ABI.
DT_MIPS_CONFLICT
This entry contains the address of the .conflict section.
DT_MIPS_LIBLIST
Chapter 12

This element contains the address of the .liblist section.

DT_MIPS_LOCAL_GOTNO
This element contains the number of local GOT entries.
DT_MIPS_CONFLICTNO
This entry contains the number of entries in the .conflict section and
is mandatory if there is a .conflict section.
DT_MIPS_RLD_MAP
This entry contains the address of the word that contains a pointer to
the dynamic linker’s object list.
DT_MIPS_SYMTABNO
This entry indicates the number of entries in the .dynsym section.
DT_MIPS_LIBLISTNO
This element indicates the number of entries in the .liblist section.
DT_MIPS_UNREFEXTNO
This field holds the index into the dynamic symbol table which is the
entry of the first external symbol that is not referenced within the
same object.
DT_MIPS_GOTSYM
This field holds the index of the first dynamic symbol table entry that
corresponds to an entry in the global offset table.
DT_MIPS_HIPAGENO
This field holds the number of page table entries in the global offset
table. A page table entry here refers to a 64K byte chunk of data
space. This field is used by profiling tools and is optional.
Entries may appear in any order, except for the relative order of the
DT_NEEDED elements and the DT_NULL element at the end of the
array. All other tag values are reserved.
The following flags are defined for DT_MIPS_FLAGS:

12-16 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

RHF_NONE 0x00000000 none

RHF_HARDWAY 0x00000001 ignore shortcut pointers
RHF_NOTPOT 0x00000002 hash size not a power of two

Chapter 12

Assembly Language Programmer’s Guide 12-17

 Chapter 12

Shared Object Dependencies

When the link editor processes an archive library, library members are
extracted and copied into the output object file. These statically linked
services are available during execution and do not involve the dynamic linker.
Shared objects also provide services which require the dynamic linker to
include the appropriate shared object files in the process image. To
accomplish this, executable and shared object files must describe their
dependencies.
The dependencies, indicated in the DT_NEEDED entries of the dynamic
structure, tell what shared objects are required for the program. The dynamic
linker builds a process image by connecting the referenced shared objects and
Chapter 12

their dependencies. When resolving symbolic references, the dynamic linker

looks first at the symbol table of the executable program, then at the symbol
tables of the DT_NEEDED entries (in order), then at the second level
DT_NEEDED entries, and so on. Shared object files must be readable by the
process.
NOTE: Even if a shared object is referenced more than once in the
dependency list, the dynamic linker only includes one instance of the object
in the process image.
Names in the dependency list are copies of either the DT_SONAME strings
or the path names of the shared objects used to build the object file. If the link
editor builds an executable file from a shared object with a DT_SONAME
entry of liba and another shared object with path name /usr/lib/libz, the
executable file contains liba and /usr/lib/libz in its dependency list.
If a shared object name has one or more slash characters in its name, such as
/usr/lib/libz, the dynamic linker uses the string as the path name. If the name
has no slashes, such as liba, the object is searched for as follows:
• First, the dynamic array tag DT_RPATH may give a string that holds
a list of directories separated by colons, such as /usr/new/lib: /usr/
local/lib. The dynamic linker searches these directories in order and
if a library is not located, also looks in the current directory.
• Second, the environment variable LD_LIBRARY_PATH may hold a
list of colon separated directories, optionally followed by a semicolon
and another directory list. These directories are searched after those
specified by DT_RPATH.
• Finally, if the library was not located in any of the directories
specified by DT_PATH or LD_LIBRARY_PATH, the dynamic
linker searches /lib, then /usr/lib, and then /usr/lib/cmplrs/cc.
MIPS defines the following environment variables:

12-18 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

_RLD_PATH path to dynamic linker (rld)

_RLD_ARGS argument list to dynamic linker
_RLD_ROOT prefix dynamic linker prepends to all paths
NOTE: For security, the dynamic linker ignores environmental search
specifications, such as LD_LIBRARY_PATH, for set-user-ID and set-group-
ID programs.

Global Offset Table (GOT)

Position-independent code cannot, in general, contain absolute virtual
addresses. Global Offset Tables (GOT) hold absolute addresses in private
data, thus making the addresses available without compromising the position-

