SPNU151V January   1998  – February 2020

 

  1.   Read This First
    1.     About This Manual
    2.     Notational Conventions
    3.     Related Documentation
    4.     Related Documentation From Texas Instruments
    5.     Trademarks
  2. 1Introduction to the Software Development Tools
    1. 1.1 Software Development Tools Overview
    2. 1.2 Compiler Interface
    3. 1.3 ANSI/ISO Standard
    4. 1.4 Output Files
    5. 1.5 Utilities
  3. 2Using the C/C++ Compiler
    1. 2.1  About the Compiler
    2. 2.2  Invoking the C/C++ Compiler
    3. 2.3  Changing the Compiler's Behavior with Options
      1. 2.3.1  Linker Options
      2. 2.3.2  Frequently Used Options
      3. 2.3.3  Miscellaneous Useful Options
      4. 2.3.4  Run-Time Model Options
      5. 2.3.5  Symbolic Debugging and Profiling Options
      6. 2.3.6  Specifying Filenames
      7. 2.3.7  Changing How the Compiler Interprets Filenames
      8. 2.3.8  Changing How the Compiler Processes C Files
      9. 2.3.9  Changing How the Compiler Interprets and Names Extensions
      10. 2.3.10 Specifying Directories
      11. 2.3.11 Assembler Options
      12. 2.3.12 Deprecated Options
    4. 2.4  Controlling the Compiler Through Environment Variables
      1. 2.4.1 Setting Default Compiler Options (TI_ARM_C_OPTION)
      2. 2.4.2 Naming One or More Alternate Directories (TI_ARM_C_DIR)
    5. 2.5  Controlling the Preprocessor
      1. 2.5.1  Predefined Macro Names
      2. 2.5.2  The Search Path for #include Files
        1. 2.5.2.1 Adding a Directory to the #include File Search Path (--include_path Option)
      3. 2.5.3  Support for the #warning and #warn Directives
      4. 2.5.4  Generating a Preprocessed Listing File (--preproc_only Option)
      5. 2.5.5  Continuing Compilation After Preprocessing (--preproc_with_compile Option)
      6. 2.5.6  Generating a Preprocessed Listing File with Comments (--preproc_with_comment Option)
      7. 2.5.7  Generating Preprocessed Listing with Line-Control Details (--preproc_with_line Option)
      8. 2.5.8  Generating Preprocessed Output for a Make Utility (--preproc_dependency Option)
      9. 2.5.9  Generating a List of Files Included with #include (--preproc_includes Option)
      10. 2.5.10 Generating a List of Macros in a File (--preproc_macros Option)
    6. 2.6  Passing Arguments to main()
    7. 2.7  Understanding Diagnostic Messages
      1. 2.7.1 Controlling Diagnostic Messages
      2. 2.7.2 How You Can Use Diagnostic Suppression Options
    8. 2.8  Other Messages
    9. 2.9  Generating Cross-Reference Listing Information (--gen_cross_reference Option)
    10. 2.10 Generating a Raw Listing File (--gen_preprocessor_listing Option)
    11. 2.11 Using Inline Function Expansion
      1. 2.11.1 Inlining Intrinsic Operators
      2. 2.11.2 Inlining Restrictions
    12. 2.12 Using Interlist
      1.      Example 1. An Interlisted Assembly Language File
    13. 2.13 Controlling Application Binary Interface
    14. 2.14 VFP Support
    15. 2.15 Enabling Entry Hook and Exit Hook Functions
  4. 3Optimizing Your Code
    1. 3.1  Invoking Optimization
    2. 3.2  Controlling Code Size Versus Speed
    3. 3.3  Performing File-Level Optimization (--opt_level=3 option)
      1. 3.3.1 Creating an Optimization Information File (--gen_opt_info Option)
    4. 3.4  Program-Level Optimization (--program_level_compile and --opt_level=3 options)
