Using and Porting GNU CC

This manual documents how to run, install and port the GNU compiler, as well as its new features and incompatibilities, and how to report bugs. It corresponds to GNU CC version 2.7.2.

• GNU GENERAL PUBLIC LICENSE
  • Preamble
  • TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
  • How to Apply These Terms to Your New Programs
• Contributors to GNU CC
• 2 Funding Free Software
• 3 Protect Your Freedom--Fight "Look And Feel"
• 4 Compile C, C++, or Objective C
• 5 GNU CC Command Options
  • 5.1 Option Summary
  • 5.2 Options Controlling the Kind of Output
  • 5.3 Compiling C++ Programs
  • 5.4 Options Controlling C Dialect
  • 5.5 Options Controlling C++ Dialect
  • 5.6 Options to Request or Suppress Warnings
  • 5.7 Options for Debugging Your Program or GNU CC
  • 5.8 Options That Control Optimization
  • 5.9 Options Controlling the Preprocessor
  • 5.10 Passing Options to the Assembler
  • 5.11 Options for Linking
  • 5.12 Options for Directory Search
  • 5.13 Specifying Target Machine and Compiler Version
  • 5.14 Hardware Models and Configurations
    • 5.14.1 M680x0 Options
    • 5.14.2 VAX Options
    • 5.14.3 SPARC Options
    • 5.14.4 Convex Options
    • 5.14.5 AMD29K Options
    • 5.14.6 ARM Options
    • 5.14.7 M88K Options
    • 5.14.8 IBM RS/6000 and PowerPC Options
    • 5.14.9 IBM RT Options
    • 5.14.10 MIPS Options
    • 5.14.11 Intel 386 Options
    • 5.14.12 HPPA Options
    • 5.14.13 Intel 960 Options
    • 5.14.14 DEC Alpha Options
    • 5.14.15 Clipper Options
    • 5.14.16 H8/300 Options
    • 5.14.17 Options for System V
  • 5.15 Options for Code Generation Conventions
  • 5.16 Environment Variables Affecting GNU CC
  • 5.17 Running Protoize
• 6 Installing GNU CC
  • 6.1 Configurations Supported by GNU CC
  • 6.2 Compilation in a Separate Directory
  • 6.3 Building and Installing a Cross-Compiler
    • 6.3.1 Steps of Cross-Compilation
    • 6.3.2 Configuring a Cross-Compiler
    • 6.3.3 Tools and Libraries for a Cross-Compiler
    • 6.3.4 `libgcc.a' and Cross-Compilers
    • 6.3.5 Cross-Compilers and Header Files
    • 6.3.6 Actually Building the Cross-Compiler
  • 6.4 Installing GNU CC on the Sun
  • 6.5 Installing GNU CC on VMS
  • 6.6 collect2
  • 6.7 Standard Header File Directories
• 7 Extensions to the C Language Family
  • 7.1 Statements and Declarations in Expressions
  • 7.2 Locally Declared Labels
  • 7.3 Labels as Values
  • 7.4 Nested Functions
  • 7.5 Constructing Function Calls
  • 7.6 Naming an Expression's Type
  • 7.7 Referring to a Type with typeof
  • 7.8 Generalized Lvalues
  • 7.9 Conditionals with Omitted Operands
  • 7.10 Double-Word Integers
  • 7.11 Complex Numbers
  • 7.12 Arrays of Length Zero
  • 7.13 Arrays of Variable Length
  • 7.14 Macros with Variable Numbers of Arguments
  • 7.15 Non-Lvalue Arrays May Have Subscripts
  • 7.16 Arithmetic on void- and Function-Pointers
  • 7.17 Non-Constant Initializers
  • 7.18 Constructor Expressions
  • 7.19 Labeled Elements in Initializers
  • 7.20 Case Ranges
  • 7.21 Cast to a Union Type
  • 7.22 Declaring Attributes of Functions
  • 7.23 Prototypes and Old-Style Function Definitions
  • 7.24 C++ Style Comments
  • 7.25 Dollar Signs in Identifier Names
  • 7.26 The Character ESC in Constants
  • 7.27 Inquiring on Alignment of Types or Variables
  • 7.28 Specifying Attributes of Variables
  • 7.29 Specifying Attributes of Types
  • 7.30 An Inline Function is As Fast As a Macro
  • 7.31 Assembler Instructions with C Expression Operands
  • 7.32 Controlling Names Used in Assembler Code
  • 7.33 Variables in Specified Registers
    • 7.33.1 Defining Global Register Variables
    • 7.33.2 Specifying Registers for Local Variables
  • 7.34 Alternate Keywords
  • 7.35 Incomplete enum Types
  • 7.36 Function Names as Strings
• 8 Extensions to the C++ Language
  • 8.1 Named Return Values in C++
  • 8.2 Minimum and Maximum Operators in C++
  • 8.3 goto and Destructors in GNU C++
  • 8.4 Declarations and Definitions in One Header
  • 8.5 Where's the Template?
  • 8.6 Type Abstraction using Signatures
• 9 Known Causes of Trouble with GNU CC
  • 9.1 Actual Bugs We Haven't Fixed Yet
  • 9.2 Installation Problems
  • 9.3 Cross-Compiler Problems
  • 9.4 Interoperation
  • 9.5 Problems Compiling Certain Programs
  • 9.6 Incompatibilities of GNU CC
  • 9.7 Fixed Header Files
  • 9.8 Standard Libraries
  • 9.9 Disappointments and Misunderstandings
  • 9.10 Common Misunderstandings with GNU C++
    • 9.10.1 Declare and Define Static Members
    • 9.10.2 Temporaries May Vanish Before You Expect
  • 9.11 Caveats of using protoize
  • 9.12 Certain Changes We Don't Want to Make