Chapter 12
independence and shareability of a program’s text. A program references its
Global Offset Table using position-independent addressing and extracts
absolute values, thus redirecting position-independent references to absolute
locations.
The Global Offset Table is split into two logically separate subtables: locals
and externals. Local entries reside in the first part of the table; these are
entries for which there are standard local relocation entries. These entries
only require relocation if they occur in a shared object and the shared object’s
memory load address differs from the virtual address of the shared object’s
loadable segments. As with the defined external entries in the Global Offset
Table, these local entries contain actual addresses.
External entries reside in the second part of the section. Each entry in the
external part of the GOT corresponds to an entry in the .dynsym section. The
first symbol in the .dynsym section corresponds to the first word of the table,
the second symbol corresponds to the second word, and so on. Each word in
the external entry part of the GOT contains the actual address for its
corresponding symbol. The external entries for defined symbols must contain
actual addresses. If an entry corresponds to an undefined symbol and the table
entry contains a zero, the entry must be resolved by the dynamic linker, even
if the dynamic linker is performing a quickstart. See the Quickstart section of
this chapter for more information.
After the system creates memory segments for a loadable object file, the
dynamic linker may process the relocation entries. The only relocation entries
remaining are type R_MIPS_REL32, referring to local entries in the GOT and
data containing addresses. The dynamic linker determines the associated
symbol (or section) values, calculates their absolute addresses, and sets the
proper values. Although the absolute addresses may be unknown when the
link editor builds an object file, the dynamic linker knows the addresses of all
memory segments and can find the correct symbols and calculate the absolute
addresses.

Assembly Language Programmer’s Guide 12-19

 Chapter 12

If a program requires direct access to the absolute address of a symbol, it uses

the appropriate GOT entry. Because the executable file and shared objects
have separate Global Offset Tables, a symbol’s address may appear in several
tables. The dynamic linker processes all necessary relocations before giving
control to the process image, thus ensuring the absolute addresses are
available during execution.
The zero entry of the .dynsym section is reserved and holds a null symbol
table entry. The corresponding zero entry in the GOT is reserved to hold the
address of the entry point in the dynamic linker to call when lazy resolving
text symbols; see the Procedure Linkage Table section in this chapter. When
a program begins execution, it must check this entry and if it is zero, the
program must invoke the dynamic linker; otherwise, the system has done so
Chapter 12

for the program as part of program loading.

If the system has not invoked the dynamic linker and the program fails to map
in a dynamic linker, or the program fails to find a dynamic linker, then the
program must execute a BREAK instruction with a code of 10. This allows
the parent program to determine the reason for failure.
The system may choose different memory segment addresses for the same
shared object in different programs; it may even choose different library
addresses for different executions of the same program. Nonetheless,
memory segments do not change addresses once the process image is
established. As long as a process exists, its memory segments reside at fixed
virtual addresses.

Calling Position Independent Functions

The GOT is used to hold addresses of position independent functions as well
as data addresses. It is not possible to resolve function calls from one
executable or shared object to another at static link time, so all of the function
address entries in the GOT would normally be resolved at execution time. The
dynamic linker would then resolve all of these undefined relocation entries at
run-time. Through the use of specially constructed pieces of code known as
stubs this run-time resolution can be deferred through a technique known as
lazy binding.
Using this technique, the link editor (or a combination of the compiler,
assembler, and link editor) builds a stub for each called function, and
allocates a GOT entry that initially points to the stub. Because of the normal
calling sequence for position independent code, the call ends up invoking the
stub the first time the call is made.
stub_xyz: .
lw t9, 0(gp)
move t7, ra

12-20 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

jal t9
li t8, .dynsym_index
The stub code loads register t9 with an entry from the GOT which contains a
well-known entry point in the dynamic linker; it also loads register t8 with the
index into the .dynsym section of the referenced external. The code saves
register ra and transfers control to the dynamic linker. The dynamic linker
determines the correct address for the called function and replaces the address
of the stub in the GOT with the address of the function.
Most undefined text references can be handled by lazy text evaluation except
when the address of a function is used other than in a jump and link
instruction. In this case, rather than the actual address of the function you

Chapter 12
would get the address of the stub.
The dynamic linker detects this usage in the following manner:
The link editor generates symbol table entries for all function references with
the st_shndx field containing SHN_UNDEF and the st_type field containing
STT_FUNC. The dynamic linker examines each symbol table entry when it
starts execution. If the st_value field for one of these symbols is non-zero then
there were only jump and link references to the function and nothing need be
done to the GOT entry; if the field is zero, then there was some other kind of
reference to the function and the GOT entry must be replaced with the actual
address of the referenced function.
The LD_BIND_NOW environment variable can also change dynamic linking
behavior. If its value is non-null, the dynamic linker evaluates all symbol
table entries of type STT_FUNC, replacing their stub addresses in the GOT
with the actual address of the referenced function.
NOTE: Lazy binding generally improves overall application performance,
because unused symbols do not incur the dynamic linking overhead.
Nevertheless, two situations make lazy binding undesirable for some
applications. First, the initial reference to a shared object function takes
longer than subsequent calls, because the dynamic linker intercepts the call to
resolve the symbol. Some applications cannot tolerate this unpredictability.
Second, if an error occurs and the dynamic linker cannot resolve the symbol,
the dynamic linker terminates the program. Under lazy binding, this might
occur at arbitrary times. Once again, some applications cannot tolerate this
unpredictability. By turning off lazy binding, the dynamic linker forces the
failure to occur during process initialization, before the application receives
control.

