C

C
Notes for Professionals

Notes for Professionals

300+ pages
of professional hints and tricks

Disclaimer
GoalKicker.com This is an unocial free book created for educational purposes and is
not aliated with ocial C group(s) or company(s).
Free Programming Books All trademarks and registered trademarks are
the property of their respective owners
Contents
About ................................................................................................................................................................................... 1
Chapter 1: Getting started with C Language .................................................................................................. 2
Section 1.1: Hello World ................................................................................................................................................. 2
Section 1.2: Original "Hello, World!" in K&R C .............................................................................................................. 4
Chapter 2: Comments ................................................................................................................................................. 6
Section 2.1: Commenting using the preprocessor ...................................................................................................... 6
Section 2.2: /* */ delimited comments ........................................................................................................................ 6
Section 2.3: // delimited comments ............................................................................................................................ 7
Section 2.4: Possible pitfall due to trigraphs .............................................................................................................. 7
Chapter 3: Data Types ............................................................................................................................................... 9
Section 3.1: Interpreting Declarations .......................................................................................................................... 9
Section 3.2: Fixed Width Integer Types (since C99) ................................................................................................. 11
Section 3.3: Integer types and constants .................................................................................................................. 11
Section 3.4: Floating Point Constants ........................................................................................................................ 12
Section 3.5: String Literals .......................................................................................................................................... 13
Chapter 4: Operators ............................................................................................................................................... 14
Section 4.1: Relational Operators ............................................................................................................................... 14
Section 4.2: Conditional Operator/Ternary Operator ............................................................................................. 15
Section 4.3: Bitwise Operators ................................................................................................................................... 16
Section 4.4: Short circuit behavior of logical operators .......................................................................................... 18
Section 4.5: Comma Operator ................................................................................................................................... 19
Section 4.6: Arithmetic Operators .............................................................................................................................. 19
Section 4.7: Access Operators ................................................................................................................................... 22
Section 4.8: sizeof Operator ....................................................................................................................................... 24
Section 4.9: Cast Operator ......................................................................................................................................... 24
Section 4.10: Function Call Operator ......................................................................................................................... 24
Section 4.11: Increment / Decrement ......................................................................................................................... 25
Section 4.12: Assignment Operators .......................................................................................................................... 25
Section 4.13: Logical Operators .................................................................................................................................. 26
Section 4.14: Pointer Arithmetic .................................................................................................................................. 27
Section 4.15: _Alignof .................................................................................................................................................. 28
Chapter 5: Boolean .................................................................................................................................................... 30
Section 5.1: Using stdbool.h ........................................................................................................................................ 30
Section 5.2: Using #define .......................................................................................................................................... 30
Section 5.3: Using the Intrinsic (built-in) Type _Bool ............................................................................................... 31
Section 5.4: Integers and pointers in Boolean expressions .................................................................................... 31
Section 5.5: Defining a bool type using typedef ...................................................................................................... 32
Chapter 6: Strings ....................................................................................................................................................... 33
Section 6.1: Tokenisation: strtok(), strtok_r() and strtok_s() .................................................................................. 33
Section 6.2: String literals ........................................................................................................................................... 35
Section 6.3: Calculate the Length: strlen() ................................................................................................................ 36
Section 6.4: Basic introduction to strings .................................................................................................................. 37
Section 6.5: Copying strings ....................................................................................................................................... 37
Section 6.6: Iterating Over the Characters in a String ............................................................................................. 40
Section 6.7: Creating Arrays of Strings ..................................................................................................................... 41
Section 6.8: Convert Strings to Number: atoi(), atof() (dangerous, don't use them) ........................................... 41
Section 6.9: string formatted data read/write ......................................................................................................... 42
 Section 6.10: Find first/last occurrence of a specific character: strchr(), strrchr() ............................................... 43
Section 6.11: Copy and Concatenation: strcpy(), strcat() ........................................................................................ 44
Section 6.12: Comparsion: strcmp(), strncmp(), strcasecmp(), strncasecmp() .................................................... 45
Section 6.13: Safely convert Strings to Number: strtoX functions .......................................................................... 47
Section 6.14: strspn and strcspn ................................................................................................................................. 48
Chapter 7: Literals for numbers, characters and strings ...................................................................... 50
Section 7.1: Floating point literals ............................................................................................................................... 50
Section 7.2: String literals ........................................................................................................................................... 50
Section 7.3: Character literals .................................................................................................................................... 50
Section 7.4: Integer literals ......................................................................................................................................... 51
Chapter 8: Compound Literals ............................................................................................................................. 53
Section 8.1: Definition/Initialisation of Compound Literals ...................................................................................... 53
Chapter 9: Bit-fields .................................................................................................................................................. 55
Section 9.1: Bit-fields .................................................................................................................................................... 55
Section 9.2: Using bit-fields as small integers .......................................................................................................... 56
Section 9.3: Bit-field alignment .................................................................................................................................. 56
Section 9.4: Don'ts for bit-fields ................................................................................................................................. 57
Section 9.5: When are bit-fields useful? .................................................................................................................... 58
Chapter 10: Arrays ...................................................................................................................................................... 60
Section 10.1: Declaring and initializing an array ....................................................................................................... 60
Section 10.2: Iterating through an array eciently and row-major order ............................................................ 61
Section 10.3: Array length ........................................................................................................................................... 62
Section 10.4: Passing multidimensional arrays to a function ................................................................................. 63
Section 10.5: Multi-dimensional arrays ...................................................................................................................... 64
Section 10.6: Define array and access array element ............................................................................................. 67
Section 10.7: Clearing array contents (zeroing) ....................................................................................................... 67
Section 10.8: Setting values in arrays ........................................................................................................................ 68
Section 10.9: Allocate and zero-initialize an array with user defined size ............................................................. 68
Section 10.10: Iterating through an array using pointers ........................................................................................ 69
Chapter 11: Linked lists ............................................................................................................................................. 71
Section 11.1: A doubly linked list .................................................................................................................................. 71
Section 11.2: Reversing a linked list ............................................................................................................................ 73
Section 11.3: Inserting a node at the nth position ..................................................................................................... 75
Section 11.4: Inserting a node at the beginning of a singly linked list .................................................................... 76
Chapter 12: Enumerations ...................................................................................................................................... 79
Section 12.1: Simple Enumeration ............................................................................................................................... 79
Section 12.2: enumeration constant without typename .......................................................................................... 80
Section 12.3: Enumeration with duplicate value ....................................................................................................... 80
Section 12.4: Typedef enum ....................................................................................................................................... 81
Chapter 13: Structs ..................................................................................................................................................... 83
Section 13.1: Flexible Array Members ......................................................................................................................... 83
Section 13.2: Typedef Structs ..................................................................................................................................... 85
Section 13.3: Pointers to structs .................................................................................................................................. 86
Section 13.4: Passing structs to functions ................................................................................................................. 88
Section 13.5: Object-based programming using structs ......................................................................................... 89
Section 13.6: Simple data structures .......................................................................................................................... 91
Chapter 14: Standard Math ................................................................................................................................... 93
Section 14.1: Power functions - pow(), powf(), powl() .............................................................................................. 93
Section 14.2: Double precision floating-point remainder: fmod() .......................................................................... 94
 Section 14.3: Single precision and long double precision floating-point remainder: fmodf(), fmodl() ............... 94
Chapter 15: Iteration Statements/Loops: for, while, do-while ............................................................ 96
Section 15.1: For loop ................................................................................................................................................... 96
Section 15.2: Loop Unrolling and Du's Device ........................................................................................................ 96
Section 15.3: While loop ............................................................................................................................................... 97
Section 15.4: Do-While loop ........................................................................................................................................ 97
Section 15.5: Structure and flow of control in a for loop ......................................................................................... 98
Section 15.6: Infinite Loops .......................................................................................................................................... 99
Chapter 16: Selection Statements .................................................................................................................... 100
Section 16.1: if () Statements ..................................................................................................................................... 100
Section 16.2: Nested if()...else VS if()..else Ladder .................................................................................................. 100
Section 16.3: switch () Statements ........................................................................................................................... 102
Section 16.4: if () ... else statements and syntax ..................................................................................................... 104
Section 16.5: if()...else Ladder Chaining two or more if () ... else statements ....................................................... 104
Chapter 17: Initialization ........................................................................................................................................ 105
Section 17.1: Initialization of Variables in C .............................................................................................................. 105
Section 17.2: Using designated initializers ............................................................................................................... 106
Section 17.3: Initializing structures and arrays of structures ................................................................................ 108
Chapter 18: Declaration vs Definition ............................................................................................................ 110
Section 18.1: Understanding Declaration and Definition ....................................................................................... 110
Chapter 19: Command-line arguments ......................................................................................................... 111
Section 19.1: Print the arguments to a program and convert to integer values ................................................. 111
Section 19.2: Printing the command line arguments ............................................................................................. 111
Section 19.3: Using GNU getopt tools ...................................................................................................................... 112
Chapter 20: Files and I/O streams .................................................................................................................. 115
Section 20.1: Open and write to file ......................................................................................................................... 115
Section 20.2: Run process ........................................................................................................................................ 116
Section 20.3: fprintf ................................................................................................................................................... 116
Section 20.4: Get lines from a file using getline() .................................................................................................. 116
Section 20.5: fscanf() ................................................................................................................................................ 120
Section 20.6: Read lines from a file ......................................................................................................................... 121
Section 20.7: Open and write to a binary file ......................................................................................................... 122
Chapter 21: Formatted Input/Output ............................................................................................................. 124
Section 21.1: Conversion Specifiers for printing ...................................................................................................... 124
Section 21.2: The printf() Function ........................................................................................................................... 125
Section 21.3: Printing format flags ........................................................................................................................... 125
Section 21.4: Printing the Value of a Pointer to an Object .................................................................................... 126
Section 21.5: Printing the Dierence of the Values of two Pointers to an Object ............................................... 127
Section 21.6: Length modifiers ................................................................................................................................. 128
Chapter 22: Pointers ................................................................................................................................................ 129
Section 22.1: Introduction ......................................................................................................................................... 129
Section 22.2: Common errors .................................................................................................................................. 131
Section 22.3: Dereferencing a Pointer .................................................................................................................... 134
Section 22.4: Dereferencing a Pointer to a struct .................................................................................................. 134
Section 22.5: Const Pointers ..................................................................................................................................... 135
Section 22.6: Function pointers ................................................................................................................................ 138
Section 22.7: Polymorphic behaviour with void pointers ...................................................................................... 139
Section 22.8: Address-of Operator ( & ) ................................................................................................................. 140
Section 22.9: Initializing Pointers ............................................................................................................................. 140
 Section 22.10: Pointer to Pointer .............................................................................................................................. 141
Section 22.11: void* pointers as arguments and return values to standard functions ....................................... 141
Section 22.12: Same Asterisk, Dierent Meanings ................................................................................................. 142
Chapter 23: Sequence points .............................................................................................................................. 144
Section 23.1: Unsequenced expressions .................................................................................................................. 144
Section 23.2: Sequenced expressions ..................................................................................................................... 144
Section 23.3: Indeterminately sequenced expressions ......................................................................................... 145
Chapter 24: Function Pointers ........................................................................................................................... 146
Section 24.1: Introduction .......................................................................................................................................... 146
Section 24.2: Returning Function Pointers from a Function ................................................................................. 146
Section 24.3: Best Practices ..................................................................................................................................... 147
Section 24.4: Assigning a Function Pointer ............................................................................................................. 149
Section 24.5: Mnemonic for writing function pointers ........................................................................................... 149
Section 24.6: Basics ................................................................................................................................................... 150
Chapter 25: Function Parameters .................................................................................................................... 152
Section 25.1: Parameters are passed by value ...................................................................................................... 152
Section 25.2: Passing in Arrays to Functions .......................................................................................................... 152
Section 25.3: Order of function parameter execution ........................................................................................... 153
Section 25.4: Using pointer parameters to return multiple values ...................................................................... 153
Section 25.5: Example of function returning struct containing values with error codes ................................... 154
Chapter 26: Pass 2D-arrays to functions ..................................................................................................... 156
Section 26.1: Pass a 2D-array to a function ........................................................................................................... 156
Section 26.2: Using flat arrays as 2D arrays .......................................................................................................... 162
Chapter 27: Error handling .................................................................................................................................. 163
Section 27.1: errno ..................................................................................................................................................... 163
Section 27.2: strerror ................................................................................................................................................. 163
Section 27.3: perror ................................................................................................................................................... 163
Chapter 28: Undefined behavior ...................................................................................................................... 165
Section 28.1: Dereferencing a pointer to variable beyond its lifetime ................................................................. 165
Section 28.2: Copying overlapping memory .......................................................................................................... 165
Section 28.3: Signed integer overflow ..................................................................................................................... 166
Section 28.4: Use of an uninitialized variable ......................................................................................................... 167
Section 28.5: Data race ............................................................................................................................................ 168
Section 28.6: Read value of pointer that was freed .............................................................................................. 169
Section 28.7: Using incorrect format specifier in printf ......................................................................................... 170
Section 28.8: Modify string literal ............................................................................................................................ 170
Section 28.9: Passing a null pointer to printf %s conversion ................................................................................ 170
Section 28.10: Modifying any object more than once between two sequence points ....................................... 171
Section 28.11: Freeing memory twice ...................................................................................................................... 172
Section 28.12: Bit shifting using negative counts or beyond the width of the type ............................................ 172
Section 28.13: Returning from a function that's declared with `_Noreturn` or `noreturn` function specifier
............................................................................................................................................................................. 173
Section 28.14: Accessing memory beyond allocated chunk ................................................................................. 174
Section 28.15: Modifying a const variable using a pointer .................................................................................... 174
Section 28.16: Reading an uninitialized object that is not backed by memory .................................................. 175
Section 28.17: Addition or subtraction of pointer not properly bounded ............................................................ 175
Section 28.18: Dereferencing a null pointer ............................................................................................................ 175
Section 28.19: Using ush on an input stream ...................................................................................................... 176
Section 28.20: Inconsistent linkage of identifiers ................................................................................................... 176
Section 28.21: Missing return statement in value returning function ................................................................... 177
 Section 28.22: Division by zero ................................................................................................................................ 177
Section 28.23: Conversion between pointer types produces incorrectly aligned result .................................... 178
Section 28.24: Modifying the string returned by getenv, strerror, and setlocale functions .............................. 179
Chapter 29: Random Number Generation ................................................................................................... 180
Section 29.1: Basic Random Number Generation .................................................................................................. 180
Section 29.2: Permuted Congruential Generator ................................................................................................... 180
Section 29.3: Xorshift Generation ............................................................................................................................ 181
Section 29.4: Restrict generation to a given range ............................................................................................... 182
Chapter 30: Preprocessor and Macros .......................................................................................................... 183
Section 30.1: Header Include Guards ....................................................................................................................... 183
Section 30.2: #if 0 to block out code sections ........................................................................................................ 186
Section 30.3: Function-like macros .......................................................................................................................... 187
Section 30.4: Source file inclusion ............................................................................................................................ 188
Section 30.5: Conditional inclusion and conditional function signature modification ....................................... 188
Section 30.6: __cplusplus for using C externals in C++ code compiled with C++ - name mangling ............... 190
Section 30.7: Token pasting ..................................................................................................................................... 191
Section 30.8: Predefined Macros ............................................................................................................................. 192
Section 30.9: Variadic arguments macro ............................................................................................................... 193
Section 30.10: Macro Replacement ......................................................................................................................... 194
Section 30.11: Error directive ..................................................................................................................................... 195
Section 30.12: FOREACH implementation ............................................................................................................... 196
Chapter 31: Signal handling ................................................................................................................................. 199
Section 31.1: Signal Handling with “signal()” ............................................................................................................ 199
Chapter 32: Variable arguments ...................................................................................................................... 201
Section 32.1: Using an explicit count argument to determine the length of the va_list .................................... 201
Section 32.2: Using terminator values to determine the end of va_list .............................................................. 202
Section 32.3: Implementing functions with a `printf()`-like interface ................................................................... 202
Section 32.4: Using a format string ......................................................................................................................... 205
Chapter 33: Assertion ............................................................................................................................................. 207
Section 33.1: Simple Assertion .................................................................................................................................. 207
Section 33.2: Static Assertion ................................................................................................................................... 207
Section 33.3: Assert Error Messages ....................................................................................................................... 208
Section 33.4: Assertion of Unreachable Code ........................................................................................................ 209
Section 33.5: Precondition and Postcondition ........................................................................................................ 209
Chapter 34: Generic selection ............................................................................................................................ 211
Section 34.1: Check whether a variable is of a certain qualified type .................................................................. 211
Section 34.2: Generic selection based on multiple arguments ............................................................................. 211
Section 34.3: Type-generic printing macro ............................................................................................................ 213
Chapter 35: X-macros ............................................................................................................................................ 214
Section 35.1: Trivial use of X-macros for printfs ..................................................................................................... 214
Section 35.2: Extension: Give the X macro as an argument ................................................................................. 214
Section 35.3: Enum Value and Identifier ................................................................................................................. 215
Section 35.4: Code generation ................................................................................................................................. 215
Chapter 36: Aliasing and eective type ....................................................................................................... 217
Section 36.1: Eective type ....................................................................................................................................... 217
Section 36.2: restrict qualification ............................................................................................................................ 217
Section 36.3: Changing bytes ................................................................................................................................... 218
Section 36.4: Character types cannot be accessed through non-character types ........................................... 219
Section 36.5: Violating the strict aliasing rules ....................................................................................................... 220
Chapter 37: Compilation ....................................................................................................................................... 221
Section 37.1: The Compiler ........................................................................................................................................ 221
Section 37.2: File Types ............................................................................................................................................ 222
Section 37.3: The Linker ............................................................................................................................................ 222
Section 37.4: The Preprocessor ............................................................................................................................... 224
Section 37.5: The Translation Phases ...................................................................................................................... 225
Chapter 38: Inline assembly ................................................................................................................................ 227
Section 38.1: gcc Inline assembly in macros ........................................................................................................... 227
Section 38.2: gcc Basic asm support ...................................................................................................................... 227
Section 38.3: gcc Extended asm support ................................................................................................................ 228
Chapter 39: Identifier Scope ............................................................................................................................... 229
Section 39.1: Function Prototype Scope .................................................................................................................. 229
Section 39.2: Block Scope ......................................................................................................................................... 230
Section 39.3: File Scope ............................................................................................................................................ 230
Section 39.4: Function scope .................................................................................................................................... 231
Chapter 40: Implicit and Explicit Conversions ........................................................................................... 232
Section 40.1: Integer Conversions in Function Calls ............................................................................................... 232
Section 40.2: Pointer Conversions in Function Calls .............................................................................................. 233
Chapter 41: Type Qualifiers ................................................................................................................................ 235
Section 41.1: Volatile variables .................................................................................................................................. 235
Section 41.2: Unmodifiable (const) variables ......................................................................................................... 236
Chapter 42: Typedef .............................................................................................................................................. 237
Section 42.1: Typedef for Structures and Unions ................................................................................................... 237
Section 42.2: Typedef for Function Pointers .......................................................................................................... 238
Section 42.3: Simple Uses of Typedef ..................................................................................................................... 239
Chapter 43: Storage Classes .............................................................................................................................. 241
Section 43.1: auto ....................................................................................................................................................... 241
Section 43.2: register ................................................................................................................................................ 241
Section 43.3: static ..................................................................................................................................................... 242
Section 43.4: typedef ................................................................................................................................................ 243
Section 43.5: extern ................................................................................................................................................... 243
Section 43.6: _Thread_local .................................................................................................................................... 244
Chapter 44: Declarations .................................................................................................................................... 246
Section 44.1: Calling a function from another C file ............................................................................................... 246
Section 44.2: Using a Global Variable ..................................................................................................................... 247
Section 44.3: Introduction ......................................................................................................................................... 247
Section 44.4: Typedef ............................................................................................................................................... 250
Section 44.5: Using Global Constants ..................................................................................................................... 250
Section 44.6: Using the right-left or spiral rule to decipher C declaration .......................................................... 252
Chapter 45: Structure Padding and Packing ............................................................................................. 256
Section 45.1: Packing structures .............................................................................................................................. 256
Section 45.2: Structure padding .............................................................................................................................. 257
Chapter 46: Memory management ................................................................................................................ 258
Section 46.1: Allocating Memory .............................................................................................................................. 258
Section 46.2: Freeing Memory ................................................................................................................................. 259
Section 46.3: Reallocating Memory ......................................................................................................................... 261
Section 46.4: realloc(ptr, 0) is not equivalent to free(ptr) ..................................................................................... 262
Section 46.5: Multidimensional arrays of variable size ......................................................................................... 262
Section 46.6: alloca: allocate memory on stack .................................................................................................... 263
 Section 46.7: User-defined memory management ............................................................................................... 264
Chapter 47: Implementation-defined behaviour ..................................................................................... 266
Section 47.1: Right shift of a negative integer ........................................................................................................ 266
Section 47.2: Assigning an out-of-range value to an integer ............................................................................... 266
Section 47.3: Allocating zero bytes .......................................................................................................................... 266
Section 47.4: Representation of signed integers ................................................................................................... 266
Chapter 48: Atomics ............................................................................................................................................... 267
Section 48.1: atomics and operators ....................................................................................................................... 267
Chapter 49: Jump Statements .......................................................................................................................... 268
Section 49.1: Using return ......................................................................................................................................... 268
Section 49.2: Using goto to jump out of nested loops .......................................................................................... 268
Section 49.3: Using break and continue ................................................................................................................. 269
Chapter 50: Create and include header files ............................................................................................. 271
Section 50.1: Introduction ......................................................................................................................................... 271
Section 50.2: Self-containment ................................................................................................................................ 271
Section 50.3: Minimality ............................................................................................................................................ 273
Section 50.4: Notation and Miscellany .................................................................................................................... 273
Section 50.5: Idempotence ....................................................................................................................................... 275
Section 50.6: Include What You Use (IWYU) ........................................................................................................... 275
Chapter 51: <ctype.h> — character classification & conversion ....................................................... 277
Section 51.1: Introduction .......................................................................................................................................... 277
Section 51.2: Classifying characters read from a stream ..................................................................................... 278
Section 51.3: Classifying characters from a string ................................................................................................. 279
Chapter 52: Side Eects ....................................................................................................................................... 280
Section 52.1: Pre/Post Increment/Decrement operators ..................................................................................... 280
Chapter 53: Multi-Character Character Sequence .................................................................................. 282
Section 53.1: Trigraphs .............................................................................................................................................. 282
Section 53.2: Digraphs .............................................................................................................................................. 282
Chapter 54: Constraints ........................................................................................................................................ 284
Section 54.1: Duplicate variable names in the same scope .................................................................................. 284
Section 54.2: Unary arithmetic operators .............................................................................................................. 284
Chapter 55: Inlining ................................................................................................................................................. 285
Section 55.1: Inlining functions used in more than one source file ....................................................................... 285
Chapter 56: Unions ................................................................................................................................................... 287
Section 56.1: Using unions to reinterpret values .................................................................................................... 287
Section 56.2: Writing to one union member and reading from another ............................................................. 287
Section 56.3: Dierence between struct and union ............................................................................................... 288
Chapter 57: Threads (native) ............................................................................................................................. 289
Section 57.1: Inititialization by one thread ............................................................................................................... 289
Section 57.2: Start several threads .......................................................................................................................... 289
Chapter 58: Multithreading ................................................................................................................................. 291
Section 58.1: C11 Threads simple example .............................................................................................................. 291
Chapter 59: Interprocess Communication (IPC) ........................................................................................ 292
Section 59.1: Semaphores ......................................................................................................................................... 292
Chapter 60: Testing frameworks ..................................................................................................................... 297
Section 60.1: Unity Test Framework ........................................................................................................................ 297
Section 60.2: CMocka ................................................................................................................................................ 297
Section 60.3: CppUTest ............................................................................................................................................. 298
Chapter 61: Valgrind ................................................................................................................................................ 300
Section 61.1: Bytes lost -- Forgetting to free ........................................................................................................... 300
Section 61.2: Most common errors encountered while using Valgrind ................................................................ 300
Section 61.3: Running Valgrind ................................................................................................................................. 301
Section 61.4: Adding flags ......................................................................................................................................... 301
Chapter 62: Common C programming idioms and developer practices ..................................... 302
Section 62.1: Comparing literal and variable .......................................................................................................... 302
Section 62.2: Do not leave the parameter list of a function blank — use void ................................................... 302
Chapter 63: Common pitfalls .............................................................................................................................. 305
Section 63.1: Mixing signed and unsigned integers in arithmetic operations ...................................................... 305
Section 63.2: Macros are simple string replacements .......................................................................................... 305
Section 63.3: Forgetting to copy the return value of realloc into a temporary .................................................. 307
Section 63.4: Forgetting to allocate one extra byte for \0 ................................................................................... 308
Section 63.5: Misunderstanding array decay ......................................................................................................... 308
Section 63.6: Forgetting to free memory (memory leaks) ................................................................................... 310
Section 63.7: Copying too much .............................................................................................................................. 311
Section 63.8: Mistakenly writing = instead of == when comparing ....................................................................... 312
Section 63.9: Newline character is not consumed in typical scanf() call ............................................................ 313
Section 63.10: Adding a semicolon to a #define .................................................................................................... 314
Section 63.11: Incautious use of semicolons ............................................................................................................ 314
Section 63.12: Undefined reference errors when linking ....................................................................................... 315
Section 63.13: Checking logical expression against 'true' ...................................................................................... 317
Section 63.14: Doing extra scaling in pointer arithmetic ....................................................................................... 318
Section 63.15: Multi-line comments cannot be nested ........................................................................................... 319
Section 63.16: Ignoring return values of library functions ..................................................................................... 321
Section 63.17: Comparing floating point numbers ................................................................................................. 321
Section 63.18: Floating point literals are of type double by default ..................................................................... 323
Section 63.19: Using character constants instead of string literals, and vice versa ........................................... 323
Section 63.20: Recursive function — missing out the base condition .................................................................. 324
Section 63.21: Overstepping array boundaries ...................................................................................................... 325
Section 63.22: Passing unadjacent arrays to functions expecting "real" multidimensional arrays ................. 326
Credits ............................................................................................................................................................................ 328
You may also like ...................................................................................................................................................... 333
About