      1. 3.4.1 Controlling Program-Level Optimization (--call_assumptions Option)
      2. 3.4.2 Optimization Considerations When Mixing C/C++ and Assembly
    5. 3.5  Automatic Inline Expansion (--auto_inline Option)
    6. 3.6  Link-Time Optimization (--opt_level=4 Option)
      1. 3.6.1 Option Handling
      2. 3.6.2 Incompatible Types
    7. 3.7  Using Feedback Directed Optimization
      1. 3.7.1 Feedback Directed Optimization
        1. 3.7.1.1 Phase 1 -- Collect Program Profile Information
        2. 3.7.1.2 Phase 2 -- Use Application Profile Information for Optimization
        3. 3.7.1.3 Generating and Using Profile Information
        4. 3.7.1.4 Example Use of Feedback Directed Optimization
        5. 3.7.1.5 The .ppdata Section
        6. 3.7.1.6 Feedback Directed Optimization and Code Size Tune
        7. 3.7.1.7 Instrumented Program Execution Overhead
        8. 3.7.1.8 Invalid Profile Data
      2. 3.7.2 Profile Data Decoder
      3. 3.7.3 Feedback Directed Optimization API
      4. 3.7.4 Feedback Directed Optimization Summary
    8. 3.8  Using Profile Information to Analyze Code Coverage
      1. 3.8.1 Code Coverage
        1. 3.8.1.1 Phase1 -- Collect Program Profile Information
        2. 3.8.1.2 Phase 2 -- Generate Code Coverage Reports
      2. 3.8.2 Related Features and Capabilities
        1. 3.8.2.1 Path Profiler
        2. 3.8.2.2 Analysis Options
        3. 3.8.2.3 Environment Variables
    9. 3.9  Accessing Aliased Variables in Optimized Code
    10. 3.10 Use Caution With asm Statements in Optimized Code
    11. 3.11 Using the Interlist Feature With Optimization
      1.      Example 1. The Function From Compiled With the -O2 and --optimizer_interlist Options
      2.      Example 2. The Function From Compiled with the --opt_level=2, --optimizer_interlist, and --c_src_interlist Options
    12. 3.12 Debugging and Profiling Optimized Code
      1. 3.12.1 Profiling Optimized Code
    13. 3.13 What Kind of Optimization Is Being Performed?
      1. 3.13.1  Cost-Based Register Allocation
      2. 3.13.2  Alias Disambiguation
      3. 3.13.3  Branch Optimizations and Control-Flow Simplification
      4. 3.13.4  Data Flow Optimizations
      5. 3.13.5  Expression Simplification
      6. 3.13.6  Inline Expansion of Functions
      7. 3.13.7  Function Symbol Aliasing
      8. 3.13.8  Induction Variables and Strength Reduction
      9. 3.13.9  Loop-Invariant Code Motion
      10. 3.13.10 Loop Rotation
      11. 3.13.11 Instruction Scheduling
      12. 3.13.12 Tail Merging
      13. 3.13.13 Autoincrement Addressing
      14. 3.13.14 Block Conditionalizing
        1.       Example 3. Block Conditionalizing C Source
        2.       Example 4. C/C++ Compiler Output for
      15. 3.13.15 Epilog Inlining
      16. 3.13.16 Removing Comparisons to Zero
      17. 3.13.17 Integer Division With Constant Divisor
      18. 3.13.18 Branch Chaining
  5. 4Linking C/C++ Code
    1. 4.1 Invoking the Linker Through the Compiler (-z Option)
      1. 4.1.1 Invoking the Linker Separately
      2. 4.1.2 Invoking the Linker as Part of the Compile Step
      3. 4.1.3 Disabling the Linker (--compile_only Compiler Option)
    2. 4.2 Linker Code Optimizations
      1. 4.2.1 Generate List of Dead Functions (--generate_dead_funcs_list Option)
      2. 4.2.2 Generating Aggregate Data Subsections (--gen_data_subsections Compiler Option)
    3. 4.3 Controlling the Linking Process
      1. 4.3.1 Including the Run-Time-Support Library