Assembly Language Programmer’s Guide 12-21

 Chapter 12

Symbols
All externally visible symbols, both defined and undefined, must be hashed
into the hash table.
Undefined symbols of type STT_FUNC which have been referenced only by
jump and link instructions may contain non-zero values in the their st_value
field denoting the stub address used for lazy evaluation for this symbol. The
run-time linker uses this to reset the GOT entry for this external to its stub
address when unlinking a shared object. All other undefined symbols must
contain zero in their st_value fields.
Defined symbols in an executable may not be preempted. The symbol table
in the executable is always searched first to resolve any symbol references.
Chapter 12

The executable may or may not contain position independent code.

Relocations
There may be only one dynamic relocation section to resolve addresses in
data and local entries in the GOT. It must be called .rel.dyn. Executables
may contain normal relocation sections in addition to a dynamic relocation
section. The normal relocation sections may contain resolutions for any
absolute values in the main program. The dynamic linker does not resolve
these or relocate the main program.
As noted previously, only R_MIPS_REL32 relocation entries are supported
in the dynamic relocation section.

Hash table
A hash table of Elf32_Word entries supplies symbol table access. The hash
table can be viewed as follows:

nbucket
nchain
bucket[0]
...
bucket[nbucket – 1]
chain[0]
...
chain[nchain – 1]

12-22 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

nbucket indicates the number of entries in the bucket array and nchain
indicates the number of entries in the chain array. Both bucket and chain hold
symbol table indexes; the entries in chain parallel the symbol table. The
number of symbol table entries should be equal to nchain; symbol tables
indexes also select chain entries.
A hashing function accepts a symbol name and returns a value that may be
used to compute a bucket index. If the hashing function returns the value X
for a name, bucket[X % nbucket] gives an index, Y, into the symbol table and
chain array. If the symbol table entry indicated is not the correct one, chain[Y]
indicates the next symbol table entry with the same hash value. The chain
links can be followed until either the desired symbol table entry is located, or

Chapter 12
the chain entry contains the value STN_UNDEF.

Initialization and Termination Functions

After the dynamic linker has created the process image and performed
relocations, each shared object gets the opportunity to execute initialization
code. The initialization functions are called in no particular order, but all
shared object initialization occurs before the executable file gains control.
Similarly, shared object may have termination functions that are executed by
the atexit(3) mechanism when the process is being terminated. The order in
which the dynamic linker calls the termination functions is unspecified.
Shared objects designate initialization and termination functions through
DT_INIT and DT_FINI entries in the dynamic structure. Typically, the code
for these functions resides in the .init and .fini sections.
NOTE: Although the atexit(3) termination processing normally is done, it is
not guaranteed to have executed upon process death. In particular, the process
does not execute the termination processing if it calls _exit(2) or if the process
dies because it received a signal that it neither caught nor ignored.

Quickstart
MIPS supports several sections which are useful for faster startup of
programs that have been linked with shared objects. Some ordering
constraints are imposed on these sections. The group of structures defined in
these sections and the ordering constraints allow the dynamic linker to
operate more efficiently. These additional sections are also used for more
complete dynamic shared object version control.
NOTE: An ABI compliant system may ignore any of the three sections
defined here, but if it supports one of these sections, it must support all three.

Assembly Language Programmer’s Guide 12-23

 Chapter 12

If an object file is relinked or relocated on secondary storage and these

sections cannot be processed, they must be deleted.

Shared Object List

A shared object list section is an array of Elf32_Lib structures which contains
information about the various dynamic shared objects used to statically link
the object file. Each shared object used has an entry in the array; each entry
has the following format:
typedef struct {
Elf32_Word l_name;
Elf32_Word l_time_stamp;
Chapter 12

Elf32_Word l_checksum;
Elf32_Word l_version;
Elf32_Word l_flags;
} Elf32_Lib;
l_name
This member specifies the name of a shared object. Its value is a
string table index. This name may be a trailing component of the path
to be used with RPATH + LD_LIBPATH, a name containing ‘/’s
which is relative to ‘.’, or it may be a full path name.
l_time_stamp
This member’s value is a 32-bit time stamp. The value can be
combined with the l_checksum value and the l_version string to form
a unique id for this shared object.
l_checksum
This member’s value is the sum of all externally visible symbols’
string names and common sizes.
l_version
This member specifies the interface version. Its value is a string table
index. The interface version is a single string containing no colons. It
is compared to a colon separated string of versions pointed to by a
dynamic section entry of the shared object. Shared objects with
matching names may be considered incompatible if the interface
version strings are deemed incompatible. An index value of zero
means no version string is specified.
l_flags
This member is a set of 1-bit flags. The following flags are defined:
LL_EXACT_MATCH0x00000001require exact match
LL_IGNORE_INT_VER0x00000002ignore interface version
LL_EXACT_MATCH

12-24 Assembly Language Programmer’s Guide

 Program Loading and Dynamic Linking

At run-time use a unique id composed of the l_time_stamp,

l_checksum, and l_version fields to demand that the run-time
dynamic shared object match exactly the shared object used at
static link time.
LL_IGNORE_INT_VER
At run-time, ignore any version incompatibility between the
dynamic shared object and the object used at static link time.
Normally, if neither LL_EXACT_MATCH nor LL_IGNORE_INT_VER
bits are set, the dynamic linker requires that the version of the dynamic shared
library match at least one of the colon separated version strings indexed by
the l_version string table index.