Please feel free to share this PDF with anyone for free,
latest version of this book can be downloaded from:
https://goalkicker.com/CBook

This C Notes for Professionals book is compiled from Stack Overflow

Documentation, the content is written by the beautiful people at Stack Overflow.
Text content is released under Creative Commons BY-SA, see credits at the end
of this book whom contributed to the various chapters. Images may be copyright
of their respective owners unless otherwise specified

This is an unofficial free book created for educational purposes and is not
affiliated with official C group(s) or company(s) nor Stack Overflow. All
trademarks and registered trademarks are the property of their respective
company owners

The information presented in this book is not guaranteed to be correct nor

accurate, use at your own risk

Please send feedback and corrections to [email protected]

GoalKicker.com – C Notes for Professionals 1

Chapter 1: Getting started with C
Language
Version Standard Publication Date
K&R n/a 1978-02-22
C89 ANSI X3.159-1989 1989-12-14
C90 ISO/IEC 9899:1990 1990-12-20
C95 ISO/IEC 9899/AMD1:1995 1995-03-30
C99 ISO/IEC 9899:1999 1999-12-16
C11 ISO/IEC 9899:2011 2011-12-15

Section 1.1: Hello World

To create a simple C program which prints "Hello, World" on the screen, use a text editor to create a new file (e.g.
hello.c — the file extension must be .c) containing the following source code:

hello.c
#include <stdio.h>

int main(void)
{
puts("Hello, World");
return 0;
}

Live demo on Coliru

Let's look at this simple program line by line

#include <stdio.h>

This line tells the compiler to include the contents of the standard library header file stdio.h in the program.
Headers are usually files containing function declarations, macros and data types, and you must include the header
file before you use them. This line includes stdio.h so it can call the function puts().

See more about headers.

int main(void)

This line starts the definition of a function. It states the name of the function (main), the type and number of
arguments it expects (void, meaning none), and the type of value that this function returns (int). Program
execution starts in the main() function.

{
…
}

The curly braces are used in pairs to indicate where a block of code begins and ends. They can be used in a lot of
ways, but in this case they indicate where the function begins and ends.

puts("Hello, World");

This line calls the puts() function to output text to standard output (the screen, by default), followed by a newline.

GoalKicker.com – C Notes for Professionals 2

The string to be output is included within the parentheses.

"Hello, World" is the string that will be written to the screen. In C, every string literal value must be inside the
double quotes "…".

See more about strings.

In C programs, every statement needs to be terminated by a semi-colon (i.e. ;).

return 0;

When we defined main(), we declared it as a function returning an int, meaning it needs to return an integer. In
this example, we are returning the integer value 0, which is used to indicate that the program exited successfully.
After the return 0; statement, the execution process will terminate.

Editing the program

Simple text editors include vim or gedit on Linux, or Notepad on Windows. Cross-platform editors also include
Visual Studio Code or Sublime Text.

The editor must create plain text files, not RTF or other any other format.

Compiling and running the program

To run the program, this source file (hello.c) first needs to be compiled into an executable file (e.g. hello on
Unix/Linux system or hello.exe on Windows). This is done using a compiler for the C language.

See more about compiling

Compile using GCC

GCC (GNU Compiler Collection) is a widely used C compiler. To use it, open a terminal, use the command line to
navigate to the source file's location and then run:

gcc hello.c -o hello

If no errors are found in the the source code (hello.c), the compiler will create a binary file, the name of which is
given by the argument to the -o command line option (hello). This is the final executable file.

We can also use the warning options -Wall -Wextra -Werror, that help to identify problems that can cause the
program to fail or produce unexpected results. They are not necessary for this simple program but this is way of
adding them:

gcc -Wall -Wextra -Werror -o hello hello.c

Using the clang compiler

To compile the program using clang you can use:

clang -Wall -Wextra -Werror -o hello hello.c

By design, the clang command line options are similar to those of GCC.

Using the Microsoft C compiler from the command line

GoalKicker.com – C Notes for Professionals 3

If using the Microsoft cl.exe compiler on a Windows system which supports Visual Studio and if all environment
variables are set, this C example may be compiled using the following command which will produce an executable
hello.exe within the directory the command is executed in (There are warning options such as /W3 for cl, roughly
analogous to -Wall etc for GCC or clang).

cl hello.c

Executing the program

Once compiled, the binary file may then be executed by typing ./hello in the terminal. Upon execution, the
compiled program will print Hello, World, followed by a newline, to the command prompt.

Section 1.2: Original "Hello, World!" in K&R C

The following is the original "Hello, World!" program from the book The C Programming Language by Brian
Kernighan and Dennis Ritchie (Ritchie was the original developer of the C programming language at Bell Labs),
referred to as "K&R":

Version = K&R
#include <stdio.h>

main()
{
printf("hello, world\n");
}

Notice that the C programming language was not standardized at the time of writing the first edition of this book
(1978), and that this program will probably not compile on most modern compilers unless they are instructed to
accept C90 code.

This very first example in the K&R book is now considered poor quality, in part because it lacks an explicit return
type for main() and in part because it lacks a return statement. The 2nd edition of the book was written for the old
C89 standard. In C89, the type of main would default to int, but the K&R example does not return a defined value
to the environment. In C99 and later standards, the return type is required, but it is safe to leave out the return
statement of main (and only main), because of a special case introduced with C99 5.1.2.2.3 — it is equivalent to
returning 0, which indicates success.

The recommended and most portable form of main for hosted systems is int main (void) when the program does
not use any command line arguments, or int main(int argc, char **argv) when the program does use the
command line arguments.

C90 §5.1.2.2.3 Program termination

A return from the initial call to the main function is equivalent to calling the exit function with the value
returned by the main function as its argument. If the main function executes a return that specifies no
value, the termination status returned to the host environment is undefined.

C90 §6.6.6.4 The return statement

If a return statement without an expression is executed, and the value of the function call is used by the

GoalKicker.com – C Notes for Professionals 4

 caller, the behavior is undefined. Reaching the } that terminates a function is equivalent to executing a
return statement without an expression.

C99 §5.1.2.2.3 Program termination

If the return type of the main function is a type compatible with int, a return from the initial call to the
main function is equivalent to calling the exit function with the value returned by the main function as its
argument; reaching the } that terminates the main function returns a value of 0. If the return type is not
compatible with int, the termination status returned to the host environment is unspecified.

GoalKicker.com – C Notes for Professionals 5

Chapter 2: Comments
Comments are used to indicate something to the person reading the code. Comments are treated like a blank by
the compiler and do not change anything in the code's actual meaning. There are two syntaxes used for comments
in C, the original /* */ and the slightly newer //. Some documentation systems use specially formatted comments
to help produce the documentation for code.

Section 2.1: Commenting using the preprocessor

Large chunks of code can also be "commented out" using the preprocessor directives #if 0 and #endif. This is
useful when the code contains multi-line comments that otherwise would not nest.

#if 0 /* Starts the "comment", anything from here on is removed by preprocessor */

/* A large amount of code with multi-line comments */

int foo()
{
/* lots of code */
...

/* ... some comment describing the if statement ... */

if (someTest) {
/* some more comments */
return 1;
}

return 0;
}

#endif /* 0 */

/* code from here on is "uncommented" (included in compiled executable) */

...

Section 2.2: /* */ delimited comments

A comment starts with a forward slash followed immediately by an asterisk (/*), and ends as soon as an asterisk
immediately followed by a forward slash (*/) is encountered. Everything in between these character combinations
is a comment and is treated as a blank (basically ignored) by the compiler.

/* this is a comment */

The comment above is a single line comment. Comments of this /* type can span multiple lines, like so:

/* this is a
multi-line
comment */

Though it is not strictly necessary, a common style convention with multi-line comments is to put leading spaces
and asterisks on the lines subsequent to the first, and the /* and */ on new lines, such that they all line up:

/*
* this is a
* multi-line
* comment

GoalKicker.com – C Notes for Professionals 6

 */

The extra asterisks do not have any functional effect on the comment as none of them have a related forward
slash.

These /* type of comments can be used on their own line, at the end of a code line, or even within lines of code:

/* this comment is on its own line */

if (x && y) { /*this comment is at the end of a line */
if ((complexCondition1) /* this comment is within a line of code */
&& (complexCondition2)) {
/* this comment is within an if, on its own line */
}
}

Comments cannot be nested. This is because any subsequent /* will be ignored (as part of the comment) and the
first */ reached will be treated as ending the comment. The comment in the following example will not work:

/* outer comment, means this is ignored => /* attempted inner comment */ <= ends the comment, not
this one => */

To comment blocks of code that contain comments of this type, that would otherwise be nested, see the
Commenting using the preprocessor example below

Section 2.3: // delimited comments

Version ≥ C99

C99 introduced the use of C++-style single-line comments. This type of comment starts with two forward slashes
and runs to the end of a line:

// this is a comment

This type of comment does not allow multi-line comments, though it is possible to make a comment block by
adding several single line comments one after the other:

// each of these lines are a single-line comment

// note how each must start with
// the double forward-slash

This type of comment may be used on its own line or at the end of a code line. However, because they run to the
end of the line, they may not be used within a code line

// this comment is on its own line

if (x && y) { // this comment is at the end of a line
// this comment is within an if, on its own line
}

Section 2.4: Possible pitfall due to trigraphs

Version ≥ C99

While writing // delimited comments, it is possible to make a typographical error that affects their expected
operation. If one types:

GoalKicker.com – C Notes for Professionals 7

int x = 20; // Why did I do this??/

The / at the end was a typo but now will get interpreted into \. This is because the ??/ forms a trigraph.