        1. 4.3.1.1 Automatic Run-Time-Support Library Selection
          1.        Example 1. Using the --issue_remarks Option
        2. 4.3.1.2 Manual Run-Time-Support Library Selection
        3. 4.3.1.3 Library Order for Searching for Symbols
      2. 4.3.2 Run-Time Initialization
      3. 4.3.3 Initialization of Cinit and Watchdog Timer Hold
      4. 4.3.4 Global Object Constructors
      5. 4.3.5 Specifying the Type of Global Variable Initialization
      6. 4.3.6 Specifying Where to Allocate Sections in Memory
      7. 4.3.7 A Sample Linker Command File
        1.       Example 2. Linker Command File
  6. 5C/C++ Language Implementation
    1. 5.1  Characteristics of ARM C
      1. 5.1.1 Implementation-Defined Behavior
    2. 5.2  Characteristics of ARM C++
    3. 5.3  Using MISRA C 2004
    4. 5.4  Using the ULP Advisor
    5. 5.5  Data Types
      1. 5.5.1 Size of Enum Types
    6. 5.6  File Encodings and Character Sets
    7. 5.7  Keywords
      1. 5.7.1 The const Keyword
      2. 5.7.2 The __interrupt Keyword
      3. 5.7.3 The volatile Keyword
        1.       Example 1. Volatile for Local Variables With setjmp
    8. 5.8  C++ Exception Handling
    9. 5.9  Register Variables and Parameters
      1. 5.9.1 Local Register Variables and Parameters
      2. 5.9.2 Global Register Variables
    10. 5.10 The __asm Statement
    11. 5.11 Pragma Directives
      1. 5.11.1  The CALLS Pragma
      2. 5.11.2  The CHECK_MISRA Pragma
      3. 5.11.3  The CHECK_ULP Pragma
      4. 5.11.4  The CODE_ALIGN Pragma
      5. 5.11.5  The CODE_SECTION Pragma
        1.       Example 2. Using the CODE_SECTION Pragma C Source File
        2.       Example 3. Generated Assembly Code From
      6. 5.11.6  The CODE_STATE Pragma
      7. 5.11.7  The DATA_ALIGN Pragma
      8. 5.11.8  The DATA_SECTION Pragma
        1.       Example 4. Using the DATA_SECTION Pragma C Source File
        2.       Example 5. Using the DATA_SECTION Pragma C++ Source File
        3.       Example 6. Using the DATA_SECTION Pragma Assembly Source File
      9. 5.11.9  The Diagnostic Message Pragmas
      10. 5.11.10 The DUAL_STATE Pragma
      11. 5.11.11 The FORCEINLINE Pragma
      12. 5.11.12 The FORCEINLINE_RECURSIVE Pragma
      13. 5.11.13 The FUNC_ALWAYS_INLINE Pragma
      14. 5.11.14 The FUNC_CANNOT_INLINE Pragma
      15. 5.11.15 The FUNC_EXT_CALLED Pragma
      16. 5.11.16 The FUNCTION_OPTIONS Pragma
      17. 5.11.17 The INTERRUPT Pragma
      18. 5.11.18 The LOCATION Pragma
      19. 5.11.19 The MUST_ITERATE Pragma
        1. 5.11.19.1 The MUST_ITERATE Pragma Syntax
        2. 5.11.19.2 Using MUST_ITERATE to Expand Compiler Knowledge of Loops
      20. 5.11.20 The NOINIT and PERSISTENT Pragmas
      21. 5.11.21 The NOINLINE Pragma
      22. 5.11.22 The NO_HOOKS Pragma
      23. 5.11.23 The once Pragma
      24. 5.11.24 The pack Pragma
      25. 5.11.25 The RESET_MISRA Pragma
      26. 5.11.26 The RESET_ULP Pragma
      27. 5.11.27 The RETAIN Pragma
      28. 5.11.28 The SET_CODE_SECTION and SET_DATA_SECTION Pragmas
        1.       Example 7. Setting Section With SET_DATA_SECTION Pragma
        2.       Example 8. Setting a Section With SET_CODE_SECTION Pragma
        3.       Example 9. Overriding SET_DATA_SECTION Setting
      29. 5.11.29 The SWI_ALIAS Pragma
        1.       Example 10. Using the SWI_ALIAS Pragma C Source File
        2.       Example 11. Generated Assembly File