Chapter 12
Conflict Section
Each .conflict section is an array of indexes into the .dynsym section. Each
index identifies a symbol whose attributes conflict with a shared object on
which it depends, either in type or size, such that this definition preempts the
shared object’s definition. The dependent shared object is identified at static
link time. The .conflict section is an array of Elf32_Conflict elements.
typedef Elf32_Addr Elf32_Conflict;

Ordering
In order to take advantage of Quickstart functionality, ordering constraints
are imposed on the .dynsym and .rel.dyn sections. The .dynsym section must
be ordered on increasing values of the st_value field. Note that this requires
the .got section to be ordered in the same way, since it must correspond to the
.dynsym section.
The .rel.dyn section must have all local entries first, followed by the external
entries. Within these sub-sections, the entries must be ordered by symbol
index. This groups each symbol’s relocations together.

Assembly Language Programmer’s Guide 12-25

 Chapter 12
Chapter 12

12-26 Assembly Language Programmer’s Guide

Instruction Summary

Appendix A
A
The tables in this appendix summarize the assembly language instruction set.
Most of the assembly language instructions have direct machine equivalents.
Refer to Appendix A and Appendix B of the MIPS RISC Architecture book
published by Prentice-Hall for detailed instruction descriptions. In the tables
in this appendix, the operand terms have the following meanings:

Operand Description
destination Destination register
address Symbolic expression (see Chapter2)
source Source register
expression An absolute value
immediate Immediate value
label Symbol label
breakcode Value that determines the break

Assembly Language Programmer’s Guide A-1

 Appendix A

Table A-1: Main Processor Instruction Summary

Description Op-code Operands
Appendix A

Load Address la destination, address

Load Byte lb
Load Byte Unsigned lbu
Load Halfword lh
Load Halfword Unsigned lhu
Load Linked ll
Load Word lw
Load Word Left lwl
Load Word Right lwr
Load Word Coprocessor z lwcz
Load Double Coprocessor z ldcz
Load Double ld
Load Linked Doubleword lld
Load Word Unsigned lwu
Load Doubleword Left ldl
Load Doubleword Right ldr
Unaligned Load Double uld
Unaligned Load Halfword ulh
Unaligned Load Halfword Unsigned ulhu
Unaligned Load Word ulw
Load Immediate li destination, expression
Load Upper Immediate lui
Restore From Exception rfe
Syscall syscall
Store Byte sb source, address
Store Conditional sc
Store Double sd
Store Conditional Doubleword scd
Store Double Left sdl
Store Double Right sdr
Store Halfword sh
Store Word Left swl
Store Word Right swr
Store Word sw
Unaligned Store Halfword ush
Unaligned Store Doubleword usd
Unaligned Store Word usw
Conditional Trap
Trap if Equal teq src1, src2
Trap if not Equal tne src1, immediate
Trap if Less Than tlt

A-2 Assembly Language Programmer’s Guide

 Table A-1: Main Processor Instruction Summary
Description Op-code Operands

Appendix A
Trap if Less than, Unsigned tltu
Trap if Greater Than or Equal tge
Trap if Greater than or Equal, Unsigned tgeu
Absolute Value abs destination,src1
Negate with Overflow neg destination/src1
Negate without Overflow negu
NOT not
Doubleword Absolute Value dabs
Doubleword Negate with Overflow dneg
Doubleword Negate without Overflow dnegu
Add with Overflow add destination, src1, src2
Add without Overflow addu destination, src1, src2
AND and destination, src1, immediate
Divide Signed div destination/src1, immediate
Divide Unsigned divu
Exclusive-OR xor
Multiply mul
Multiply with Overflow mulo
Multiply with Overflow Unsigned mulou
NOT OR nor
OR or
Set Equal seq
Set Greater sgt
Set Greater/Equal sge
Set Greater/Equal Unsigned sgeu
Set Greater Unsigned sgtu
Set Less slt
Set Less/Equal sle
Set Less/Equal Unsigned sleu
Set Less Unsigned sltu
Set Not Equal sne
Subtract with Overflow sub
Subtract without Overflow subu
Remainder Signed rem
Remainder Unsigned remu
Rotate Left rol
Rotate Right ror
Shift Right Arithmetic sra
Shift Left Logical sll
Shift Right Logical srl
Multiply mult src1,src2
Multiply Unsigned multu