The ??/ trigraph is actually a longhand notation for \, which is the line continuation symbol. This means that the
compiler thinks the next line is a continuation of the current line, that is, a continuation of the comment, which may
not be what is intended.

int foo = 20; // Start at 20 ??/

int bar = 0;

// The following will cause a compilation error (undeclared variable 'bar')

// because 'int bar = 0;' is part of the comment on the preceding line
bar += foo;

GoalKicker.com – C Notes for Professionals 8

Chapter 3: Data Types
Section 3.1: Interpreting Declarations
A distinctive syntactic peculiarity of C is that declarations mirror the use of the declared object as it would be in a
normal expression.

The following set of operators with identical precedence and associativity are reused in declarators, namely:

the unary * "dereference" operator which denotes a pointer;

the binary [] "array subscription" operator which denotes an array;
the (1+n)-ary () "function call" operator which denotes a function;
the () grouping parentheses which override the precedence and associativity of the rest of the listed
operators.

The above three operators have the following precedence and associativity:

Operator Relative Precedence Associativity

[] (array subscription) 1 Left-to-right
() (function call) 1 Left-to-right
* (dereference) 2 Right-to-left

When interpreting declarations, one has to start from the identifier outwards and apply the adjacent operators in
the correct order as per the above table. Each application of an operator can be substituted with the following
English words:

Expression Interpretation
thing[X] an array of size X of...
thing(t1, t2, t3) a function taking t1, t2, t3 and returning...
*thing a pointer to...

It follows that the beginning of the English interpretation will always start with the identifier and will end with the
type that stands on the left-hand side of the declaration.

Examples
char *names[20];

[] takes precedence over *, so the interpretation is: names is an array of size 20 of a pointer to char.

char (*place)[10];

In case of using parentheses to override the precedence, the * is applied first: place is a pointer to an array of size
10 of char.

int fn(long, short);

There is no precedence to worry about here: fn is a function taking long, short and returning int.

int *fn(void);

The () is applied first: fn is a function taking void and returning a pointer to int.

GoalKicker.com – C Notes for Professionals 9

int (*fp)(void);

Overriding the precedence of (): fp is a pointer to a function taking void and returning int.

int arr[5][8];

Multidimensional arrays are not an exception to the rule; the [] operators are applied in left-to-right order
according to the associativity in the table: arr is an array of size 5 of an array of size 8 of int.

int **ptr;

The two dereference operators have equal precedence, so the associativity takes effect. The operators are applied
in right-to-left order: ptr is a pointer to a pointer to an int.

Multiple Declarations
The comma can be used as a separator (*not* acting like the comma operator) in order to delimit multiple
declarations within a single statement. The following statement contains five declarations:
int fn(void), *ptr, (*fp)(int), arr[10][20], num;

The declared objects in the above example are:

fn: a function taking void and returning int;

ptr: a pointer to an int;
fp: a pointer to a function taking int and returning int;
arr: an array of size 10 of an array of size 20 of int;
num: int.

Alternative Interpretation

Because declarations mirror use, a declaration can also be interpreted in terms of the operators that could be
applied over the object and the final resulting type of that expression. The type that stands on the left-hand side is
the final result that is yielded after applying all operators.

/*
* Subscripting "arr" and dereferencing it yields a "char" result.
* Particularly: *arr[5] is of type "char".
*/
char *arr[20];

/*
* Calling "fn" yields an "int" result.
* Particularly: fn('b') is of type "int".
*/
int fn(char);

/*
* Dereferencing "fp" and then calling it yields an "int" result.
* Particularly: (*fp)() is of type "int".
*/
int (*fp)(void);

/*
* Subscripting "strings" twice and dereferencing it yields a "char" result.
* Particularly: *strings[5][15] is of type "char"
*/
char *strings[10][20];

GoalKicker.com – C Notes for Professionals 10

Section 3.2: Fixed Width Integer Types (since C99)
Version ≥ C99

The header <stdint.h> provides several fixed-width integer type definitions. These types are optional and only
provided if the platform has an integer type of the corresponding width, and if the corresponding signed type has a
two's complement representation of negative values.

See the remarks section for usage hints of fixed width types.

/* commonly used types include */

uint32_t u32 = 32; /* exactly 32-bits wide */

uint8_t u8 = 255; /* exactly 8-bits wide */

int64_t i64 = -65 /* exactly 64 bit in two's complement representation */

Section 3.3: Integer types and constants

Signed integers can be of these types (the int after short, or long is optional):

signed char c = 127; /* required to be 1 byte, see remarks for further information. */
signed short int si = 32767; /* required to be at least 16 bits. */
signed int i = 32767; /* required to be at least 16 bits */
signed long int li = 2147483647; /* required to be at least 32 bits. */
Version ≥ C99
signed long long int li = 2147483647; /* required to be at least 64 bits */

Each of these signed integer types has an unsigned version.

unsigned int i = 65535;

unsigned short = 2767;
unsigned char = 255;

For all types but char the signed version is assumed if the signed or unsigned part is omitted. The type char
constitutes a third character type, different from signed char and unsigned char and the signedness (or not)
depends on the platform.

Different types of integer constants (called literals in C jargon) can be written in different bases, and different width,
based on their prefix or suffix.

/* the following variables are initialized to the same value: */

int d = 42; /* decimal constant (base10) */
int o = 052; /* octal constant (base8) */
int x = 0xaf; /* hexadecimal constants (base16) */
int X = 0XAf; /* (letters 'a' through 'f' (case insensitive) represent 10 through 15) */

Decimal constants are always signed. Hexadecimal constants start with 0x or 0X and octal constants start just with
a 0. The latter two are signed or unsigned depending on whether the value fits into the signed type or not.

/* suffixes to describe width and signedness : */

long int i = 0x32; /* no suffix represent int, or long int */
unsigned int ui = 65535u; /* u or U represent unsigned int, or long int */
long int li = 65536l; /* l or L represent long int */

Without a suffix the constant has the first type that fits its value, that is a decimal constant that is larger than

GoalKicker.com – C Notes for Professionals 11

INT_MAX is of type long if possible, or long long otherwise.

The header file <limits.h> describes the limits of integers as follows. Their implementation-defined values shall be
equal or greater in magnitude (absolute value) to those shown below, with the same sign.

Macro Type Value

CHAR_BIT smallest object that is not a bit-field (byte) 8
SCHAR_MIN signed char -127 / -(27 - 1)
SCHAR_MAX signed char +127 / 27 - 1
UCHAR_MAX unsigned char 255 / 28 - 1
CHAR_MIN char see below
CHAR_MAX char see below
SHRT_MIN short int -32767 / -(215 - 1)
SHRT_MAX short int +32767 / 215 - 1
USHRT_MAX unsigned short int 65535 / 216 - 1
INT_MIN int -32767 / -(215 - 1)
INT_MAX int +32767 / 215 - 1
UINT_MAX unsigned int 65535 / 216 - 1
LONG_MIN long int -2147483647 / -(231 - 1)
LONG_MAX long int +2147483647 / 231 - 1
ULONG_MAX unsigned long int 4294967295 / 232 - 1
Version ≥ C99
Macro Type Value
LLONG_MIN long long int -9223372036854775807 / -(263 - 1)
LLONG_MAX long long int +9223372036854775807 / 263 - 1
ULLONG_MAX unsigned long long int 18446744073709551615 / 264 - 1

If the value of an object of type char sign-extends when used in an expression, the value of CHAR_MIN shall be the
same as that of SCHAR_MIN and the value of CHAR_MAX shall be the same as that of SCHAR_MAX . If the value of an
object of type char does not sign-extend when used in an expression, the value of CHAR_MIN shall be 0 and the
value of CHAR_MAX shall be the same as that of UCHAR_MAX.

Version ≥ C99

The C99 standard added a new header, <stdint.h>, which contains definitions for fixed width integers. See the
fixed width integer example for a more in-depth explanation.

Section 3.4: Floating Point Constants

The C language has three mandatory real floating point types, float, double, and long double.

float f = 0.314f; /* suffix f or F denotes type float */

double d = 0.314; /* no suffix denotes double */
long double ld = 0.314l; /* suffix l or L denotes long double */

/* the different parts of a floating point definition are optional */

double x = 1.; /* valid, fractional part is optional */
double y = .1; /* valid, whole-number part is optional */

/* they can also defined in scientific notation */

double sd = 1.2e3; /* decimal fraction 1.2 is scaled by 10^3, that is 1200.0 */

GoalKicker.com – C Notes for Professionals 12

The header <float.h> defines various limits for floating point operations.

Floating point arithmetic is implementation defined. However, most modern platforms (arm, x86, x86_64, MIPS) use
IEEE 754 floating point operations.

C also has three optional complex floating point types that are derived from the above.

Section 3.5: String Literals

A string literal in C is a sequence of chars, terminated by a literal zero.

char* str = "hello, world"; /* string literal */

/* string literals can be used to initialize arrays */

char a1[] = "abc"; /* a1 is char[4] holding {'a','b','c','\0'} */
char a2[4] = "abc"; /* same as a1 */
char a3[3] = "abc"; /* a1 is char[3] holding {'a','b','c'}, missing the '\0' */

String literals are not modifiable (and in fact may be placed in read-only memory such as .rodata). Attempting to
alter their values results in undefined behaviour.

char* s = "foobar";
s[0] = 'F'; /* undefined behaviour */

/* it's good practice to denote string literals as such, by using `const` */

char const* s1 = "foobar";
s1[0] = 'F'; /* compiler error! */

Multiple string literals are concatenated at compile time, which means you can write construct like these.

Version < C99

/* only two narrow or two wide string literals may be concatenated */
char* s = "Hello, " "World";
Version ≥ C99
/* since C99, more than two can be concatenated */
/* concatenation is implementation defined */
char* s1 = "Hello" ", " "World";

/* common usages are concatenations of format strings */

char* fmt = "%" PRId16; /* PRId16 macro since C99 */

String literals, same as character constants, support different character sets.

/* normal string literal, of type char[] */

char* s1 = "abc";

/* wide character string literal, of type wchar_t[] */

wchar_t* s2 = L"abc";
Version ≥ C11
/* UTF-8 string literal, of type char[] */
char* s3 = u8"abc";

/* 16-bit wide string literal, of type char16_t[] */

char16_t* s4 = u"abc";

/* 32-bit wide string literal, of type char32_t[] */

char32_t* s5 = U"abc";

GoalKicker.com – C Notes for Professionals 13

Chapter 4: Operators
An operator in a programming language is a symbol that tells the compiler or interpreter to perform a specific
mathematical, relational or logical operation and produce a final result.

C has many powerful operators. Many C operators are binary operators, which means they have two operands. For
example, in a / b, / is a binary operator that accepts two operands (a, b). There are some unary operators which
take one operand (for example: ~, ++), and only one ternary operator ? :.

Section 4.1: Relational Operators

Relational operators check if a specific relation between two operands is true. The result is evaluated to 1 (which
means true) or 0 (which means false). This result is often used to affect control flow (via if, while, for), but can also
be stored in variables.

Equals "=="

Checks whether the supplied operands are equal.

1 == 0; /* evaluates to 0. */
1 == 1; /* evaluates to 1. */

int x = 5;
int y = 5;
int *xptr = &x, *yptr = &y;
xptr == yptr; /* evaluates to 0, the operands hold different location addresses. */
*xptr == *yptr; /* evaluates to 1, the operands point at locations that hold the same value. */

Attention: This operator should not be confused with the assignment operator (=)!

Not equals "!="

Checks whether the supplied operands are not equal.

1 != 0; /* evaluates to 1. */
1 != 1; /* evaluates to 0. */

int x = 5;
int y = 5;
int *xptr = &x, *yptr = &y;
xptr != yptr; /* evaluates to 1, the operands hold different location addresses. */
*xptr != *yptr; /* evaluates to 0, the operands point at locations that hold the same value. */

This operator effectively returns the opposite result to that of the equals (==) operator.

Not "!"

Check whether an object is equal to 0.

The ! can also be used directly with a variable as follows:

!someVal

This has the same effect as:

GoalKicker.com – C Notes for Professionals 14

someVal == 0

Greater than ">"

Checks whether the left hand operand has a greater value than the right hand operand

5 > 4 /* evaluates to 1. */
4 > 5 /* evaluates to 0. */
4 > 4 /* evaluates to 0. */

Less than "<"

Checks whether the left hand operand has a smaller value than the right hand operand

5 < 4 /* evaluates to 0. */
4 < 5 /* evaluates to 1. */
4 < 4 /* evaluates to 0. */

Greater than or equal ">="

Checks whether the left hand operand has a greater or equal value to the right operand.

5 >= 4 /* evaluates to 1. */
4 >= 5 /* evaluates to 0. */
4 >= 4 /* evaluates to 1. */

Less than or equal "<="

Checks whether the left hand operand has a smaller or equal value to the right operand.

5 <= 4 /* evaluates to 0. */
4 <= 5 /* evaluates to 1. */
4 <= 4 /* evaluates to 1. */

Section 4.2: Conditional Operator/Ternary Operator

Evaluates its first operand, and, if the resulting value is not equal to zero, evaluates its second operand. Otherwise,
it evaluates its third operand, as shown in the following example:

a = b ? c : d;

is equivalent to:

if (b)
a = c;
else
a = d;

This pseudo-code represents it : condition ? value_if_true : value_if_false. Each value can be the result of
an evaluated expression.

int x = 5;
int y = 42;
printf("%i, %i\n", 1 ? x : y, 0 ? x : y); /* Outputs "5, 42" */

GoalKicker.com – C Notes for Professionals 15

The conditional operator can be nested. For example, the following code determines the bigger of three numbers:

big= a > b ? (a > c ? a : c)

: (b > c ? b : c);

The following example writes even integers to one file and odd integers to another file:

#include<stdio.h>

int main()
{
FILE *even, *odds;
int n = 10;
size_t k = 0;

even = fopen("even.txt", "w");

odds = fopen("odds.txt", "w");

for(k = 1; k < n + 1; k++)

{
k%2==0 ? fprintf(even, "\t%5d\n", k)
: fprintf(odds, "\t%5d\n", k);
}
fclose(even);
fclose(odds);

return 0;
}

The conditional operator associates from right to left. Consider the following:

exp1 ? exp2 : exp3 ? exp4 : exp5

As the association is from right to left, the above expression is evaluated as

exp1 ? exp2 : ( exp3 ? exp4 : exp5 )

Section 4.3: Bitwise Operators

Bitwise operators can be used to perform bit level operation on variables.
Below is a list of all six bitwise operators supported in C:

Symbol Operator
& bitwise AND
| bitwise inclusive OR
^ bitwise exclusive OR (XOR)
~ bitwise not (one's complement)
<< logical left shift
>> logical right shift

Following program illustrates the use of all bitwise operators:

#include <stdio.h>

int main(void)
{

GoalKicker.com – C Notes for Professionals 16

 unsigned int a = 29; /* 29 = 0001 1101 */
unsigned int b = 48; /* 48 = 0011 0000 */
int c = 0;

c = a & b; /* 32 = 0001 0000 */

printf("%d & %d = %d\n", a, b, c );

c = a | b; /* 61 = 0011 1101 */
printf("%d | %d = %d\n", a, b, c );

c = a ^ b; /* 45 = 0010 1101 */
printf("%d ^ %d = %d\n", a, b, c );

c = ~a; /* -30 = 1110 0010 */

printf("~%d = %d\n", a, c );

c = a << 2; /* 116 = 0111 0100 */

printf("%d << 2 = %d\n", a, c );

c = a >> 2; /* 7 = 0000 0111 */

printf("%d >> 2 = %d\n", a, c );

return 0;
}

Bitwise operations with signed types should be avoided because the sign bit of such a bit representation has a
particular meaning. Particular restrictions apply to the shift operators:

Left shifting a 1 bit into the signed bit is erroneous and leads to undefined behavior.

Right shifting a negative value (with sign bit 1) is implementation defined and therefore not portable.

If the value of the right operand of a shift operator is negative or is greater than or equal to the width of the
promoted left operand, the behavior is undefined.

Masking:

Masking refers to the process of extracting the desired bits from (or transforming the desired bits in) a variable by
using logical bitwise operations. The operand (a constant or variable) that is used to perform masking is called a
mask.

Masking is used in many different ways:

To decide the bit pattern of an integer variable.

To copy a portion of a given bit pattern to a new variable, while the remainder of the new variable is filled
with 0s (using bitwise AND)
To copy a portion of a given bit pattern to a new variable, while the remainder of the new variable is filled
with 1s (using bitwise OR).
To copy a portion of a given bit pattern to a new variable, while the remainder of the original bit pattern is
inverted within the new variable (using bitwise exclusive OR).