      30. 5.11.30 The TASK Pragma
      31. 5.11.31 The UNROLL Pragma
      32. 5.11.32 The WEAK Pragma
    12. 5.12 The _Pragma Operator
    13. 5.13 Application Binary Interface
    14. 5.14 ARM Instruction Intrinsics
    15. 5.15 Object File Symbol Naming Conventions (Linknames)
    16. 5.16 Changing the ANSI/ISO C/C++ Language Mode
      1. 5.16.1 C99 Support (--c99)
      2. 5.16.2 C11 Support (--c11)
      3. 5.16.3 Strict ANSI Mode and Relaxed ANSI Mode (--strict_ansi and --relaxed_ansi)
    17. 5.17 GNU, Clang, and ACLE Language Extensions
      1. 5.17.1 Extensions
      2. 5.17.2 Function Attributes
      3. 5.17.3 Variable Attributes
      4. 5.17.4 Type Attributes
      5. 5.17.5 Built-In Functions
    18. 5.18 AUTOSAR
    19. 5.19 Compiler Limits
  7. 6Run-Time Environment
    1. 6.1  Memory Model
      1. 6.1.1 Sections
      2. 6.1.2 C/C++ System Stack
      3. 6.1.3 Dynamic Memory Allocation
    2. 6.2  Object Representation
      1. 6.2.1 Data Type Storage
        1. 6.2.1.1 char and short Data Types (signed and unsigned)
        2. 6.2.1.2 float, int, and long Data Types (signed and unsigned)
        3. 6.2.1.3 double, long double, and long long Data Types (signed and unsigned)
        4. 6.2.1.4 Pointer to Data Member Types
        5. 6.2.1.5 Pointer to Member Function Types
        6. 6.2.1.6 Structure and Array Alignment
      2. 6.2.2 Bit Fields
      3. 6.2.3 Character String Constants
    3. 6.3  Register Conventions
    4. 6.4  Function Structure and Calling Conventions
      1. 6.4.1 How a Function Makes a Call
      2. 6.4.2 How a Called Function Responds
      3. 6.4.3 C Exception Handler Calling Convention
      4. 6.4.4 Accessing Arguments and Local Variables
    5. 6.5  Accessing Linker Symbols in C and C++
    6. 6.6  Interfacing C and C++ With Assembly Language
      1. 6.6.1 Using Assembly Language Modules With C/C++ Code
      2. 6.6.2 Accessing Assembly Language Functions From C/C++
        1.       Example 1. Calling an Assembly Language Function From a C/C++ Program
        2.       Example 2. Assembly Language Program Called by
      3. 6.6.3 Accessing Assembly Language Variables From C/C++
        1. 6.6.3.1 Accessing Assembly Language Global Variables
          1.        Example 3. Assembly Language Variable Program
          2.        Example 4. C Program to Access Assembly Language From
        2. 6.6.3.2 Accessing Assembly Language Constants
          1.        Example 5. Accessing an Assembly Language Constant From C
          2.        Example 6. Assembly Language Program for
      4. 6.6.4 Sharing C/C++ Header Files With Assembly Source
      5. 6.6.5 Using Inline Assembly Language
      6. 6.6.6 Modifying Compiler Output
    7. 6.7  Interrupt Handling
      1. 6.7.1 Saving Registers During Interrupts
      2. 6.7.2 Using C/C++ Interrupt Routines
      3. 6.7.3 Using Assembly Language Interrupt Routines
      4. 6.7.4 How to Map Interrupt Routines to Interrupt Vectors
        1.       Example 7. Sample intvecs.asm File
      5. 6.7.5 Using Software Interrupts
      6. 6.7.6 Other Interrupt Information
    8. 6.8  Intrinsic Run-Time-Support Arithmetic and Conversion Routines
      1. 6.8.1 CPSR Register and Interrupt Intrinsics
    9. 6.9  Built-In Functions
    10. 6.10 System Initialization
      1. 6.10.1 Boot Hook Functions for System Pre-Initialization
      2. 6.10.2 Run-Time Stack