Assembly Language Programmer’s Guide A-3

 Appendix A

Table A-1: Main Processor Instruction Summary

Description Op-code Operands
Doubleword Multiply dmult src1, src2
Appendix A

Doubleword Multiply Unsigned dmultu src1, immediate

Doubleword Add with Overflow dadd destination, src1, src2
Doubleword Add without Overflow daddu destination, src1, src2
Doubleword Divide Signed ddiv destination, src1, immediate
Doubleword Divide Unsigned ddivu destination/src1, immediate
Doubleword Multiply dmul
Doubleword Multiply with Overflow dmulo
Doubleword Multiply with Overflow Unsigned dmulou
Doubleword Subtract with Overflow dsub
Doubleword Subtract without Overflow dsubu
Doubleword Remainder Signed drem
Doubleword Remainder Unsigned dremu
Doubleword Rotate Left drol
Doubleword Rotate Right dror
Doubleword Shift Right Arithmetic dsra
Doubleword Shift Left Logical dsll
Doubleword Shift Right Logical dsrl
Jump j address
Jump and Link jal address, target
return,target
Branch on Equal beq src1,src2,label
Branch on Greater bgt src1, immediate,label
Branch on Greater/Equal bge
Branch on Greater/Equal Unsigned bgeu
Branch on Greater Unsigned bgtu
Branch on Less blt
Branch on Less/Equal ble
Branch on Less/Equal Unsigned bleu
Branch on Less Unsigned bltu
Branch on Not Equal bne
Branch b label
Branch and Link bal
Branch Coprocessor z False bczf
Branch Coprocessor z True bczt
Branch Coprocessor z FalseLikely bczfl
Branch Coprocessor z TrueLikely bcztl
Move From Coprocessor z mfcz dest-gpr, source
Move To Coprocessor z mtcz src-gpr, destination
Doubleword Move From Coprocessor z dmfcz
Doubleword Move To Coprocessor z dmtcz
Coprocessor z Operation cz expression

A-4 Assembly Language Programmer’s Guide

 Table A-1: Main Processor Instruction Summary
Description Op-code Operands

Appendix A
Control From Coprocessor z cfcz dest-gpr, source
Control To Coprocessor z ctcz src-gpr, destination
Branch on Equal Likely beql src1,src2,label
Branch on Greater Likely immediate,label bgtl src1,
Branch on Greater/Equal Likely bgel
Branch on Greater/Equal Unsigned Likely bgeul
Branch on Greater Unsigned Likely bgtul
Branch on Less Likely bltl
Branch on Less/Equal Likely blel
Branch on Less/Equal Unsigned Likely bleul
Branch on Less Unsigned Likely bltul
Branch on Not Equal Likely bnel
Branch on Equal to Zero beqz src1,label
Branch on Greater/Equal Zero bgez
Branch on Greater Than Zero bgtz
Branch on Greater or Equal to Zero and Link bgezal
Branch on Less Than Zero and Link bltzal
Branch on Less/Equal Zero blez
Branch on Less Than Zero bltz
Branch on Not Equal to Zero bnez
Branch on Equal to Zero Likely beqzl
Branch on Greater/Equal Zero Likely bgezl
Branch on Greater Than Zero Likely bgtzl
Branch on Greater or Equal to Zero and Link bgezall
Likely
Branch on Less Than Zero and Link Likely bltzall
Branch on Less/Equal Zero Likely blezl
Branch on Less Than Zero Likely bltzl
Branch on Not Equal to Zero Likely bnezl
Break break breakcode
Exception Return eret
Restore From Exception rfe
Syscall syscall
Move From HI Register mfhi register
Move To HI Register mthi
Move From LO Register mflo
Move To LO Register mtlo
Move move destination,src1

Assembly Language Programmer’s Guide A-5

 Appendix A

Table A-2: System Coprocessor Instruction Summary

Description Op-code Operand
Appendix A

Cache cache
Translation Lookaside Buffer Probe tlbp
Translation Lookaside Buffer Read tlbr
Translation Lookaside Buffer Write Random tlbwr
Translation Lookaside Write Index tlbwi
Synchronize sync

Table A-3: Floating Point Instruction Summary

Description Op-code Operand
Load Fp
Double l.d destination, address
Single l.s
Load Immediate Fp
Double li.d destination, floating-point constant
Single li.s
Store Fp
Double s.d source, address
Single s.s
Absolute Value Fp
Double abs.d destination, src1
Single abs.s
Negate Fp
Double neg.d
Single neg.s
Add Fp
Double add.d destination, src1, src2
Single add.s
Divide Fp
Double div.d
Single div.s
Multiply Fp
Double mul.d
Single mul.s
Subtract Fp
Double sub.d
Single sub.s
Convert Source to Specified Fp Precision
Double to Single Fp cvt.s.d destination, src1
Fixed Point to Single Fp cvt.s.w
Single to Double Fp cvt.d.s