The following function uses a mask to display the bit pattern of a variable:

#include <limits.h>
void bit_pattern(int u)
{
int i, x, word;
unsigned mask = 1;

GoalKicker.com – C Notes for Professionals 17

 word = CHAR_BIT * sizeof(int);
mask = mask << (word - 1); /* shift 1 to the leftmost position */
for(i = 1; i <= word; i++)
{
x = (u & mask) ? 1 : 0; /* identify the bit */
printf("%d", x); /* print bit value */
mask >>= 1; /* shift mask to the right by 1 bit */
}
}

Section 4.4: Short circuit behavior of logical operators

Short circuiting is a functionality that skips evaluating parts of a (if/while/...) condition when able. In case of a logical
operation on two operands, the first operand is evaluated (to true or false) and if there is a verdict (i.e first operand
is false when using &&, first operand is true when using ||) the second operand is not evaluated.

Example:

#include <stdio.h>

int main(void) {
int a = 20;
int b = -5;

/* here 'b == -5' is not evaluated,

since a 'a != 20' is false. */
if (a != 20 && b == -5) {
printf("I won't be printed!\n");
}

return 0;
}

Check it out yourself:

#include <stdio.h>

int print(int i) {
printf("print function %d\n", i);
return i;
}

int main(void) {
int a = 20;

/* here 'print(a)' is not called,

since a 'a != 20' is false. */
if (a != 20 && print(a)) {
printf("I won't be printed!\n");
}

/* here 'print(a)' is called,

since a 'a == 20' is true. */
if (a == 20 && print(a)) {
printf("I will be printed!\n");
}

return 0;
}

GoalKicker.com – C Notes for Professionals 18

Output:

$ ./a.out
print function 20
I will be printed!

Short circuiting is important, when you want to avoid evaluating terms that are (computationally) costly. Moreover,
it can heavily affect the flow of your program like in this case: Why does this program print "forked!" 4 times?

Section 4.5: Comma Operator

Evaluates its left operand, discards the resulting value, and then evaluates its rights operand and result yields the
value of its rightmost operand.

int x = 42, y = 42;

printf("%i\n", (x *= 2, y)); /* Outputs "42". */

The comma operator introduces a sequence point between its operands.

Note that the comma used in functions calls that separate arguments is NOT the comma operator, rather it's called a
separator which is different from the comma operator. Hence, it doesn't have the properties of the comma operator.

The above printf() call contains both the comma operator and the separator.

printf("%i\n", (x *= 2, y)); /* Outputs "42". */

/* ^ ^ this is a comma operator */
/* this is a separator */

The comma operator is often used in the initialization section as well as in the updating section of a for loop. For
example:

for(k = 1; k < 10; printf("\%d\\n", k), k += 2); /*outputs the odd numbers below 9/*

/* outputs sum to first 9 natural numbers */

for(sumk = 1, k = 1; k < 10; k++, sumk += k)
printf("\%5d\%5d\\n", k, sumk);

Section 4.6: Arithmetic Operators

Basic Arithmetic

Return a value that is the result of applying the left hand operand to the right hand operand, using the associated
mathematical operation. Normal mathematical rules of commutation apply (i.e. addition and multiplication are
commutative, subtraction, division and modulus are not).

Addition Operator

The addition operator (+) is used to add two operands together. Example:

#include <stdio.h>

int main(void)
{
int a = 5;
int b = 7;

GoalKicker.com – C Notes for Professionals 19

 int c = a + b; /* c now holds the value 12 */

printf("%d + %d = %d",a,b,c); /* will output "5 + 7 = 12" */

return 0;
}

Subtraction Operator

The subtraction operator (-) is used to subtract the second operand from the first. Example:

#include <stdio.h>

int main(void)
{
int a = 10;
int b = 7;

int c = a - b; /* c now holds the value 3 */

printf("%d - %d = %d",a,b,c); /* will output "10 - 7 = 3" */

return 0;
}

Multiplication Operator

The multiplication operator (*) is used to multiply both operands. Example:

#include <stdio.h>

int main(void)
{
int a = 5;
int b = 7;

int c = a * b; /* c now holds the value 35 */

printf("%d * %d = %d",a,b,c); /* will output "5 * 7 = 35" */

return 0;
}

Not to be confused with the * dereference operator.

Division Operator

The division operator (/) divides the first operand by the second. If both operands of the division are integers, it will
return an integer value and discard the remainder (use the modulo operator % for calculating and acquiring the
remainder).

If one of the operands is a floating point value, the result is an approximation of the fraction.

Example:

#include <stdio.h>

int main (void)

GoalKicker.com – C Notes for Professionals 20

{
int a = 19 / 2 ; /* a holds value 9 */
int b = 18 / 2 ; /* b holds value 9 */
int c = 255 / 2; /* c holds value 127 */
int d = 44 / 4 ; /* d holds value 11 */
double e = 19 / 2.0 ; /* e holds value 9.5 */
double f = 18.0 / 2 ; /* f holds value 9.0 */
double g = 255 / 2.0; /* g holds value 127.5 */
double h = 45.0 / 4 ; /* h holds value 11.25 */

printf("19 / 2 = %d\n", a); /* Will output "19 / 2 = 9" */

printf("18 / 2 = %d\n", b); /* Will output "18 / 2 = 9" */
printf("255 / 2 = %d\n", c); /* Will output "255 / 2 = 127" */
printf("44 / 4 = %d\n", d); /* Will output "44 / 4 = 11" */
printf("19 / 2.0 = %g\n", e); /* Will output "19 / 2.0 = 9.5" */
printf("18.0 / 2 = %g\n", f); /* Will output "18.0 / 2 = 9" */
printf("255 / 2.0 = %g\n", g); /* Will output "255 / 2.0 = 127.5" */
printf("45.0 / 4 = %g\n", h); /* Will output "45.0 / 4 = 11.25" */

return 0;
}

Modulo Operator

The modulo operator (%) receives integer operands only, and is used to calculate the remainder after the first
operand is divided by the second. Example:

#include <stdio.h>

int main (void) {

int a = 25 % 2; /* a holds value 1 */
int b = 24 % 2; /* b holds value 0 */
int c = 155 % 5; /* c holds value 0 */
int d = 49 % 25; /* d holds value 24 */

printf("25 % 2 = %d\n", a); /* Will output "25 % 2 = 1" */

printf("24 % 2 = %d\n", b); /* Will output "24 % 2 = 0" */
printf("155 % 5 = %d\n", c); /* Will output "155 % 5 = 0" */
printf("49 % 25 = %d\n", d); /* Will output "49 % 25 = 24" */

return 0;
}

Increment / Decrement Operators

The increment (a++) and decrement (

a--

) operators are different in that they change the value of the variable you apply them to without an assignment
operator. You can use increment and decrement operators either before or after the variable. The placement of the
operator changes the timing of the incrementation/decrementation of the value to before or after assigning it to
the variable. Example:
#include <stdio.h>

int main(void)
{
int a = 1;
int b = 4;

GoalKicker.com – C Notes for Professionals 21

 int c = 1;
int d = 4;

a++;
printf("a = %d\n",a); /* Will output "a = 2" */
b--;
printf("b = %d\n",b); /* Will output "b = 3" */

if (++c > 1) { /* c is incremented by 1 before being compared in the condition */

printf("This will print\n"); /* This is printed */
} else {
printf("This will never print\n"); /* This is not printed */
}

if (d-- < 4) { /* d is decremented after being compared */

printf("This will never print\n"); /* This is not printed */
} else {
printf("This will print\n"); /* This is printed */
}
}

As the example for c and d shows, both operators have two forms, as prefix notation and postfix notation. Both
have the same effect in incrementing (++) or decrementing (--) the variable, but differ by the value they return:
prefix operations do the operation first and then return the value, whereas postfix operations first determine the
value that is to be returned, and then do the operation.

Because of this potentially counter-intuitive behaviour, the use of increment/decrement operators inside
expressions is controversial.

Section 4.7: Access Operators

The member access operators (dot . and arrow ->) are used to access a member of a struct.

Member of object

Evaluates into the lvalue denoting the object that is a member of the accessed object.

struct MyStruct
{
int x;
int y;
};

struct MyStruct myObject;

myObject.x = 42;
myObject.y = 123;

printf(".x = %i, .y = %i\n", myObject.x, myObject.y); /* Outputs ".x = 42, .y = 123". */

Member of pointed-to object

Syntactic sugar for dereferencing followed by member access. Effectively, an expression of the form x->y is
shorthand for (*x).y — but the arrow operator is much clearer, especially if the structure pointers are nested.

struct MyStruct
{
int x;
int y;

GoalKicker.com – C Notes for Professionals 22

};

struct MyStruct myObject;

struct MyStruct *p = &myObject;

p->x = 42;
p->y = 123;

printf(".x = %i, .y = %i\n", p->x, p->y); /* Outputs ".x = 42, .y = 123". */

printf(".x = %i, .y = %i\n", myObject.x, myObject.y); /* Also outputs ".x = 42, .y = 123". */

Address-of

The unary & operator is the address of operator. It evaluates the given expression, where the resulting object must
be an lvalue. Then, it evaluates into an object whose type is a pointer to the resulting object's type, and contains the
address of the resulting object.

int x = 3;
int *p = &x;
printf("%p = %p\n", （void *)&x, (void *)p); /* Outputs "A = A", for some implementation-defined A.
*/

Dereference

The unary * operator dereferences a pointer. It evaluates into the lvalue resulting from dereferencing the pointer
that results from evaluating the given expression.

int x = 42;
int *p = &x;
printf("x = %d, *p = %d\n", x, *p); /* Outputs "x = 42, *p = 42". */

*p = 123;
printf("x = %d, *p = %d\n", x, *p); /* Outputs "x = 123, *p = 123". */

Indexing

Indexing is syntactic sugar for pointer addition followed by dereferencing. Effectively, an expression of the form
a[i] is equivalent to *(a + i) — but the explicit subscript notation is preferred.

int arr[] = { 1, 2, 3, 4, 5 };
printf("arr[2] = %i\n", arr[2]); /* Outputs "arr[2] = 3". */

Interchangeability of indexing

Adding a pointer to an integer is a commutative operation (i.e. the order of the operands does not change the
result) so pointer + integer == integer + pointer.

A consequence of this is that arr[3] and 3[arr] are equivalent.

printf("3[arr] = %i\n", 3[arr]); /* Outputs "3[arr] = 4". */

Usage of an expression 3[arr] instead of arr[3] is generally not recommended, as it affects code readability. It
tends to be a popular in obfuscated programming contests.

GoalKicker.com – C Notes for Professionals 23

Section 4.8: sizeof Operator
With a type as operand

Evaluates into the size in bytes, of type size_t, of objects of the given type. Requires parentheses around the type.

printf("%zu\n", sizeof(int)); /* Valid, outputs the size of an int object, which is platform-
dependent. */
printf("%zu\n", sizeof int); /* Invalid, types as arguments need to be surrounded by parentheses! */

With an expression as operand

Evaluates into the size in bytes, of type size_t, of objects of the type of the given expression. The expression itself
is not evaluated. Parentheses are not required; however, because the given expression must be unary, it's
considered best practice to always use them.

char ch = 'a';
printf("%zu\n", sizeof(ch)); /* Valid, will output the size of a char object, which is always 1 for
all platforms. */
printf("%zu\n", sizeof ch); /* Valid, will output the size of a char object, which is always 1 for
all platforms. */

Section 4.9: Cast Operator

Performs an explicit conversion into the given type from the value resulting from evaluating the given expression.

int x = 3;
int y = 4;
printf("%f\n", (double)x / y); /* Outputs "0.750000". */

Here the value of x is converted to a double, the division promotes the value of y to double, too, and the result of
the division, a double is passed to printf for printing.

Section 4.10: Function Call Operator

The first operand must be a function pointer (a function designator is also acceptable because it will be converted
to a pointer to the function), identifying the function to call, and all other operands, if any, are collectively known as
the function call's arguments. Evaluates into the return value resulting from calling the appropriate function with
the respective arguments.

int myFunction(int x, int y)

{
return x * 2 + y;
}

int (*fn)(int, int) = &myFunction;

int x = 42;
int y = 123;

printf("(*fn)(%i, %i) = %i\n", x, y, (*fn)(x, y)); /* Outputs "fn(42, 123) = 207". */

printf("fn(%i, %i) = %i\n", x, y, fn(x, y)); /* Another form: you don't need to dereference
explicitly */

GoalKicker.com – C Notes for Professionals 24

Section 4.11: Increment / Decrement
The increment and decrement operators exist in prefix and postfix form.

int a = 1;
int b = 1;
int tmp = 0;

tmp = ++a; /* increments a by one, and returns new value; a == 2, tmp == 2 */

tmp = a++; /* increments a by one, but returns old value; a == 3, tmp == 2 */
tmp = --b; /* decrements b by one, and returns new value; b == 0, tmp == 0 */
tmp = b--; /* decrements b by one, but returns old value; b == -1, tmp == 0 */

Note that arithmetic operations do not introduce sequence points, so certain expressions with ++ or -- operators
may introduce undefined behaviour.

Section 4.12: Assignment Operators

Assigns the value of the right-hand operand to the storage location named by the left-hand operand, and returns
the value.

int x = 5; /* Variable x holds the value 5. Returns 5. */

char y = 'c'; /* Variable y holds the value 99. Returns 99
* (as the character 'c' is represented in the ASCII table with 99).
*/
float z = 1.5; /* variable z holds the value 1.5. Returns 1.5. */
char const* s = "foo"; /* Variable s holds the address of the first character of the string 'foo'.
*/

Several arithmetical operations have a compound assignment operator.

a += b /* equal to: a = a + b */
a -= b /* equal to: a = a - b */
a *= b /* equal to: a = a * b */
a /= b /* equal to: a = a / b */
a %= b /* equal to: a = a % b */
a &= b /* equal to: a = a & b */
a |= b /* equal to: a = a | b */
a ^= b /* equal to: a = a ^ b */
a <<= b /* equal to: a = a << b */
a >>= b /* equal to: a = a >> b */

One important feature of these compound assignments is that the expression on the left hand side (a) is only
evaluated once. E.g if p is a pointer

*p += 27;

dereferences p only once, whereas the following does so twice.

*p = *p + 27;

It should also be noted that the result of an assignment such as a = b is what is known as an rvalue. Thus, the
assignment actually has a value which can then be assigned to another variable. This allows the chaining of
assignments to set multiple variables in a single statement.

This rvalue can be used in the controlling expressions of if statements (or loops or switch statements) that guard

GoalKicker.com – C Notes for Professionals 25

some code on the result of another expression or function call. For example:

char *buffer;
if ((buffer = malloc(1024)) != NULL)
{
/* do something with buffer */
free(buffer);
}
else
{
/* report allocation failure */
}

Because of this, care must be taken to avoid a common typo which can lead to mysterious bugs.

int a = 2;
/* ... */
if (a = 1)
/* Delete all files on my hard drive */

This will have disastrous results, as a = 1 will always evaluate to 1 and thus the controlling expression of the if
statement will always be true (read more about this common pitfall here). The author almost certainly meant to use
the equality operator (==) as shown below:

int a = 2;
/* ... */
if (a == 1)
/* Delete all files on my hard drive */

Operator Associativity

int a, b = 1, c = 2;
a = b = c;

This assigns c to b, which returns b, which is than assigned to a. This happens because all assignment-operators
have right associativity, that means the rightmost operation in the expression is evaluated first, and proceeds from
right to left.

Section 4.13: Logical Operators

Logical AND

Performs a logical boolean AND-ing of the two operands returning 1 if both of the operands are non-zero. The
logical AND operator is of type int.

0 && 0 /* Returns 0. */
0 && 1 /* Returns 0. */
2 && 0 /* Returns 0. */
2 && 3 /* Returns 1. */

Logical OR

Performs a logical boolean OR-ing of the two operands returning 1 if any of the operands are non-zero. The logical
OR operator is of type int.

0 || 0 /* Returns 0. */

GoalKicker.com – C Notes for Professionals 26

0 || 1 /* Returns 1. */
2 || 0 /* Returns 1. */
2 || 3 /* Returns 1. */

Logical NOT

Performs a logical negation. The logical NOT operator is of type int. The NOT operator checks if at least one bit is
equal to 1, if so it returns 0. Else it returns 1;

!1 /* Returns 0. */
!5 /* Returns 0. */
!0 /* Returns 1. */

Short-Circuit Evaluation

There are some crucial properties common to both && and ||:

the left-hand operand (LHS) is fully evaluated before the right-hand operand (RHS) is evaluated at all,
there is a sequence point between the evaluation of the left-hand operand and the right-hand operand,
and, most importantly, the right-hand operand is not evaluated at all if the result of the left-hand operand
determines the overall result.

This means that:

if the LHS evaluates to 'true' (non-zero), the RHS of || will not be evaluated (because the result of 'true OR
anything' is 'true'),
if the LHS evaluates to 'false' (zero), the RHS of && will not be evaluated (because the result of 'false AND
anything' is 'false').

This is important as it permits you to write code such as:

const char *name_for_value(int value)

{
static const char *names[] = { "zero", "one", "two", "three", };
enum { NUM_NAMES = sizeof(names) / sizeof(names[0]) };
return (value >= 0 && value < NUM_NAMES) ? names[value] : "infinity";
}

If a negative value is passed to the function, the value >= 0 term evaluates to false and the value < NUM_NAMES
term is not evaluated.

Section 4.14: Pointer Arithmetic

Pointer addition

Given a pointer and a scalar type N, evaluates into a pointer to the Nth element of the pointed-to type that directly
succeeds the pointed-to object in memory.

int arr[] = {1, 2, 3, 4, 5};

printf("*(arr + 3) = %i\n", *(arr + 3)); /* Outputs "4", arr's fourth element. */

It does not matter if the pointer is used as the operand value or the scalar value. This means that things such as 3 +
arr are valid. If arr[k] is the k+1 member of an array, then arr+k is a pointer to arr[k]. In other words, arr or
arr+0 is a pointer to arr[0], arr+1 is a pointer to arr[2], and so on. In general, *(arr+k) is same as arr[k].