      3. 6.10.3 Automatic Initialization of Variables
        1. 6.10.3.1 Zero Initializing Variables
        2. 6.10.3.2 Direct Initialization
        3. 6.10.3.3 Autoinitialization of Variables at Run Time
        4. 6.10.3.4 Autoinitialization Tables
          1. 6.10.3.4.1 Length Followed by Data Format
          2. 6.10.3.4.2 Zero Initialization Format
          3. 6.10.3.4.3 Run Length Encoded (RLE) Format
          4. 6.10.3.4.4 Lempel-Ziv-Storer-Szymanski Compression (LZSS) Format
          5. 6.10.3.4.5 Sample C Code to Process the C Autoinitialization Table
            1.         Example 8. Processing the C Autoinitialization Table
        5. 6.10.3.5 Initialization of Variables at Load Time
        6. 6.10.3.6 Global Constructors
      4. 6.10.4 Initialization Tables
        1.       Example 9. Initialized Variables Defined in C
        2.       Example 10. Initialized Information for Variables Defined in
    11. 6.11 Dual-State Interworking Under TIABI (Deprecated)
      1. 6.11.1 Level of Dual-State Support
      2. 6.11.2 Implementation
        1. 6.11.2.1 Naming Conventions for Entry Points
        2. 6.11.2.2 Indirect Calls
          1.        Example 11. C Code Compiled for 16-BIS State: sum( )
          2.        Example 12. 16-Bit Assembly Program for
          3.        Example 13. C Code Compiled for 32-BIS State: sum( )
          4.        Example 14. 32-Bit Assembly Program for
  8. 7Using Run-Time-Support Functions and Building Libraries
    1. 7.1 C and C++ Run-Time Support Libraries
      1. 7.1.1 Linking Code With the Object Library
      2. 7.1.2 Header Files
      3. 7.1.3 Modifying a Library Function
      4. 7.1.4 Support for String Handling
      5. 7.1.5 Minimal Support for Internationalization
      6. 7.1.6 Allowable Number of Open Files
      7. 7.1.7 Nonstandard Header Files in the Source Tree
      8. 7.1.8 Library Naming Conventions
    2. 7.2 The C I/O Functions
      1. 7.2.1 High-Level I/O Functions
        1. 7.2.1.1 Formatting and the Format Conversion Buffer
      2. 7.2.2 Overview of Low-Level I/O Implementation
      3. 7.2.3 Device-Driver Level I/O Functions
      4. 7.2.4 Adding a User-Defined Device Driver for C I/O
        1.       Example 1. Mapping Default Streams to Device
      5. 7.2.5 The device Prefix
        1.       Example 2. Program for C I/O Device
    3. 7.3 Handling Reentrancy (_register_lock() and _register_unlock() Functions)
    4. 7.4 Library-Build Process
      1. 7.4.1 Required Non-Texas Instruments Software
      2. 7.4.2 Using the Library-Build Process
        1. 7.4.2.1 Automatic Standard Library Rebuilding by the Linker
        2. 7.4.2.2 Invoking mklib Manually
          1. 7.4.2.2.1 Building Standard Libraries
          2. 7.4.2.2.2 Shared or Read-Only Library Directory
          3. 7.4.2.2.3 Building Libraries With Custom Options
          4. 7.4.2.2.4 The mklib Program Option Summary
      3. 7.4.3 Extending mklib
        1. 7.4.3.1 Underlying Mechanism
        2. 7.4.3.2 Libraries From Other Vendors
  9. 8C++ Name Demangler
    1. 8.1 Invoking the C++ Name Demangler
    2. 8.2 Sample Usage of the C++ Name Demangler
      1.      Example 1. C++ Code for calories_in_a_banana
      2.      Example 2. Resulting Assembly for calories_in_a_banana
      3.      Example 3. Result After Running the C++ Name Demangler
  10.   A Glossary
    1.     A.1 Terminology
  11.   B Revision History
    1.     B.1 Recent Revisions