A-6 Assembly Language Programmer’s Guide

 Table A-3: Floating Point Instruction Summary
Description Op-code Operand

Appendix A
Fixed Point to Double Fp cvt.d.w
Single to Fixed Point Fp cvt.w.s
Double to Fixed Point Fp cvt.w.d
Long Fixed Point to Single Fp cvt.s.l
Long Fixed Point to Double FP cvt.d.l
Single to Long Fixed Point FP cvt.l.s
Double to Long Fixed Point FP cvt.l.d
Truncate and Round Operations
Truncate to Single Fp trunc.w.s destination, src, gpr
Truncate to Double Fp trunc.w.d
Round to Single Fp round.w.s
Round to Double Fp round.w.d
Ceiling to Double Fp ceil.w.d
Ceiling to Single Fp ceil.w.s
Ceiling to Double Fp, Unsigned ceilu.w.d
Ceiling to Single Fp, Unsigned ceilu.w.s
Floor to Double Fp floor.w.d
Floor to Single Fp floor.w.s
Floor to Double Fp, Unsigned flooru.w.d
Floor to Single Fp, Unsigned flooru.w.s
Round to Double Fp, Unsigned roundu.w.d
Round to Single Fp, Unsigned roundu.w.s
Truncate to Double Fp, Unsigned truncu.w.d
Truncate to Single Fp, Unsigned truncu.w.s
Truncate Single to Long Fixed Point trunc.l.s destination, src, gpr
Truncate Double to Long Fixed Point trunc.l.d
Round Single to Long Fixed Point round.l.s
Round Double to Long Fixed Point round.l.d
Ceiling Single to Long Fixed Point ceil.l.s
Ceiling Double to Long Fixed Point ceil.l.d
Floor Single to Long Fixed Point floor.l.s
Floor Double to Long Fixed Point floor.l.d
Compare F
Double c.f.d src1,src2
Single c.f.s
Compare UN
Double c.un.d
Single c.un.s
*Compare EQ
Double c.eq.d
Single c.eq.s
Compare UEQ

Assembly Language Programmer’s Guide A-7

 Appendix A

Table A-3: Floating Point Instruction Summary

Description Op-code Operand
Double c.ueq.d
Appendix A

Single c.ueq.s
Compare OLT
Double c.olt.d
Single c.olt.s
Compare ULT
Double c.ult.d
Single c.ult.s
Compare OLE
Double c.ole.d
Single c.ole.s
Compare ULE
Double c.ule.d
Single c.ule.s
Compare SF
Double c.sf.d
Single c.sf.s
Compare NGLE
Double c.ngle.d src1, src2
Single c.ngle.s
Compare SEQ
Double c.seq.d
Single c.seq.s
Compare NGL
Double c.ngl.d
Single c.ngl.s
*Compare LT
Double c.lt.d
Single c.lt.s
Compare NGE
Double c.nge.d
Single c.nge.s
*Compare LE
Double c.le.d
Single c.le.s
Compare NGT
Double c.ngt.d
Single c.ngt.s
Move FP
Single mov.s destination,src1
Double mov.d

A-8 Assembly Language Programmer’s Guide

Basic Machine Definition

Appendix B
The assembly language instructions described in this book are distinct from
the actual machine instructions.
Generally, the assembly language instructions match the machine
instructions; however, in some cases the assembly language instruction are
macros that generate more than one machine instruction (the assembly
language multiplication instructions are examples).
Some machine instructions are not available as assembly language
instructions. For example, the jr machine instruction is not a valid assembly
language instruction. However, the j assembly language instruction with a
register operand gets translated into the jr machine instruction by the
assembler.
You can, in most instances, consider the assembly instructions as machine
instructions; however, for routines that require tight coding for performance
reasons, you must be aware of the assembly instructions that generate more
than one machine language instruction, as described in this appendix.

Load and Store Instructions

If you use an address as an operand in an assembler Load or Store instruction
and the address references a data item that is not addressable through register
$gp or the data item does not have an absolute address in the range –
32768...32767, the assembler instruction generates a lui (load upper
immediate) machine instruction and generates the appropriate offset to $at.
The assembler then uses $at as the index address for the reference. This
condition occurs when the address has a relocatable external name offset (or
index) from where the offset began.

Assembly Language Programmer’s Guide B-1

 Appendix B

The assembler’s la (load address) instruction generates an addiu (add

unsigned immediate) machine instruction. If the address requires it, the la
instruction also generates a lui (load upper immediate) machine instruction.
The machine requires the la instruction because la couples relocatable
information with the instruction for symbolic addresses.
Depending on the expression’s value, the assembler’s Ii (load immediate)
instruction can generate one or two machine instructions. For values in the –
32768...65535 range or for values that have zeros as the 16 least-significant
bits, the li instruction generates a single machine instruction; otherwise it
generates two machine instructions.
Appendix B

Computational Instructions
If a computational instruction immediate value falls outside the 0...65535
range for Logical ANDs, Logical ORs, or Logical XORs (exclusive or), the
immediate field causes the machine to explicitly load a constant to a
temporary register. Other instructions generate a single machine instruction
when a value falls in the –32768...32767 range.
The assembler’s seq (set equal) and sne (set not equal) instructions generate
three machine instructions each.
If one operand is a literal outside the range –32768...32767, the assembler’s
sge (set greater than or equal to) and sle (set less/equal) instructions generate
two machine instructions each.
The assembler’s mulo and mulou (multiply) instructions generate machine
instructions to test for overflow and to move the result to a general register;
if the destination register is $0, the check and move are not generated.
The assembler’s mul (multiply unsigned) instruction generates a machine
instruction to move the result to a general register; if the destination register
is $0, the move and divide–by–zero checking is not generated. The
assembler’s divide instructions, div (divide with overflow) and divu (divide
without overflow), generate machine instructions to check for division by
zero and to move the quotient into a general register; if the destination register
is $0, the move is not generated.
The assembler’s rem (signed) and remu (unsigned) instructions also generate
multiple instructions.
The rotate instructions ror (rotate right) and rol (rotate left) generate three
machine instructions each.
The abs (absolute value) instruction generates three machine instructions.