GoalKicker.com – C Notes for Professionals 27

Unlike the usual arithmetic, addition of 1 to a pointer to an int will add 4 bytes to the current address value. As
array names are constant pointers, + is the only operator we can use to access the members of an array via pointer
notation using the array name. However, by defining a pointer to an array, we can get more flexibility to process the
data in an array. For example, we can print the members of an array as follows:

#include<stdio.h>
static const size_t N = 5

int main()
{
size_t k = 0;
int arr[] = {1, 2, 3, 4, 5};
for(k = 0; k < N; k++)
{
printf("\n\t%d", *(arr + k));
}
return 0;
}

By defining a pointer to the array, the above program is equivalent to the following:

#include<stdio.h>
static const size_t N = 5

int main()
{
size_t k = 0;
int arr[] = {1, 2, 3, 4, 5};
int *ptr = arr; /* or int *ptr = &arr[0]; */
for(k = 0; k < N; k++)
{
printf("\n\t%d", ptr[k]);
/* or printf("\n\t%d", *(ptr + k)); */
/* or printf("\n\t%d", *ptr++); */
}
return 0;
}

See that the members of the array arr are accessed using the operators + and ++. The other operators that can be
used with the pointer ptr are - and --.

Pointer subtraction

Given two pointers to the same type, evaluates into an object of type ptrdiff_t that holds the scalar value that
must be added to the second pointer in order to obtain the value of the first pointer.

int arr[] = {1, 2, 3, 4, 5};

int *p = &arr[2];
int *q = &arr[3];
ptrdiff_t diff = q - p;

printf("q - p = %ti\n", diff); /* Outputs "1". */

printf("*(p + (q - p)) = %d\n", *(p + diff)); /* Outputs "4". */

Section 4.15: _Alignof

Version ≥ C11

GoalKicker.com – C Notes for Professionals 28

Queries the alignment requirement for the specified type. The alignment requirement is a positive integral power
of 2 representing the number of bytes between which two objects of the type may be allocated. In C, the alignment
requirement is measured in size_t.

The type name may not be an incomplete type nor a function type. If an array is used as the type, the type of the
array element is used.

This operator is often accessed through the convenience macro alignof from <stdalign.h>.

int main(void)
{
printf("Alignment of char = %zu\n", alignof(char));
printf("Alignment of max_align_t = %zu\n", alignof(max_align_t));
printf("alignof(float[10]) = %zu\n", alignof(float[10]));
printf("alignof(struct{char c; int n;}) = %zu\n",
alignof(struct {char c; int n;}));
}

Possible Output:

Alignment of char = 1
Alignment of max_align_t = 16
alignof(float[10]) = 4
alignof(struct{char c; int n;}) = 4

http://en.cppreference.com/w/c/language/_Alignof

GoalKicker.com – C Notes for Professionals 29

Chapter 5: Boolean
Section 5.1: Using stdbool.h
Version ≥ C99

Using the system header file stdbool.h allows you to use bool as a Boolean data type. true evaluates to 1 and
false evaluates to 0.

#include <stdio.h>
#include <stdbool.h>

int main(void) {
bool x = true; /* equivalent to bool x = 1; */
bool y = false; /* equivalent to bool y = 0; */
if (x) /* Functionally equivalent to if (x != 0) or if (x != false) */
{
puts("This will print!");
}
if (!y) /* Functionally equivalent to if (y == 0) or if (y == false) */
{
puts("This will also print!");
}
}

bool is just a nice spelling for the data type _Bool. It has special rules when numbers or pointers are converted to it.

Section 5.2: Using #define

C of all versions, will effectively treat any integer value other than 0 as true for comparison operators and the
integer value 0 as false. If you don't have _Bool or bool as of C99 available, you could simulate a Boolean data type
in C using #define macros, and you might still find such things in legacy code.

#include <stdio.h>

#define bool int

#define true 1
#define false 0

int main(void) {
bool x = true; /* Equivalent to int x = 1; */
bool y = false; /* Equivalent to int y = 0; */
if (x) /* Functionally equivalent to if (x != 0) or if (x != false) */
{
puts("This will print!");
}
if (!y) /* Functionally equivalent to if (y == 0) or if (y == false) */
{
puts("This will also print!");
}
}

Don't introduce this in new code since the definition of these macros might clash with modern uses of
<stdbool.h>.

GoalKicker.com – C Notes for Professionals 30

Section 5.3: Using the Intrinsic (built-in) Type _Bool
Version ≥ C99

Added in the C standard version C99, _Bool is also a native C data type. It is capable of holding the values 0 (for
false) and 1 (for true).

#include <stdio.h>

int main(void) {
_Bool x = 1;
_Bool y = 0;
if(x) /* Equivalent to if (x == 1) */
{
puts("This will print!");
}
if (!y) /* Equivalent to if (y == 0) */
{
puts("This will also print!");
}
}

_Bool is an integer type but has special rules for conversions from other types. The result is analogous to the usage
of other types in if expressions. In the following

_Bool z = X;

If X has an arithmetic type (is any kind of number), z becomes 0 if X == 0. Otherwise z becomes 1.
If X has a pointer type, z becomes 0 if X is a null pointer and 1 otherwise.

To use nicer spellings bool, false and true you need to use <stdbool.h>.

Section 5.4: Integers and pointers in Boolean expressions

All integers or pointers can be used in an expression that is interpreted as "truth value".

int main(int argc, char* argv[]) {

if (argc % 4) {
puts("arguments number is not divisible by 4");
} else {
puts("argument number is divisible by 4");
}
...

The expression argc % 4 is evaluated and leads to one of the values 0, 1, 2 or 3. The first, 0 is the only value that is
"false" and brings execution into the else part. All other values are "true" and go into the if part.

double* A = malloc(n*sizeof *A);

if (!A) {
perror("allocation problems");
exit(EXIT_FAILURE);
}

Here the pointer A is evaluated and if it is a null pointer, an error is detected and the program exits.

Many people prefer to write something as A == NULL, instead, but if you have such pointer comparisons as part of

GoalKicker.com – C Notes for Professionals 31

other complicated expressions, things become quickly difficult to read.

char const* s = ....; /* some pointer that we receive */

if (s != NULL && s[0] != '\0' && isalpha(s[0])) {
printf("this starts well, %c is alphabetic\n", s[0]);
}

For this to check, you'd have to scan a complicated code in the expression and be sure about operator preference.

char const* s = ....; /* some pointer that we receive */

if (s && s[0] && isalpha(s[0])) {
printf("this starts well, %c is alphabetic\n", s[0]);
}

is relatively easy to capture: if the pointer is valid we check if the first character is non-zero and then check if it is a
letter.

Section 5.5: Defining a bool type using typedef

Considering that most debuggers are not aware of #define macros, but can check enum constants, it may be
desirable to do something like this:

#if __STDC_VERSION__ < 199900L

typedef enum { false, true } bool;
/* Modern C code might expect these to be macros. */
# ifndef bool
# define bool bool
# endif
# ifndef true
# define true true
# endif
# ifndef false
# define false false
# endif
#else
# include <stdbool.h>
#endif

/* Somewhere later in the code ... */

bool b = true;

This allows compilers for historic versions of C to function, but remains forward compatible if the code is compiled
with a modern C compiler.

For more information on typedef, see Typedef, for more on enum see Enumerations

GoalKicker.com – C Notes for Professionals 32

Chapter 6: Strings
In C, a string is not an intrinsic type. A C-string is the convention to have a one-dimensional array of characters
which is terminated by a null-character, by a '\0'.

This means that a C-string with a content of "abc" will have four characters 'a', 'b', 'c' and '\0'.

See the basic introduction to strings example.

Section 6.1: Tokenisation: strtok(), strtok_r() and strtok_s()

The function strtok breaks a string into a smaller strings, or tokens, using a set of delimiters.

#include <stdio.h>
#include <string.h>

int main(void)
{
int toknum = 0;
char src[] = "Hello,, world!";
const char delimiters[] = ", !";
char *token = strtok(src, delimiters);
while (token != NULL)
{
printf("%d: [%s]\n", ++toknum, token);
token = strtok(NULL, delimiters);
}
/* source is now "Hello\0, world\0\0" */
}

Output:

1: [Hello]
2: [world]

The string of delimiters may contain one or more delimiters and different delimiter strings may be used with each
call to strtok.

Calls to strtok to continue tokenizing the same source string should not pass the source string again, but instead
pass NULL as the first argument. If the same source string is passed then the first token will instead be re-tokenized.
That is, given the same delimiters, strtok would simply return the first token again.

Note that as strtok does not allocate new memory for the tokens, it modifies the source string. That is, in the above
example, the string src will be manipulated to produce the tokens that are referenced by the pointer returned by
the calls to strtok. This means that the source string cannot be const (so it can't be a string literal). It also means
that the identity of the delimiting byte is lost (i.e. in the example the "," and "!" are effectively deleted from the
source string and you cannot tell which delimiter character matched).

Note also that multiple consecutive delimiters in the source string are treated as one; in the example, the second
comma is ignored.

strtok is neither thread safe nor re-entrant because it uses a static buffer while parsing. This means that if a
function calls strtok, no function that it calls while it is using strtok can also use strtok, and it cannot be called by
any function that is itself using strtok.

GoalKicker.com – C Notes for Professionals 33

An example that demonstrates the problems caused by the fact that strtokis not re-entrant is as follows:

char src[] = "1.2,3.5,4.2";

char *first = strtok(src, ",");

do
{
char *part;
/* Nested calls to strtok do not work as desired */
printf("[%s]\n", first);
part = strtok(first, ".");
while (part != NULL)
{
printf(" [%s]\n", part);
part = strtok(NULL, ".");
}
} while ((first = strtok(NULL, ",")) != NULL);

Output:

[1.2]
[1]
[2]

The expected operation is that the outer do while loop should create three tokens consisting of each decimal
number string ("1.2", "3.5", "4.2"), for each of which the strtok calls for the inner loop should split it into
separate digit strings ("1", "2", "3", "5", "4", "2").

However, because strtok is not re-entrant, this does not occur. Instead the first strtok correctly creates the "1.2\0"
token, and the inner loop correctly creates the tokens "1" and "2". But then the strtok in the outer loop is at the
end of the string used by the inner loop, and returns NULL immediately. The second and third substrings of the src
array are not analyzed at all.

Version < C11

The standard C libraries do not contain a thread-safe or re-entrant version but some others do, such as POSIX'
strtok_r. Note that on MSVC the strtok equivalent, strtok_s is thread-safe.

Version ≥ C11

C11 has an optional part, Annex K, that offers a thread-safe and re-entrant version named strtok_s. You can test
for the feature with __STDC_LIB_EXT1__. This optional part is not widely supported.

The strtok_s function differs from the POSIX strtok_r function by guarding against storing outside of the string
being tokenized, and by checking runtime constraints. On correctly written programs, though, the strtok_s and
strtok_r behave the same.

Using strtok_s with the example now yields the correct response, like so:

/* you have to announce that you want to use Annex K */

#define __STDC_WANT_LIB_EXT1__ 1
#include <string.h>

#ifndef __STDC_LIB_EXT1__
# error "we need strtok_s from Annex K"
#endif

GoalKicker.com – C Notes for Professionals 34

char src[] = "1.2,3.5,4.2";
char *next = NULL;
char *first = strtok_s(src, ",", &next);

do
{
char *part;
char *posn;

printf("[%s]\n", first);
part = strtok_s(first, ".", &posn);
while (part != NULL)
{
printf(" [%s]\n", part);
part = strtok_s(NULL, ".", &posn);
}
}
while ((first = strtok_s(NULL, ",", &next)) != NULL);

And the output will be:

[1.2]
[1]
[2]
[3.5]
[3]
[5]
[4.2]
[4]
[2]

Section 6.2: String literals

String literals represent null-terminated, static-duration arrays of char. Because they have static storage duration, a
string literal or a pointer to the same underlying array can safely be used in several ways that a pointer to an
automatic array cannot. For example, returning a string literal from a function has well-defined behavior:

const char *get_hello() {

return "Hello, World!"; /* safe */
}

For historical reasons, the elements of the array corresponding to a string literal are not formally const.
Nevertheless, any attempt to modify them has undefined behavior. Typically, a program that attempts to modify
the array corresponding to a string literal will crash or otherwise malfunction.

char *foo = "hello";

foo[0] = 'y'; /* Undefined behavior - BAD! */

Where a pointer points to a string literal -- or where it sometimes may do -- it is advisable to declare that pointer's
referent const to avoid engaging such undefined behavior accidentally.

const char *foo = "hello";

/* GOOD: can't modify the string pointed to by foo */

On the other hand, a pointer to or into the underlying array of a string literal is not itself inherently special; its value
can freely be modified to point to something else:

GoalKicker.com – C Notes for Professionals 35

char *foo = "hello";
foo = "World!"; /* OK - we're just changing what foo points to */

Furthermore, although initializers for char arrays can have the same form as string literals, use of such an initializer
does not confer the characteristics of a string literal on the initialized array. The initializer simply designates the
length and initial contents of the array. In particular, the elements are modifiable if not explicitly declared const:

char foo[] = "hello";

foo[0] = 'y'; /* OK! */

Section 6.3: Calculate the Length: strlen()

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

int main(int argc, char **argv)

{
/* Exit if no second argument is found. */
if (argc != 2)
{
puts("Argument missing.");
return EXIT_FAILURE;
}

size_t len = strlen(argv[1]);

printf("The length of the second argument is %zu.\\n", len);

return EXIT_SUCCESS;
}

This program computes the length of its second input argument and stores the result in len. It then prints that
length to the terminal. For example, when run with the parameters program_name "Hello, world!", the program
will output The length of the second argument is 13. because the string Hello, world! is 13 characters long.

strlen counts all the bytes from the beginning of the string up to, but not including, the terminating NUL
character, '\\0'. As such, it can only be used when the string is guaranteed to be NUL-terminated.

Also keep in mind that if the string contains any Unicode characters, strlen will not tell you how many characters
are in the string (since some characters may be multiple bytes long). In such cases, you need to count the
characters (i.e., code units) yourself. Consider the output of the following example:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

int main(void)
{
char asciiString[50] = "Hello world!";
char utf8String[50] = "Γειά σου Κόσμε!"; /* "Hello World!" in Greek */

printf("asciiString has %zu bytes in the array\\n", sizeof(asciiString));

printf("utf8String has %zu bytes in the array\\n", sizeof(utf8String));
printf("\\"%s\\" is %zu bytes\\n", asciiString, strlen(asciiString));
printf("\\"%s\\" is %zu bytes\\n", utf8String, strlen(utf8String));
}

GoalKicker.com – C Notes for Professionals 36

Output:

asciiString has 50 bytes in the array

utf8String has 50 bytes in the array
"Hello world!" is 12 bytes
"Γειά σου Κόσμε!" is 27 bytes

Section 6.4: Basic introduction to strings

In C, a string is a sequence of characters that is terminated by a null character ('\0').

We can create strings using string literals, which are sequences of characters surrounded by double quotation
marks; for example, take the string literal "hello world". String literals are automatically null-terminated.

We can create strings using several methods. For instance, we can declare a char * and initialize it to point to the
first character of a string:

char * string = "hello world";

When initializing a char * to a string constant as above, the string itself is usually allocated in read-only data;
string is a pointer to the first element of the array, which is the character 'h'.

Since the string literal is allocated in read-only memory, it is non-modifiable1. Any attempt to modify it will lead to
undefined behaviour, so it's better to add const to get a compile-time error like this

char const * string = "hello world";

It has similar effect2 as

char const string_arr[] = "hello world";

To create a modifiable string, you can declare a character array and initialize its contents using a string literal, like
so:

char modifiable_string[] = "hello world";

This is equivalent to the following:

char modifiable_string[] = {'h', 'e', 'l', 'l', 'o', ' ', 'w', 'o', 'r', 'l', 'd', '\0'};

Since the second version uses brace-enclosed initializer, the string is not automatically null-terminated unless a
'\0' character is included explicitly in the character array usually as its last element.

1 Non-modifiable implies that the characters in the string literal can't be modified, but remember that the pointer
string can be modified (can point somewhere else or can be incremented or decremented).

2 Both strings have similar effect in a sense that characters of both strings can't be modified. It should be noted
that string is a pointer to char and it is a modifiable l-value so it can be incremented or point to some other
location while the array string_arr is a non-modifiable l-value, it can't be modified.

Section 6.5: Copying strings

Pointer assignments do not copy strings

GoalKicker.com – C Notes for Professionals 37

You can use the = operator to copy integers, but you cannot use the = operator to copy strings in C. Strings in C are
represented as arrays of characters with a terminating null-character, so using the = operator will only save the
address (pointer) of a string.

#include <stdio.h>

int main(void) {
int a = 10, b;
char c[] = "abc", *d;

b = a; /* Integer is copied */
a = 20; /* Modifying a leaves b unchanged - b is a 'deep copy' of a */
printf("%d %d\n", a, b); /* "20 10" will be printed */

d = c;
/* Only copies the address of the string -
there is still only one string stored in memory */

c[1] = 'x';
/* Modifies the original string - d[1] = 'x' will do exactly the same thing */

printf("%s %s\n", c, d); /* "axc axc" will be printed */

return 0;
}

The above example compiled because we used char *d rather than char d[3]. Using the latter would cause a
compiler error. You cannot assign to arrays in C.

#include <stdio.h>

int main(void) {
char a[] = "abc";
char b[8];

b = a; /* compile error */
printf("%s\n", b);

return 0;
}

Copying strings using standard functions

strcpy()

To actually copy strings, strcpy() function is available in string.h. Enough space must be allocated for the
destination before copying.