  • 9.13 Warning Messages and Error Messages
• 10 Reporting Bugs
  • 10.1 Have You Found a Bug?
  • 10.2 Where to Report Bugs
  • 10.3 How to Report Bugs
  • 10.4 Sending Patches for GNU CC
• 11 How To Get Help with GNU CC
• 12 Using GNU CC on VMS
  • 12.1 Include Files and VMS
  • 12.2 Global Declarations and VMS
  • 12.3 Other VMS Issues
• 13 GNU CC and Portability
• 14 Interfacing to GNU CC Output
• 15 Passes and Files of the Compiler
• 16 RTL Representation
  • 16.1 RTL Object Types
  • 16.2 Access to Operands
  • 16.3 Flags in an RTL Expression
  • 16.4 Machine Modes
  • 16.5 Constant Expression Types
  • 16.6 Registers and Memory
  • 16.7 RTL Expressions for Arithmetic
  • 16.8 Comparison Operations
  • 16.9 Bit Fields
  • 16.10 Conversions
  • 16.11 Declarations
  • 16.12 Side Effect Expressions
  • 16.13 Embedded Side-Effects on Addresses
  • 16.14 Assembler Instructions as Expressions
  • 16.15 Insns
  • 16.16 RTL Representation of Function-Call Insns
  • 16.17 Structure Sharing Assumptions
  • 16.18 Reading RTL
• 17 Machine Descriptions
  • 17.1 Everything about Instruction Patterns
  • 17.2 Example of define_insn
  • 17.3 RTL Template
  • 17.4 Output Templates and Operand Substitution
  • 17.5 C Statements for Assembler Output
  • 17.6 Operand Constraints
    • 17.6.1 Simple Constraints
    • 17.6.2 Multiple Alternative Constraints
    • 17.6.3 Register Class Preferences
    • 17.6.4 Constraint Modifier Characters
    • 17.6.5 Constraints for Particular Machines
    • 17.6.6 Not Using Constraints
  • 17.7 Standard Pattern Names For Generation
  • 17.8 When the Order of Patterns Matters
  • 17.9 Interdependence of Patterns
  • 17.10 Defining Jump Instruction Patterns
  • 17.11 Canonicalization of Instructions
  • 17.12 Machine-Specific Peephole Optimizers
  • 17.13 Defining RTL Sequences for Code Generation
  • 17.14 Defining How to Split Instructions
  • 17.15 Instruction Attributes
    • 17.15.1 Defining Attributes and their Values
    • 17.15.2 Attribute Expressions
    • 17.15.3 Assigning Attribute Values to Insns
    • 17.15.4 Example of Attribute Specifications
    • 17.15.5 Computing the Length of an Insn
    • 17.15.6 Constant Attributes
    • 17.15.7 Delay Slot Scheduling
    • 17.15.8 Specifying Function Units
• 18 Target Description Macros
  • 18.1 Controlling the Compilation Driver, `gcc'
  • 18.2 Run-time Target Specification
  • 18.3 Storage Layout
  • 18.4 Layout of Source Language Data Types
  • 18.5 Register Usage
    • 18.5.1 Basic Characteristics of Registers
    • 18.5.2 Order of Allocation of Registers
    • 18.5.3 How Values Fit in Registers
    • 18.5.4 Handling Leaf Functions
    • 18.5.5 Registers That Form a Stack
    • 18.5.6 Obsolete Macros for Controlling Register Usage
  • 18.6 Register Classes
  • 18.7 Stack Layout and Calling Conventions
    • 18.7.1 Basic Stack Layout
    • 18.7.2 Registers That Address the Stack Frame
    • 18.7.3 Eliminating Frame Pointer and Arg Pointer
    • 18.7.4 Passing Function Arguments on the Stack
    • 18.7.5 Passing Arguments in Registers
    • 18.7.6 How Scalar Function Values Are Returned
    • 18.7.7 How Large Values Are Returned
    • 18.7.8 Caller-Saves Register Allocation
    • 18.7.9 Function Entry and Exit
    • 18.7.10 Generating Code for Profiling
  • 18.8 Implementing the Varargs Macros
  • 18.9 Trampolines for Nested Functions
  • 18.10 Implicit Calls to Library Routines
  • 18.11 Addressing Modes
  • 18.12 Condition Code Status
  • 18.13 Describing Relative Costs of Operations
  • 18.14 Dividing the Output into Sections (Texts, Data, ...)
  • 18.15 Position Independent Code
  • 18.16 Defining the Output Assembler Language
    • 18.16.1 The Overall Framework of an Assembler File
    • 18.16.2 Output of Data
    • 18.16.3 Output of Uninitialized Variables
    • 18.16.4 Output and Generation of Labels
    • 18.16.5 How Initialization Functions Are Handled
    • 18.16.6 Macros Controlling Initialization Routines
    • 18.16.7 Output of Assembler Instructions
    • 18.16.8 Output of Dispatch Tables
    • 18.16.9 Assembler Commands for Alignment
  • 18.17 Controlling Debugging Information Format
    • 18.17.1 Macros Affecting All Debugging Formats
    • 18.17.2 Specific Options for DBX Output
    • 18.17.3 Open-Ended Hooks for DBX Format
    • 18.17.4 File Names in DBX Format
    • 18.17.5 Macros for SDB and DWARF Output
  • 18.18 Cross Compilation and Floating Point
  • 18.19 Miscellaneous Parameters
• 19 The Configuration File
• 20 Makefile Fragments
  • 20.1 The Target Makefile Fragment
  • 20.2 The Host Makefile Fragment
• Index

This document was generated on 8 June 2000 using the texi2html translator version 1.51a.