B-2 Assembly Language Programmer’s Guide

Branch Instructions
If the immediate value is not zero, the branch instructions beq (branch on
equal) and bne (branch on not equal), each generate a load literal machine
instruction. The relational instructions generate a slt (set less than) machine
instruction to determine whether one register is less than or greater than
another. Relational instructions can reorder the operands and branch on either
zero or not zero as required to do an operation.

Coprocessor Instructions
For symbolic addresses, the coprocessor interface Load and Store

Appendix B
instructions, lcz (load coprocessor z) and scz (store coprocessor z) can
generate a lui (load upper immediate) machine instruction.

Special Instructions
The assembler’s break instruction packs the breakcode operand in unused
register fields. An operating system convention determines the position.

Assembly Language Programmer’s Guide B-3

 Appendix B
Appendix B

B-4 Assembly Language Programmer’s Guide

Index

Symbols .sdata 8-6

(symbolic equate) 8-8 .set 8-7
.aent name, symno 8-1 .space 8-8
.alias 8-1 .struct 8-8
.align 8-1 .text 8-8
.ascii 8-2 .verstamp 8-8
.asciiz 8-2 .vreg 8-8
.asm0 8-2 .word 8-8
.bgnb 8-2
.byte 8-2 A
.comm 8-2 address
.data 8-2 description 2-3
.double 8-3 descriptions 2-3
.dword 8-3 format 2-2
.end 8-3 addressing 2-1
.endb 8-3 alignment 2-1
.endr 8-3 aligned data
.ent 8-3 load and store instructions 2-1
.err 8-3 alignment 2-1
.extern name expression 8-3 addressing 2-1
.file 8-4 allocation
.float 8-4 memory 7-15
.fmask 8-4 archive files 9-30
.frame 8-4 assembler 2-1
.galive 8-4 tokens 4-1
.gjaldef 8-4 assembler processing 9-16
.gjrlive 8-4 auxiliary symbols 10-5
.globl 8-4 format 10-17
.half 8-5
.lab 8-5 B
.lcomm 8-5 base address 12-4
.livereg 8-5 basic machine definition B-1
.loc 8-6 branch instructions B-3
.mask 8-6 filling delay slots 5-1
.option 8-6
.rdata 8-6
.repeat 8-6

Assembly Language Programmer’s Guide Index-1

Index

C dynamic linking 12-1, 12-9

COFF 9-1 dynamic section 12-11
comments 4-2
company address ii E
computational instructions 5-1, 5-10, B-2 ELF header 11-3
descriptions - table 5-13 ELF symbol table 11-18
format 5-10 endian
conflict section 12-26 Big-endian (figure) 1-2
constants 4-2 little-endian (figure) 1-2
floating point 4-3 endianness 1-1
scalar 4-3 exception
string 4-4 division by zero 6-17
convention unimplemented operation 6-19
linkage and register use 7-2 exception trap processing 6-16
conventions exceptions 3-1
data types 4-8 floating point 3-2
expression operators 4-8 main processor 3-1
expressions 4-7 execption
lexical 4-1 inexact 6-19
linkage 7-1 invalid operation 6-17
precedence 4-7 overflow 6-18
statements 4-6 trap processing 6-16
coprocessor instruction underflow 6-18
notation 6-1 execution and linking
coprocessor instruction set 6-1 format 11-1
coprocessor instructions B-3 expression
coprocessor interface instructions 5-27 type propagation 4-10
description of 5-28 expression operators 4-8
copyrights ii expressions 4-7
counters precedence 4-7
sections and locations 4-5 external relocation entries 9-15
customer service telephone numbers ii external strings 10-6
cycles per instruction 5-2 external symbols 10-7
format 10-21
D
data types F
conventions 4-8 f_magic 9-5
demad paged files 9-25 file descriptor 10-6
dense numbers file descriptor table
symbol table 10-4 format 10-20
description file header
address 2-3 format 9-4
descriptions magic field 9-5
load instructions 5-4 flags 9-10
division by zero 6-17 flags (f_flags) 9-5