#include <stdio.h>
#include <string.h>

int main(void) {
char a[] = "abc";
char b[8];

strcpy(b, a); /* think "b special equals a" */

printf("%s\n", b); /* "abc" will be printed */

return 0;
}
Version ≥ C99

GoalKicker.com – C Notes for Professionals 38

snprintf()

To avoid buffer overrun, snprintf() may be used. It is not the best solution performance-wise since it has to parse
the template string, but it is the only buffer limit-safe function for copying strings readily-available in standard
library, that can be used without any extra steps.

#include <stdio.h>
#include <string.h>

int main(void) {
char a[] = "012345678901234567890";
char b[8];

#if 0
strcpy(b, a); /* causes buffer overrun (undefined behavior), so do not execute this here! */
#endif

snprintf(b, sizeof(b), "%s", a); /* does not cause buffer overrun */

printf("%s\n", b); /* "0123456" will be printed */

return 0;
}
strncat()

A second option, with better performance, is to use strncat() (a buffer overflow checking version of strcat()) - it
takes a third argument that tells it the maximum number of bytes to copy:

char dest[32];

dest[0] = '\0';
strncat(dest, source, sizeof(dest) - 1);
/* copies up to the first (sizeof(dest) - 1) elements of source into dest,
then puts a \0 on the end of dest */

Note that this formulation use sizeof(dest) - 1; this is crucial because strncat() always adds a null byte (good),
but doesn't count that in the size of the string (a cause of confusion and buffer overwrites).

Also note that the alternative — concatenating after a non-empty string — is even more fraught. Consider:

char dst[24] = "Clownfish: ";

char src[] = "Marvin and Nemo";
size_t len = strlen(dst);

strncat(dst, src, sizeof(dst) - len - 1);

printf("%zu: [%s]\n", strlen(dst), dst);

The output is:

23: [Clownfish: Marvin and N]

Note, though, that the size specified as the length was not the size of the destination array, but the amount of space
left in it, not counting the terminal null byte. This can cause big overwriting problems. It is also a bit wasteful; to
specify the length argument correctly, you know the length of the data in the destination, so you could instead
specify the address of the null byte at the end of the existing content, saving strncat() from rescanning it:

strcpy(dst, "Clownfish: ");

GoalKicker.com – C Notes for Professionals 39

 assert(len < sizeof(dst) - 1);
strncat(dst + len, src, sizeof(dst) - len - 1);
printf("%zu: [%s]\n", strlen(dst), dst);

This produces the same output as before, but strncat() doesn't have to scan over the existing content of dst
before it starts copying.

strncpy()

The last option is the strncpy() function. Although you might think it should come first, it is a rather deceptive
function that has two main gotchas:

1. If copying via strncpy() hits the buffer limit, a terminating null-character won't be written.
2. strncpy() always completely fills the destination, with null bytes if necessary.

(Such quirky implementation is historical and was initially intended for handling UNIX file names)

The only correct way to use it is to manually ensure null-termination:

strncpy(b, a, sizeof(b)); /* the third parameter is destination buffer size */

b[sizeof(b)/sizeof(*b) - 1] = '\0'; /* terminate the string */
printf("%s\n", b); /* "0123456" will be printed */

Even then, if you have a big buffer it becomes very inefficient to use strncpy() because of additional null padding.

Section 6.6: Iterating Over the Characters in a String

If we know the length of the string, we can use a for loop to iterate over its characters:

char * string = "hello world"; /* This 11 chars long, excluding the 0-terminator. */
size_t i = 0;
for (; i < 11; i++) {
printf("%c\n", string[i]); /* Print each character of the string. */
}

Alternatively, we can use the standard function strlen() to get the length of a string if we don't know what the
string is:

size_t length = strlen(string);

size_t i = 0;
for (; i < length; i++) {
printf("%c\n", string[i]); /* Print each character of the string. */
}

Finally, we can take advantage of the fact that strings in C are guaranteed to be null-terminated (which we already
did when passing it to strlen() in the previous example ;-)). We can iterate over the array regardless of its size and
stop iterating once we reach a null-character:

size_t i = 0;
while (string[i] != '\0') { /* Stop looping when we reach the null-character. */
printf("%c\n", string[i]); /* Print each character of the string. */
i++;
}

GoalKicker.com – C Notes for Professionals 40

Section 6.7: Creating Arrays of Strings
An array of strings can mean a couple of things:

1. An array whose elements are char *s

2. An array whose elements are arrays of chars

We can create an array of character pointers like so:

char * string_array[] = {
"foo",
"bar",
"baz"
};

Remember: when we assign string literals to char *, the strings themselves are allocated in read-only memory.
However, the array string_array is allocated in read/write memory. This means that we can modify the pointers in
the array, but we cannot modify the strings they point to.

In C, the parameter to main argv (the array of command-line arguments passed when the program was run) is an
array of char *: char * argv[].

We can also create arrays of character arrays. Since strings are arrays of characters, an array of strings is simply an
array whose elements are arrays of characters:

char modifiable_string_array_literals[][4] = {
"foo",
"bar",
"baz"
};

This is equivalent to:

char modifiable_string_array[][4] = {
{'f', 'o', 'o', '\0'},
{'b', 'a', 'r', '\0'},
{'b', 'a', 'z', '\0'}
};

Note that we specify 4 as the size of the second dimension of the array; each of the strings in our array is actually 4
bytes since we must include the null-terminating character.

Section 6.8: Convert Strings to Number: atoi(), atof()

(dangerous, don't use them)
Warning: The functions atoi, atol, atoll and atof are inherently unsafe, because: If the value of the result cannot be
represented, the behavior is undefined. (7.20.1p1)

#include <stdio.h>
#include <stdlib.h>

int main(int argc, char** argv)

{
int val;
if (argc < 2)
{

GoalKicker.com – C Notes for Professionals 41

 printf("Usage: %s <integer>\n", argv[0]);
return 0;
}

val = atoi(argv[1]);

printf("String value = %s, Int value = %d\n", argv[1], val);

return 0;
}

When the string to be converted is a valid decimal integer that is in range, the function works:

$ ./atoi 100
String value = 100, Int value = 100
$ ./atoi 200
String value = 200, Int value = 200

For strings that start with a number, followed by something else, only the initial number is parsed:

$ ./atoi 0x200
0
$ ./atoi 0123x300
123

In all other cases, the behavior is undefined:

$ ./atoi hello
Formatting the hard disk...

Because of the ambiguities above and this undefined behavior, the atoi family of functions should never be used.

To convert to long int, use strtol() instead of atol().

To convert to double, use strtod() instead of atof().

Version ≥ C99

To convert to long long int, use strtoll() instead of atoll().

Section 6.9: string formatted data read/write

Write formatted data to string

int sprintf ( char * str, const char * format, ... );

use sprintf function to write float data to string.

#include <stdio.h>
int main ()
{
char buffer [50];
double PI = 3.1415926;
sprintf (buffer, "PI = %.7f", PI);
printf ("%s\n",buffer);
return 0;
}

GoalKicker.com – C Notes for Professionals 42

Read formatted data from string

int sscanf ( const char * s, const char * format, ...);

use sscanffunction to parse formatted data.

#include <stdio.h>
int main ()
{
char sentence []="date : 06-06-2012";
char str [50];
int year;
int month;
int day;
sscanf (sentence,"%s : %2d-%2d-%4d", str, &day, &month, &year);
printf ("%s -> %02d-%02d-%4d\n",str, day, month, year);
return 0;
}

Section 6.10: Find first/last occurrence of a specific character:

strchr(), strrchr()
The strchr and strrchr functions find a character in a string, that is in a NUL-terminated character array. strchr
return a pointer to the first occurrence and strrchr to the last one.

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

int main(void)
{
char toSearchFor = 'A';

/* Exit if no second argument is found. */

if (argc != 2)
{
printf("Argument missing.\n");
return EXIT_FAILURE;
}

{
char *firstOcc = strchr(argv[1], toSearchFor);
if (firstOcc != NULL)
{
printf("First position of %c in %s is %td.\n",
toSearchFor, argv[1], firstOcc-argv[1]); /* A pointer difference's result
is a signed integer and uses the length modifier 't'. */
}
else
{
printf("%c is not in %s.\n", toSearchFor, argv[1]);
}
}

{
char *lastOcc = strrchr(argv[1], toSearchFor);
if (lastOcc != NULL)
{
printf("Last position of %c in %s is %td.\n",

GoalKicker.com – C Notes for Professionals 43

 toSearchFor, argv[1], lastOcc-argv[1]);
}
}

return EXIT_SUCCESS;
}

Outputs (after having generate an executable named pos):

$ ./pos AAAAAAA
First position of A in AAAAAAA is 0.
Last position of A in AAAAAAA is 6.
$ ./pos BAbbbbbAccccAAAAzzz
First position of A in BAbbbbbAccccAAAAzzz is 1.
Last position of A in BAbbbbbAccccAAAAzzz is 15.
$ ./pos qwerty
A is not in qwerty.

One common use for strrchr is to extract a file name from a path. For example to extract myfile.txt from
C:\Users\eak\myfile.txt:

char *getFileName(const char *path)

{
char *pend;

if ((pend = strrchr(path, '\')) != NULL)

return pend + 1;

return NULL;
}

Section 6.11: Copy and Concatenation: strcpy(), strcat()

#include <stdio.h>
#include <string.h>

int main(void)
{
/* Always ensure that your string is large enough to contain the characters
* and a terminating NUL character ('\0')!
*/
char mystring[10];

/* Copy "foo" into `mystring`, until a NUL character is encountered. */

strcpy(mystring, "foo");
printf("%s\n", mystring);

/* At this point, we used 4 chars of `mystring`, the 3 characters of "foo",

* and the NUL terminating byte.
*/

/* Append "bar" to `mystring`. */

strcat(mystring, "bar");
printf("%s\n", mystring);

/* We now use 7 characters of `mystring`: "foo" requires 3, "bar" requires 3

* and there is a terminating NUL character ('\0') at the end.
*/

GoalKicker.com – C Notes for Professionals 44

 /* Copy "bar" into `mystring`, overwriting the former contents. */
strcpy(mystring, "bar");
printf("%s\n", mystring);

return 0;
}

Outputs:

foo
foobar
bar

If you append to or from or copy from an existing string, ensure it is NUL-terminated!

String literals (e.g. "foo") will always be NUL-terminated by the compiler.

Section 6.12: Comparsion: strcmp(), strncmp(), strcasecmp(),

strncasecmp()
The strcase*-functions are not Standard C, but a POSIX extension.

The strcmp function lexicographically compare two null-terminated character arrays. The functions return a
negative value if the first argument appears before the second in lexicographical order, zero if they compare equal,
or positive if the first argument appears after the second in lexicographical order.

#include <stdio.h>
#include <string.h>

void compare(char const *lhs, char const *rhs)

{
int result = strcmp(lhs, rhs); // compute comparison once
if (result < 0) {
printf("%s comes before %s\n", lhs, rhs);
} else if (result == 0) {
printf("%s equals %s\n", lhs, rhs);
} else { // last case: result > 0
printf("%s comes after %s\n", lhs, rhs);
}
}

int main(void)
{
compare("BBB", "BBB");
compare("BBB", "CCCCC");
compare("BBB", "AAAAAA");
return 0;
}

Outputs:

BBB equals BBB

BBB comes before CCCCC
BBB comes after AAAAAA

As strcmp, strcasecmp function also compares lexicographically its arguments after translating each character to its
lowercase correspondent:

GoalKicker.com – C Notes for Professionals 45

#include <stdio.h>
#include <string.h>

void compare(char const *lhs, char const *rhs)

{
int result = strcasecmp(lhs, rhs); // compute case-insensitive comparison once
if (result < 0) {
printf("%s comes before %s\n", lhs, rhs);
} else if (result == 0) {
printf("%s equals %s\n", lhs, rhs);
} else { // last case: result > 0
printf("%s comes after %s\n", lhs, rhs);
}
}

int main(void)
{
compare("BBB", "bBB");
compare("BBB", "ccCCC");
compare("BBB", "aaaaaa");
return 0;
}

Outputs:

BBB equals bBB

BBB comes before ccCCC
BBB comes after aaaaaa

strncmp and strncasecmp compare at most n characters:

#include <stdio.h>
#include <string.h>

void compare(char const *lhs, char const *rhs, int n)

{
int result = strncmp(lhs, rhs, n); // compute comparison once
if (result < 0) {
printf("%s comes before %s\n", lhs, rhs);
} else if (result == 0) {
printf("%s equals %s\n", lhs, rhs);
} else { // last case: result > 0
printf("%s comes after %s\n", lhs, rhs);
}
}

int main(void)
{
compare("BBB", "Bb", 1);
compare("BBB", "Bb", 2);
compare("BBB", "Bb", 3);
return 0;
}

Outputs:

BBB equals Bb
BBB comes before Bb
BBB comes before Bb

GoalKicker.com – C Notes for Professionals 46

Section 6.13: Safely convert Strings to Number: strtoX
functions
Version ≥ C99

Since C99 the C library has a set of safe conversion functions that interpret a string as a number. Their names are of
the form strtoX, where X is one of l, ul, d, etc to determine the target type of the conversion

double strtod(char const* p, char** endptr);

long double strtold(char const* p, char** endptr);

They provide checking that a conversion had an over- or underflow:

double ret = strtod(argv[1], 0); /* attempt conversion */

/* check the conversion result. */

if ((ret == HUGE_VAL || ret == -HUGE_VAL) && errno == ERANGE)
return; /* numeric overflow in in string */
else if (ret == HUGE_VAL && errno == ERANGE)
return; /* numeric underflow in in string */

/* At this point we know that everything went fine so ret may be used */

If the string in fact contains no number at all, this usage of strtod returns 0.0.

If this is not satisfactory, the additional parameter endptr can be used. It is a pointer to pointer that will be pointed
to the end of the detected number in the string. If it is set to 0, as above, or NULL, it is simply ignored.

This endptr parameter provides indicates if there has been a successful conversion and if so, where the number
ended:

char *check = 0;
double ret = strtod(argv[1], &check); /* attempt conversion */

/* check the conversion result. */

if (argv[1] == check)
return; /* No number was detected in string */
else if ((ret == HUGE_VAL || ret == -HUGE_VAL) && errno == ERANGE)
return; /* numeric overflow in in string */
else if (ret == HUGE_VAL && errno == ERANGE)
return; /* numeric underflow in in string */

/* At this point we know that everything went fine so ret may be used */

There are analogous functions to convert to the wider integer types:

long strtol(char const* p, char** endptr, int nbase);

long long strtoll(char const* p, char** endptr, int nbase);
unsigned long strtoul(char const* p, char** endptr, int nbase);
unsigned long long strtoull(char const* p, char** endptr, int nbase);

These functions have a third parameter nbase that holds the number base in which the number is written.

long a = strtol("101", 0, 2 ); /* a = 5L */
long b = strtol("101", 0, 8 ); /* b = 65L */
long c = strtol("101", 0, 10); /* c = 101L */
long d = strtol("101", 0, 16); /* d = 257L */
long e = strtol("101", 0, 0 ); /* e = 101L */

GoalKicker.com – C Notes for Professionals 47

long f = strtol("0101", 0, 0 ); /* f = 65L */
long g = strtol("0x101", 0, 0 ); /* g = 257L */

The special value 0 for nbase means the string is interpreted in the same way as number literals are interpreted in a
C program: a prefix of 0x corresponds to a hexadecimal representation, otherwise a leading 0 is octal and all other
numbers are seen as decimal.

Thus the most practical way to interpret a command-line argument as a number would be

int main(int argc, char* argv[] {

if (argc < 1)
return EXIT_FAILURE; /* No number given. */

/* use strtoull because size_t may be wide */

size_t mySize = strtoull(argv[1], 0, 0);

/* then check conversion results. */

...

return EXIT_SUCCESS;
}

This means that the program can be called with a parameter in octal, decimal or hexadecimal.

Section 6.14: strspn and strcspn

Given a string, strspn calculates the length of the initial substring (span) consisting solely of a specific list of
characters. strcspn is similar, except it calculates the length of the initial substring consisting of any characters
except those listed:

/*
Provided a string of "tokens" delimited by "separators", print the tokens along
with the token separators that get skipped.
*/
#include <stdio.h>
#include <string.h>

int main(void)
{
const char sepchars[] = ",.;!?";
char foo[] = ";ball call,.fall gall hall!?.,";
char *s;
int n;

for (s = foo; *s != 0; /*empty*/) {

/* Get the number of token separator characters. */
n = (int)strspn(s, sepchars);

if (n > 0)
printf("skipping separators: << %.*s >> (length=%d)\n", n, s, n);

/* Actually skip the separators now. */

s += n;

/* Get the number of token (non-separator) characters. */

n = (int)strcspn(s, sepchars);

GoalKicker.com – C Notes for Professionals 48

 if (n > 0)
printf("token found: << %.*s >> (length=%d)\n", n, s, n);

/* Skip the token now. */

s += n;
}

printf("== token list exhausted ==\n");

return 0;
}

Analogous functions using wide-character strings are wcsspn and wcscspn; they're used the same way.

GoalKicker.com – C Notes for Professionals 49

Chapter 7: Literals for numbers,
characters and strings
Section 7.1: Floating point literals
Floating point literals are used to represent signed real numbers. The following suffixes can be used to specify type
of a literal:

Suffix Type Examples

none double 3.1415926 -3E6
f, F float 3.1415926f 2.1E-6F
l, L long double 3.1415926L 1E126L

In order to use these suffixes, the literal must be a floating point literal. For example, 3f is an error, since 3 is an
integer literal, while 3.f or 3.0f are correct. For long double, the recommendation is to always use capital L for the
sake of readability.