2.11 Using Inline Function Expansion

When an inline function is called, a copy of the C/C++ source code for the function is inserted at the point of the call. This is known as inline function expansion, commonly called function inlining or just inlining. Inline function expansion can speed up execution by eliminating function call overhead. This is particularly beneficial for very small functions that are called frequently. Function inlining involves a tradeoff between execution speed and code size, because the code is duplicated at each function call site. Large functions that are called in many places are poor candidates for inlining.

NOTE

Excessive Inlining Can Degrade Performance

Excessive inlining can make the compiler dramatically slower and degrade the performance of generated code.

Function inlining is triggered by the following situations:

  • The use of built-in intrinsic operations. Intrinsic operations look like function calls, and are inlined automatically, even though no function body exists.
  • Use of the inline keyword or the equivalent __inline keyword. Functions declared with the inline keyword may be inlined by the compiler if you set --opt_level=0 or greater. The inline keyword is a suggestion from the programmer to the compiler. Even if your optimization level is high, inlining is still optional for the compiler. The compiler decides whether to inline a function based on the length of the function, the number of times it is called, your --opt_for_speed setting, and any contents of the function that disqualify it from inlining (see Section 2.11.2). Functions can be inlined at --opt_level=0 or above if the function body is visible in the same module or if -pm is also used and the function is visible in one of the modules being compiled. Functions may be inlined at link time if the file containing the definition and the call site were both compiled with --opt_level=4. Functions defined as both static and inline are more likely to be inlined.
  • When --opt_level=3 or greater is used, the compiler may automatically inline eligible functions even if they are not declared as inline functions. The same list of decision factors listed for functions explicitly defined with the inline keyword is used. For more about automatic function inlining, see Section 3.5.
  • The pragma FUNC_ALWAYS_INLINE (Section 5.11.13) and the equivalent always_inline attribute (Section 5.17.2) force a function to be inlined (where it is legal to do so) unless --opt_level=off. That is, the pragma FUNC_ALWAYS_INLINE forces function inlining even if the function is not declared as inline and the --opt_level=0 or --opt_level=1.
  • The FORCEINLINE pragma (Section 5.11.11) forces functions to be inlined in the annotated statement. That is, it has no effect on those functions in general, only on function calls in a single statement. The FORCEINLINE_RECURSIVE pragma forces inlining not only of calls visible in the statement, but also in the inlined bodies of calls from that statement.
  • The --disable_inlining option prevents any inlining. The pragma FUNC_CANNOT_INLINE prevents a function from being inlined. The NOINLINE pragma prevents calls within a single statement from being inlined. (NOINLINE is the inverse of the FORCEINLINE pragma.)

NOTE

Function Inlining Can Greatly Increase Code Size

Function inlining increases code size, especially inlining a function that is called in a number of places. Function inlining is optimal for functions that are called only from a small number of places and for small functions.

The semantics of the inline keyword in C code follow the C99 standard. The semantics of the inline keyword in C++ code follow the C++ standard.

The inline keyword is supported in all C++ modes, in relaxed ANSI mode for all C standards, and in strict ANSI mode for C99 and C11. It is disabled in strict ANSI mode for C89, because it is a language extension that could conflict with a strictly conforming program. If you want to define inline functions while in strict ANSI C89 mode, use the alternate keyword __inline.

Compiler options that affect inlining are: --opt_level, --auto_inline, --remove_hooks_when_inlining, --opt_for_speed, and --disable_inlining.