Index-2 Assembly Language Programmer’s Guide

 Index

floating point coprocessor interface - description

computational - description 6-7 5-27, 5-28
computational - format 6-4 coprocessor interface format 5-27
control register 6-15 floating point 6-2
exceptions 3-2 instruction notation 5-2
instruction format 6-3 jump and branch 5-21
instructions 6-2 load and store 5-3
load and store 6-3 load and store - unaligned data 2-1
move instruction - description of 6-13 miscellaneous tasks 5-25
move instructions - format 6-13 pipeline 5-2
relational instruction - description 6-11 reorganization rules 5-2
relational instruction formats 6-10 special 5-25
relational operations 6-8 invalid operation exception 6-17
rounding 6-20 issues 4-6
floating point constants 4-3
floating point registers - table 7-2 J
format jump and branch instructions 5-1, 5-21
address 2-2 descriptions 5-23
formats formats 5-21
load and store 5-3
K
G keyword statements 4-7
-G value
link editor 4-6 L
general registers 1-3 label definitions
global data area 11-23 statements 4-6
global offset table 12-19 leaf routines 7-3
global pointer tables 9-11 lexical conventions 4-1
LIBMAGIC 9-1
H LIBMAGIC Files 9-28
hash table 12-23 line numbers
format 10-9
I symbol table 10-4
identifiers 4-2 link editor
impure format files 9-23 -G option 4-6
inexact exception 6-19 link editor defined symbols 9-31
initialization functions 12-23 link editor processing 9-16
instruction set 5-1 linkage
coprocessor 6-1 conventions 7-1
instruction summary A-1 program design 7-2
instructions register use 7-2
classes of 5-1 load 2-1
computational 5-10 load and store
constraints and rules 5-2 floating point 6-3
coprocessor interface 5-27 load and store instructions 5-3, B-1

Assembly Language Programmer’s Guide Index-3

Index

formats 5-3 P
load instructions performance 5-2
delayed 5-1 maximizing 5-2
description 5-4 pipeline
lb (load byte) 2-2 instruction 5-2
lbu (load byte unsigned) 2-2 position independent functions 12-20
lh (load halfword) 2-1 precedence in expressions 4-7
lhu (load halfword unsigned) 2-1 procedure descriptor table 10-4
lw (load word) 2-1 format 10-13
lwl (load word left) 2-1 program design
lwr (load word right) 2-1 linkage 7-2
ulh (unaligned load halfword program header 12-2
unsigned) 2-1 program interpreter
ulh (unaligned load halfword) 2-1 dynamic linking 12-9
ulw (unaligned load word) 2-1 program loading 12-1, 12-6
loading object Files 9-29 pseudo op-codes 8-1
local strings 10-6
local symbols 10-4 Q
fomat 10-13 quickstart 12-24

M R
memory allocation 7-15 Register 1-1
move instructions register 1-1
floating point 6-13 endianness 1-1
format 1-1
N register information 11-25
NMAGIC Files 9-24 registers
NMAGIC, 9-1 general 1-3
noalias 8-6 special 1-5
non-leaf routines 7-3 relational operations
nop 8-6 floating point 6-8
null statements 4-7 relative file descriptor 10-7
relocation 11-26
O relocation table 9-15
object file relocation type 9-16
format 9-1 relocations 12-22
object file format 11-2 runtime procedure table symbols 9-32
object files 9-22
OMAGIC 9-1 S
OMAGIC Files 9-23 scalar constants 4-3
optional header 9-7 section data 9-12
magic field 9-8 section header 11-8
ordering 12-26 section header table 11-7
overflow exception 6-18 section headers 9-8
section name 9-9

Index-4 Assembly Language Programmer’s Guide

 Index

section relocation 9-15 T

segment contents 12-5 target shared library files 9-28
segment permissions 12-4 telephone numbers
shape of data 7-7 customer service ii
shared libraries 9-12 termination functions 12-23
objects using 9-28 tokens
shared object dependencies 12-18 comments 4-2
shared object list 12-24 constants 4-2
shared text files 9-24 identifiers 4-2
shown 2-2 trademarks ii
special instructions 5-1, 5-25, B-3 type propagation in expression 4-10
description 5-25
format 5-25 U
special registers 1-5 ucode objects 9-29
special sections 11-14 unaligned data
stack frame 7-3 load and store instructions 2-1
stack organization- figure 7-4 underflow exception 6-18
statements unimplemented operation exception 6-19
keyword 4-7
label definitions 4-6 V
null 4-7 value 4-6
storage class (st) constants 10-16
store instructions Z
description 5-7 ZMAGIC 9-1
description - table 5-7 ZMAGIC Files 9-25
format 5-3
sb (store byte) 2-2
sh (store halfword) 2-2
sw (store word) 2-2
swl (store word left) 2-1
swr (store word right) 2-1
ush (unaligned store halfword) 2-1
usw (unaligned store word) 2-1
string constants 4-4
string tables 11-18
symbol table 10-1
symbol type 11-21
symbol type (st) 10-16
symbol values 11-22
symbolic header 10-3
format 10-8
symbols 12-22
system control
instruction descriptions 6-14
instruction formats 6-13

Assembly Language Programmer’s Guide Index-5

Index

Index-6 Assembly Language Programmer’s Guide