Section 7.2: String literals

String literals are used to specify arrays of characters. They are sequences of characters enclosed within double
quotes (e.g. "abcd" and have the type char*).

The L prefix makes the literal a wide character array, of type wchar_t*. For example, L"abcd".

Since C11, there are other encoding prefixes, similar to L:

prefix base type encoding

none char platform dependent
L wchar_t platform dependent
u8 char UTF-8
u char16_t usually UTF-16
U char32_t usually UTF-32

For the latter two, it can be queried with feature test macros if the encoding is effectively the corresponding UTF
encoding.

Section 7.3: Character literals

Character literals are a special type of integer literals that are used to represent one character. They are enclosed in
single quotes, e.g. 'a' and have the type int. The value of the literal is an integer value according to the machine's
character set. They do not allow suffixes.

The L prefix before a character literal makes it a wide character of type wchar_t. Likewise since C11 u and U prefixes
make it wide characters of type char16_t and char32_t, respectively.

When intending to represent certain special characters, such as a character that is non-printing, escape sequences
are used. Escape sequences use a sequence of characters that are translated into another character. All escape
sequences consist of two or more characters, the first of which is a backslash \. The characters immediately
following the backslash determine what character literal the sequence is interpreted as.

Escape Sequence Represented Character

GoalKicker.com – C Notes for Professionals 50

\b Backspace
\f Form feed
\n Line feed (new line)
\r Carriage return
\t Horizontal tab
\v Vertical tab
\\ Backslash
\' Single quotation mark
\" Double quotation mark
\? Question mark
\nnn Octal value
\xnn... Hexadecimal value
Version ≥ C89
Escape Sequence Represented Character
\a Alert (beep, bell)
Version ≥ C99
Escape Sequence Represented Character
\unnnn Universal character name
\Unnnnnnnn Universal character name

A universal character name is a Unicode code point. A universal character name may map to more than one
character. The digits n are interpreted as hexadecimal digits. Depending on the UTF encoding in use, a universal
character name sequence may result in a code point that consists of multiple characters, instead of a single normal
char character.

When using the line feed escape sequence in text mode I/O, it is converted to the OS-specific newline byte or byte
sequence.

The question mark escape sequence is used to avoid trigraphs. For example, ??/ is compiled as the trigraph
representing a backslash character '\', but using ?\?/ would result in the string "??/".

There may be one, two or three octal numerals n in the octal value escape sequence.

Section 7.4: Integer literals

Integer literals are used to provide integral values. Three numerical bases are supported, indicated by prefixes:

Base Prefix Example

Decimal None 5
Octal 0 0345
Hexadecimal 0x or 0X 0x12AB, 0X12AB, 0x12ab, 0x12Ab

Note that this writing doesn't include any sign, so integer literals are always positive. Something like -1 is treated as
an expression that has one integer literal (1) that is negated with a -

The type of a decimal integer literal is the first data type that can fit the value from int and long. Since C99, long
long is also supported for very large literals.

The type of an octal or hexadecimal integer literal is the first data type that can fit the value from int, unsigned,
long, and unsigned long. Since C99, long long and unsigned long long are also supported for very large literals.

GoalKicker.com – C Notes for Professionals 51

Using various suffixes, the default type of a literal can be changed.

Suffix Explanation
L, l long int
LL, ll (since C99) long long int
U, u unsigned

The U and L/LL suffixes can be combined in any order and case. It is an error to duplicate suffixes (e.g. provide two
U suffixes) even if they have different cases.

GoalKicker.com – C Notes for Professionals 52

Chapter 8: Compound Literals
Section 8.1: Definition/Initialisation of Compound Literals
A compound literal is an unnamed object which is created in the scope where is defined. The concept was first
introduced in C99 standard. An example for compound literal is

Examples from C standard, C11-§6.5.2.5/9:

int *p = (int [2]){ 2, 4 };

p is initialized to the address of the first element of an unnamed array of two ints.

The compound literal is an lvalue. The storage duration of the unnamed object is either static (if the literal appears
at file scope) or automatic (if the literal appears at block scope), and in the latter case the object's lifetime ends
when control leaves the enclosing block.

void f(void)
{
int *p;
/*...*/
p = (int [2]){ *p };
/*...*/
}

p is assigned the address of the first element of an array of two ints, the first having the value previously
pointed to by p and the second, zero.[...]

Here p remains valid until the end of the block.

Compound literal with designators

(also from C11)

struct point {
unsigned x;
unsigned y;
};

extern void drawline(struct point, struct point);

// used somewhere like this

drawline((struct point){.x=1, .y=1}, (struct point){.x=3, .y=4});

A fictive function drawline receives two arguments of type struct point. The first has coordinate values x == 1
and y == 1, whereas the second has x == 3 and y == 4

Compound literal without specifying array length

int *p = (int []){ 1, 2, 3};

In this case the size of the array is no specified then it will be determined by the length of the initializer.

Compound literal having length of initializer less than array size specified
int *p = (int [10]){1, 2, 3};

GoalKicker.com – C Notes for Professionals 53

rest of the elements of compound literal will be initialized to 0 implicitly.

Read-only compound literal

Note that a compound literal is an lvalue and therefore it's elements can be modifiable. A read-only compound
literal can be specified using const qualifier as (const int[]){1,2}.

Compound literal containing arbitrary expressions

Inside a function, a compound literal, as for any initialization since C99, can have arbitrary expressions.

void foo()
{
int *p;
int i = 2; j = 5;
/*...*/
p = (int [2]){ i+j, i*j };
/*...*/
}

GoalKicker.com – C Notes for Professionals 54

Chapter 9: Bit-fields
Parameter Description
type-specifier signed, unsigned, int or _Bool
identifier The name for this field in the structure
size The number of bits to use for this field

Most variables in C have a size that is an integral number of bytes. Bit-fields are a part of a structure that don't
necessarily occupy a integral number of bytes; they can any number of bits. Multiple bit-fields can be packed into a
single storage unit. They are a part of standard C, but there are many aspects that are implementation defined.
They are one of the least portable parts of C.

Section 9.1: Bit-fields

A simple bit-field can be used to describe things that may have a specific number of bits involved.

struct encoderPosition {
unsigned int encoderCounts : 23;
unsigned int encoderTurns : 4;
unsigned int _reserved : 5;
};

In this example we consider an encoder with 23 bits of single precision and 4 bits to describe multi-turn. Bit-fields
are often used when interfacing with hardware that outputs data associated with specific number of bits. Another
example could be communication with an FPGA, where the FPGA writes data into your memory in 32 bit sections
allowing for hardware reads:

struct FPGAInfo {
union {
struct bits {
unsigned int bulb1On : 1;
unsigned int bulb2On : 1;
unsigned int bulb1Off : 1;
unsigned int bulb2Off : 1;
unsigned int jetOn : 1;
};
unsigned int data;
};
};

For this example we have shown a commonly used construct to be able to access the data in its individual bits, or to
write the data packet as a whole (emulating what the FPGA might do). We could then access the bits like this:

FPGAInfo fInfo;
fInfo.data = 0xFF34F;
if (fInfo.bits.bulb1On) {
printf("Bulb 1 is on\n");
}

This is valid, but as per the C99 standard 6.7.2.1, item 10:

The order of allocation of bit-fields within a unit (high-order to low-order or low-order to high-order) is
implementation-defined.

GoalKicker.com – C Notes for Professionals 55

You need to be aware of endianness when defining bit-fields in this way. As such it may be necessary to use a
preprocessor directive to check for the endianness of the machine. An example of this follows:

typedef union {
struct bits {
#if defined(WIN32) || defined(LITTLE_ENDIAN)
uint8_t commFailure :1;
uint8_t hardwareFailure :1;
uint8_t _reserved :6;
#else
uint8_t _reserved :6;
uint8_t hardwareFailure :1;
uint8_t commFailure :1;
#endif
};
uint8_t data;
} hardwareStatus;

Section 9.2: Using bit-fields as small integers

#include <stdio.h>

int main(void)
{
/* define a small bit-field that can hold values from 0 .. 7 */
struct
{
unsigned int uint3: 3;
} small;

/* extract the right 3 bits from a value */

unsigned int value = 255 - 2; /* Binary 11111101 */
small.uint3 = value; /* Binary 101 */
printf("%d", small.uint3);

/* This is in effect an infinite loop */

for (small.uint3 = 0; small.uint3 < 8; small.uint3++)
{
printf("%d\n", small.uint3);
}

return 0;
}

Section 9.3: Bit-field alignment

Bit-fields give an ability to declare structure fields that are smaller than the character width. Bit-fields are
implemented with byte-level or word-level mask. The following example results in a structure of 8 bytes.

struct C
{
short s; /* 2 bytes */
char c; /* 1 byte */
int bit1 : 1; /* 1 bit */
int nib : 4; /* 4 bits padded up to boundary of 8 bits. Thus 3 bits are padded */
int sept : 7; /* 7 Bits septet, padded up to boundary of 32 bits. */
};

GoalKicker.com – C Notes for Professionals 56

The comments describe one possible layout, but because the standard says the alignment of the addressable storage
unit is unspecified, other layouts are also possible.

An unnamed bit-field may be of any size, but they can't be initialized or referenced.

A zero-width bit-field cannot be given a name and aligns the next field to the boundary defined by the datatype of
the bit-field. This is achieved by padding bits between the bit-fields.

The size of structure 'A' is 1 byte.

struct A
{
unsigned char c1 : 3;
unsigned char c2 : 4;
unsigned char c3 : 1;
};

In structure B, the first unnamed bit-field skips 2 bits; the zero width bit-field after c2 causes c3 to start from the
char boundary (so 3 bits are skipped between c2 and c3. There are 3 padding bits after c4. Thus the size of the
structure is 2 bytes.

struct B
{
unsigned char c1 : 1;
unsigned char : 2; /* Skips 2 bits in the layout */
unsigned char c2 : 2;
unsigned char : 0; /* Causes padding up to next container boundary */
unsigned char c3 : 4;
unsigned char c4 : 1;
};

Section 9.4: Don'ts for bit-fields

1. Arrays of bit-fields, pointers to bit-fields and functions returning bit-fields are not allowed.
2. The address operator (&) cannot be applied to bit-field members.
3. The data type of a bit-field must be wide enough to contain the size of the field.
4. The sizeof() operator cannot be applied to a bit-field.
5. There is no way to create a typedef for a bit-field in isolation (though you can certainly create a typedef for a
structure containing bit-fields).

typedef struct mybitfield

{
unsigned char c1 : 20; /* incorrect, see point 3 */
unsigned char c2 : 4; /* correct */
unsigned char c3 : 1;
unsigned int x[10]: 5; /* incorrect, see point 1 */
} A;

int SomeFunction(void)
{
// Somewhere in the code
A a = { … };
printf("Address of a.c2 is %p\n", &a.c2); /* incorrect, see point 2 */
printf("Size of a.c2 is %zu\n", sizeof(a.c2)); /* incorrect, see point 4 */
}

GoalKicker.com – C Notes for Professionals 57

Section 9.5: When are bit-fields useful?
A bit-field is used to club together many variables into one object, similar to a structure. This allows for reduced
memory usage and is especially useful in an embedded environment.

e.g. consider the following variables having the ranges as given below.
a --> range 0 - 3
b --> range 0 - 1
c --> range 0 - 7
d --> range 0 - 1
e --> range 0 - 1

If we declare these variables separately, then each has to be at least an 8-bit integer and the total space required
will be 5 bytes. Moreover the variables will not use the entire range of an 8 bit unsigned integer (0-255). Here we
can use bit-fields.

typedef struct {
unsigned int a:2;
unsigned int b:1;
unsigned int c:3;
unsigned int d:1;
unsigned int e:1;
} bit_a;

The bit-fields in the structure are accessed the same as any other structure. The programmer needs to take care
that the variables are written in range. If out of range the behaviour is undefined.

int main(void)
{
bit_a bita_var;
bita_var.a = 2; // to write into element a
printf ("%d",bita_var.a); // to read from element a.
return 0;
}

Often the programmer wants to zero the set of bit-fields. This can be done element by element, but there is second
method. Simply create a union of the structure above with an unsigned type that is greater than, or equal to, the
size of the structure. Then the entire set of bit-fields may be zeroed by zeroing this unsigned integer.

typedef union {
struct {
unsigned int a:2;
unsigned int b:1;
unsigned int c:3;
unsigned int d:1;
unsigned int e:1;
};
uint8_t data;
} union_bit;

Usage is as follows

int main(void)
{
union_bit un_bit;
un_bit.data = 0x00; // clear the whole bit-field
un_bit.a = 2; // write into element a

GoalKicker.com – C Notes for Professionals 58

 printf ("%d",un_bit.a); // read from element a.
return 0;
}

In conclusion, bit-fields are commonly used in memory constrained situations where you have a lot of variables
which can take on limited ranges.

GoalKicker.com – C Notes for Professionals 59

Chapter 10: Arrays
Arrays are derived data types, representing an ordered collection of values ("elements") of another type. Most
arrays in C have a fixed number of elements of any one type, and its representation stores the elements
contiguously in memory without gaps or padding. C allows multidimensional arrays whose elements are other
arrays, and also arrays of pointers.

C supports dynamically allocated arrays whose size is determined at run time. C99 and later supports variable
length arrays or VLAs.

Section 10.1: Declaring and initializing an array

The general syntax for declaring a one-dimensional array is

type arrName[size];

where type could be any built-in type or user-defined types such as structures, arrName is a user-defined identifier,
and size is an integer constant.

Declaring an array (an array of 10 int variables in this case) is done like this:

int array[10];

it now holds indeterminate values. To ensure it holds zero values while declaring, you can do this:

int array[10] = {0};

Arrays can also have initializers, this example declares an array of 10 int's, where the first 3 int's will contain the
values 1, 2, 3, all other values will be zero:

int array[10] = {1, 2, 3};

In the above method of initialization, the first value in the list will be assigned to the first member of the array, the
second value will be assigned to the second member of the array and so on. If the list size is smaller than the array
size, then as in the above example, the remaining members of the array will be initialized to zeros. With designated
list initialization (ISO C99), explicit initialization of the array members is possible. For example,

int array[5] = {[2] = 5, [1] = 2, [4] = 9}; /* array is {0, 2, 5, 0, 9} */

In most cases, the compiler can deduce the length of the array for you, this can be achieved by leaving the square
brackets empty:

int array[] = {1, 2, 3}; /* an array of 3 int's */

int array[] = {[3] = 8, [0] = 9}; /* size is 4 */

Declaring an array of zero length is not allowed.

Version ≥ C99 Version < C11

Variable Length Arrays (VLA for short) were added in C99, and made optional in C11. They are equal to normal
arrays, with one, important, difference: The length doesn't have to be known at compile time. VLA's have automatic
storage duration. Only pointers to VLA's can have static storage duration.

GoalKicker.com – C Notes for Professionals 60

size_t m = calc_length(); /* calculate array length at runtime */
int vla[m]; /* create array with calculated length */

Important:

VLA's are potentially dangerous. If the array vla in the example above requires more space on the stack than
available, the stack will overflow. Usage of VLA's is therefore often discouraged in style guides and by books and
exercises.

Section 10.2: Iterating through an array eciently and row-

major order
Arrays in C can be seen as a contiguous chunk of memory. More precisely, the last dimension of the array is the
contiguous part. We call this the row-major order. Understanding this and the fact that a cache fault loads a
complete cache line into the cache when accessing uncached data to prevent subsequent cache faults, we can see
why accessing an array of dimension 10000x10000 with array[0][0] would potentially load array[0][1] in cache,
but accessing array[1][0] right after would generate a second cache fault, since it is sizeof(type)*10000 bytes
away from array[0][0], and therefore certainly not on the same cache line. Which is why iterating like this is
inefficient:

#define ARRLEN 10000

int array[ARRLEN][ARRLEN];

size_t i, j;
for (i = 0; i < ARRLEN; ++i)
{
for(j = 0; j < ARRLEN; ++j)
{
array[j][i] = 0;
}
}

And iterating like this is more efficient:

#define ARRLEN 10000

int array[ARRLEN][ARRLEN];

size_t i, j;
for (i = 0; i < ARRLEN; ++i)
{
for(j = 0; j < ARRLEN; ++j)
{
array[i][j] = 0;
}
}

In the same vein, this is why when dealing with an array with one dimension and multiple indexes (let's say 2
dimensions here for simplicity with indexes i and j) it is important to iterate through the array like this:

#define DIM_X 10
#define DIM_Y 20

int array[DIM_X*DIM_Y];

size_t i, j;
for (i = 0; i < DIM_X; ++i)

GoalKicker.com – C Notes for Professionals 61

{
for(j = 0; j < DIM_Y; ++j)
{
array[i*DIM_Y+j] = 0;
}
}

Or with 3 dimensions and indexes i,j and k:

#define DIM_X 10
#define DIM_Y 20
#define DIM_Z 30

int array[DIM_X*DIM_Y*DIM_Z];

size_t i, j, k;
for (i = 0; i < DIM_X; ++i)
{
for(j = 0; j < DIM_Y; ++j)
{
for (k = 0; k < DIM_Z; ++k)
{
array[i*DIM_Y*DIM_Z+j*DIM_Z+k] = 0;
}
}
}

Or in a more generic way, when we have an array with N1 x N2 x ... x Nd elements, d dimensions and indices noted
as n1,n2,...,nd the offset is calculated like this

Picture/formula taken from: https://en.wikipedia.org/wiki/Row-major_order

Section 10.3: Array length

Arrays have fixed lengths that are known within the scope of their declarations. Nevertheless, it is possible and
sometimes convenient to calculate array lengths. In particular, this can make code more flexible when the array
length is determined automatically from an initializer:

int array[] = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 };

/* size of `array` in bytes */

size_t size = sizeof(array);

/* number of elements in `array` */

size_t length = sizeof(array) / sizeof(array[0]);

However, in most contexts where an array appears in an expression, it is automatically converted to ("decays to") a
pointer to its first element. The case where an array is the operand of the sizeof operator is one of a small number
of exceptions. The resulting pointer is not itself an array, and it does not carry any information about the length of
the array from which it was derived. Therefore, if that length is needed in conjunction with the pointer, such as
when the pointer is passed to a function, then it must be conveyed separately.

For example, suppose we want to write a function to return the last element of an array of int. Continuing from the

GoalKicker.com – C Notes for Professionals 62

above, we might call it like so:

/* array will decay to a pointer, so the length must be passed separately */

int last = get_last(array, length);

The function could be implemented like this:

int get_last(int input[], size_t length) {

return input[length - 1];
}

Note in particular that although the declaration of parameter input resembles that of an array, it in fact declares
input as a pointer (to int). It is exactly equivalent to declaring input as int *input. The same would be true even
if a dimension were given. This is possible because arrays cannot ever be actual arguments to functions (they decay
to pointers when they appear in function call expressions), and it can be viewed as mnemonic.

It is a very common error to attempt to determine array size from a pointer, which cannot work. DO NOT DO THIS:

int BAD_get_last(int input[]) {

/* INCORRECTLY COMPUTES THE LENGTH OF THE ARRAY INTO WHICH input POINTS: */
size_t length = sizeof(input) / sizeof(input[0]));

return input[length - 1]; /* Oops -- not the droid we are looking for */
}

In fact, that particular error is so common that some compilers recognize it and warn about it. clang, for instance,
will emit the following warning:

warning: sizeof on array function parameter will return size of 'int *' instead of 'int []' [-
Wsizeof-array-argument]
int length = sizeof(input) / sizeof(input[0]);
^
note: declared here
int BAD_get_last(int input[])
^

Section 10.4: Passing multidimensional arrays to a function

Multidimensional arrays follow the same rules as single-dimensional arrays when passing them to a function.
However the combination of decay-to-pointer, operator precedence, and the two different ways to declare a
multidimensional array (array of arrays vs array of pointers) may make the declaration of such functions non-
intuitive. The following example shows the correct ways to pass multidimensional arrays.

#include <assert.h>
#include <stdlib.h>

/* When passing a multidimensional array (i.e. an array of arrays) to a

function, it decays into a pointer to the first element as usual. But only
the top level decays, so what is passed is a pointer to an array of some fixed
size (4 in this case). */
void f(int x[][4]) {
assert(sizeof(*x) == sizeof(int) * 4);
}

/* This prototype is equivalent to f(int x[][4]).

The parentheses around *x are required because [index] has a higher
precedence than *expr, thus int *x[4] would normally be equivalent to int

GoalKicker.com – C Notes for Professionals 63

 *(x[4]), i.e. an array of 4 pointers to int. But if it's declared as a
function parameter, it decays into a pointer and becomes int **x,
which is not compatable with x[2][4]. */
void g(int (*x)[4]) {
assert(sizeof(*x) == sizeof(int) * 4);
}

/* An array of pointers may be passed to this, since it'll decay into a pointer
to pointer, but an array of arrays may not. */
void h(int **x) {
assert(sizeof(*x) == sizeof(int*));
}

int main(void) {
int foo[2][4];
f(foo);
g(foo);

/* Here we're dynamically creating an array of pointers. Note that the

size of each dimension is not part of the datatype, and so the type
system just treats it as a pointer to pointer, not a pointer to array
or array of arrays. */
int **bar = malloc(sizeof(*bar) * 2);
assert(bar);
for (size_t i = 0; i < 2; i++) {
bar[i] = malloc(sizeof(*bar[i]) * 4);
assert(bar[i]);
}

h(bar);

for (size_t i = 0; i < 2; i++) {

free(bar[i]);
}
free(bar);
}

See also

Passing in Arrays to Functions

Section 10.5: Multi-dimensional arrays

The C programming language allows multidimensional arrays. Here is the general form of a multidimensional array
declaration −

type name[size1][size2]...[sizeN];

For example, the following declaration creates a three dimensional (5 x 10 x 4) integer array:

int arr[5][10][4];

Two-dimensional Arrays

The simplest form of multidimensional array is the two-dimensional array. A two-dimensional array is, in essence, a
list of one-dimensional arrays. To declare a two-dimensional integer array of dimensions m x n, we can write as
follows:

GoalKicker.com – C Notes for Professionals 64

type arrayName[m][n];

Where type can be any valid C data type (int, float, etc.) and arrayName can be any valid C identifier. A two-
dimensional array can be visualized as a table with m rows and n columns. Note: The order does matter in C. The
array int a[4][3] is not the same as the array int a[3][4]. The number of rows comes first as C is a row-major
language.

A two-dimensional array a, which contains three rows and four columns can be shown as follows:

Thus, every element in the array a is identified by an element name of the form a[i][j], where a is the name of the
array, i represents which row, and j represents which column. Recall that rows and columns are zero indexed. This
is very similar to mathematical notation for subscripting 2-D matrices.

Initializing Two-Dimensional Arrays

Multidimensional arrays may be initialized by specifying bracketed values for each row. The following define an
array with 3 rows where each row has 4 columns.

int a[3][4] = {
{0, 1, 2, 3} , /* initializers for row indexed by 0 */
{4, 5, 6, 7} , /* initializers for row indexed by 1 */
{8, 9, 10, 11} /* initializers for row indexed by 2 */
};

The nested braces, which indicate the intended row, are optional. The following initialization is equivalent to the
previous example:

int a[3][4] = {0,1,2,3,4,5,6,7,8,9,10,11};

While the method of creating arrays with nested braces is optional, it is strongly encouraged as it is more readable
and clearer.

Accessing Two-Dimensional Array Elements

An element in a two-dimensional array is accessed by using the subscripts, i.e., row index and column index of the
array. For example −

int val = a[2][3];

The above statement will take the 4th element from the 3rd row of the array. Let us check the following program
where we have used a nested loop to handle a two-dimensional array:

#include <stdio.h>

int main () {

/* an array with 5 rows and 2 columns*/

int a[5][2] = { {0,0}, {1,2}, {2,4}, {3,6},{4,8}};

GoalKicker.com – C Notes for Professionals 65

 int i, j;

/* output each array element's value */

for ( i = 0; i < 5; i++ ) {

for ( j = 0; j < 2; j++ ) {

printf("a[%d][%d] = %d\n", i,j, a[i][j] );
}
}

return 0;
}

When the above code is compiled and executed, it produces the following result:

a[0][0]: 0
a[0][1]: 0
a[1][0]: 1
a[1][1]: 2
a[2][0]: 2
a[2][1]: 4
a[3][0]: 3
a[3][1]: 6
a[4][0]: 4
a[4][1]: 8

Three-Dimensional array:

A 3D array is essentially an array of arrays of arrays: it's an array or collection of 2D arrays, and a 2D array is an
array of 1D arrays.

3D array memory map:

Initializing a 3D Array:

double cprogram[3][2][4]={
{{-0.1, 0.22, 0.3, 4.3}, {2.3, 4.7, -0.9, 2}},
{{0.9, 3.6, 4.5, 4}, {1.2, 2.4, 0.22, -1}},
{{8.2, 3.12, 34.2, 0.1}, {2.1, 3.2, 4.3, -2.0}}
};

We can have arrays with any number of dimensions, although it is likely that most of the arrays that are created will
be of one or two dimensions.

GoalKicker.com – C Notes for Professionals 66

Section 10.6: Define array and access array element
#include <stdio.h>

#define ARRLEN (10)

int main (void)

{

int n[ ARRLEN ]; /* n is an array of 10 integers */

size_t i, j; /* Use size_t to address memory, that is to index arrays, as its guaranteed to
be wide enough to address all of the possible available memory.
Using signed integers to do so should be considered a special use case,
and should be restricted to the uncommon case of being in the need of
negative indexes. */

/* Initialize elements of array n. */

for ( i = 0; i < ARRLEN ; i++ )
{
n[ i ] = i + 100; /* Set element at location i to i + 100. */
}

/* Output each array element's value. */

for (j = 0; j < ARRLEN ; j++ )
{
printf("Element[%zu] = %d\n", j, n[j] );
}

return 0;
}

Section 10.7: Clearing array contents (zeroing)

Sometimes it's necessary to set an array to zero, after the initialization has been done.

#include <stdlib.h> /* for EXIT_SUCCESS */

#define ARRLEN (10)

int main(void)
{
int array[ARRLEN]; /* Allocated but not initialised, as not defined static or global. */

size_t i;
for(i = 0; i < ARRLEN; ++i)
{
array[i] = 0;
}

return EXIT_SUCCESS;
}

An common short cut to the above loop is to use memset() from <string.h>. Passing array as shown below makes
it decay to a pointer to its 1st element.

memset(array, 0, ARRLEN * sizeof (int)); /* Use size explicitly provided type (int here). */

or

GoalKicker.com – C Notes for Professionals 67

memset(array, 0, ARRLEN * sizeof *array); /* Use size of type the pointer is pointing to. */

As in this example array is an array and not just a pointer to an array's 1st element (see Array length on why this is
important) a third option to 0-out the array is possible:

memset(array, 0, sizeof array); /* Use size of the array itself. */

Section 10.8: Setting values in arrays

Accessing array values is generally done through square brackets:

int val;
int array[10];

/* Setting the value of the fifth element to 5: */

array[4] = 5;

/* The above is equal to: */

*(array + 4) = 5;

/* Reading the value of the fifth element: */

val = array[4];

As a side effect of the operands to the + operator being exchangeable (--> commutative law) the following is
equivalent:

*(array + 4) = 5;
*(4 + array) = 5;

so as well the next statements are equivalent:

array[4] = 5;
4[array] = 5; /* Weird but valid C ... */

and those two as well:

val = array[4];
val = 4[array]; /* Weird but valid C ... */

C doesn't perform any boundary checks, accessing contents outside of the declared array is undefined (Accessing
memory beyond allocated chunk ):

int val;
int array[10];

array[4] = 5; /* ok */
val = array[4]; /* ok */
array[19] = 20; /* undefined behavior */
val = array[15]; /* undefined behavior */

Section 10.9: Allocate and zero-initialize an array with user

defined size
#include <stdio.h>
#include <stdlib.h>

GoalKicker.com – C Notes for Professionals 68

int main (void)
{
int * pdata;
size_t n;

printf ("Enter the size of the array: ");

fflush(stdout); /* Make sure the prompt gets printed to buffered stdout. */

if (1 != scanf("%zu", &n)) /* If zu is not supported (Windows?) use lu. */

{
fprintf("scanf() did not read a in proper value.\n");
exit(EXIT_FAILURE);
}

pdata = calloc(n, sizeof *pdata);

if (NULL == pdata)
{
perror("calloc() failed"); /* Print error. */
exit(EXIT_FAILURE);
}

free(pdata); /* Clean up. */

return EXIT_SUCCESS;
}

This program tries to scan in an unsigned integer value from standard input, allocate a block of memory for an
array of n elements of type int by calling the calloc() function. The memory is initialized to all zeros by the latter.

In case of success the memory is releases by the call to free().

Section 10.10: Iterating through an array using pointers

#include <stdio.h>
#define SIZE (10)
int main()
{
size_t i = 0;
int *p = NULL;
int a[SIZE];

/* Setting up the values to be i*i */

for(i = 0; i < SIZE; ++i)
{
a[i] = i * i;
}

/* Reading the values using pointers */

for(p = a; p < a + SIZE; ++p)
{
printf("%d\n", *p);
}

return 0;
}

Here, in the initialization of p in the first for loop condition, the array a decays to a pointer to its first element, as it
would in almost all places where such an array variable is used.

Then, the ++p performs pointer arithmetic on the pointer p and walks one by one through the elements of the

GoalKicker.com – C Notes for Professionals 69

array, and refers to them by dereferencing them with *p.

GoalKicker.com – C Notes for Professionals 70

Chapter 11: Linked lists
Section 11.1: A doubly linked list
An example of code showing how nodes can be inserted at a doubly linked list, how the list can easily be reversed,
and how it can be printed in reverse.

#include <stdio.h>
#include <stdlib.h>

/* This data is not always stored in a structure, but it is sometimes for ease of use */
struct Node {
/* Sometimes a key is also stored and used in the functions */
int data;
struct Node* next;
struct Node* previous;
};

void insert_at_beginning(struct Node **pheadNode, int value);

void insert_at_end(struct Node **pheadNode, int value);

void print_list(struct Node *headNode);

void print_list_backwards(struct Node *headNode);

void free_list(struct Node *headNode);

int main(void) {
/* Sometimes in a doubly linked list the last node is also stored */
struct Node *head = NULL;

printf("Insert a node at the beginning of the list.\n");

insert_at_beginning(&head, 5);
print_list(head);

printf("Insert a node at the beginning, and then print the list backwards\n");
insert_at_beginning(&head, 10);
print_list_backwards(head);

printf("Insert a node at the end, and then print the list forwards.\n");

insert_at_end(&head, 15);
print_list(head);

free_list(head);

return 0;
}

void print_list_backwards(struct Node *headNode) {

if (NULL == headNode)
{
return;
}
/*
Iterate through the list, and once we get to the end, iterate backwards to print
out the items in reverse order (this is done with the pointer to the previous node).
This can be done even more easily if a pointer to the last node is stored.
*/
struct Node *i = headNode;
while (i->next != NULL) {

GoalKicker.com – C Notes for Professionals 71

 i = i->next; /* Move to the end of the list */
}

while (i != NULL) {
printf("Value: %d\n", i->data);
i = i->previous;
}
}

void print_list(struct Node *headNode) {

/* Iterate through the list and print out the data member of each node */
struct Node *i;
for (i = headNode; i != NULL; i = i->next) {
printf("Value: %d\n", i->data);
}
}

void insert_at_beginning(struct Node **pheadNode, int value) {

struct Node *currentNode;

if (NULL == pheadNode)
{
return;
}
/*
This is done similarly to how we insert a node at the beginning of a singly linked
list, instead we set the previous member of the structure as well
*/
currentNode = malloc(sizeof *currentNode);

currentNode->next = NULL;
currentNode->previous = NULL;
currentNode->data = value;

if (*pheadNode == NULL) { /* The list is empty */

*pheadNode = currentNode;
return;
}

currentNode->next = *pheadNode;
(*pheadNode)->previous = currentNode;
*pheadNode = currentNode;
}

void insert_at_end(struct Node **pheadNode, int value) {

struct Node *currentNode;

if (NULL == pheadNode)
{
return;
}

/*
This can, again be done easily by being able to have the previous element. It
would also be even more useful to have a pointer to the last node, which is commonly
used.
*/

currentNode = malloc(sizeof *currentNode);

struct Node *i = *pheadNode;

currentNode->data = value;

GoalKicker.com – C Notes for Professionals 72

 currentNode->next = NULL;
currentNode->previous = NULL;

if (*pheadNode == NULL) {
*pheadNode = currentNode;
return;
}

while (i->next != NULL) { /* Go to the end of the list */

i = i->next;
}

i->next = currentNode;
currentNode->previous = i;
}

void free_list(struct Node *node) {

while (node != NULL) {
struct Node *next = node->next;
free(node);
node = next;
}
}

Note that sometimes, storing a pointer to the last node is useful (it is more efficient to simply be able to jump
straight to the end of the list than to need to iterate through to the end):

struct Node *lastNode = NULL;

In which case, updating it upon changes to the list is needed.

Sometimes, a key is also used to identify elements. It is simply a member of the Node structure:

struct Node {
int data;
int key;
struct Node* next;
struct Node* previous;
};

The key is then used when any tasks are performed on a specific element, like deleting elements.

Section 11.2: Reversing a linked list

You can also perform this task recursively, but I have chosen in this example to use an iterative approach. This task
is useful if you are inserting all of your nodes at the beginning of a linked list. Here is an example:

#include <stdio.h>
#include <stdlib.h>

#define NUM_ITEMS 10

struct Node {
int data;
struct Node *next;
};

void insert_node(struct Node **headNode, int nodeValue, int position);

void print_list(struct Node *headNode);

GoalKicker.com – C Notes for Professionals 73

void reverse_list(struct Node **headNode);

int main(void) {
int i;
struct Node *head = NULL;

for(i = 1; i <= NUM_ITEMS; i++) {

insert_node(&head, i, i);
}
print_list(head);

printf("I will now reverse the linked list\n");

reverse_list(&head);
print_list(head);
return 0;
}

void print_list(struct Node *headNode) {

struct Node *iterator;

for(iterator = headNode; iterator != NULL; iterator = iterator->next) {

printf("Value: %d\n", iterator->data);
}
}

void insert_node(struct Node **headNode, int nodeValue, int position) {

int i;
struct Node *currentNode = (struct Node *)malloc(sizeof(struct Node));
struct Node *nodeBeforePosition = *headNode;

currentNode->data = nodeValue;

if(position == 1) {
currentNode->next = *headNode;
*headNode = currentNode;
return;
}

for (i = 0; i < position - 2; i++) {

nodeBeforePosition = nodeBeforePosition->next;
}

currentNode->next = nodeBeforePosition->next;
nodeBeforePosition->next = currentNode;
}

void reverse_list(struct Node **headNode) {

struct Node *iterator = *headNode;
struct Node *previousNode = NULL;
struct Node *nextNode = NULL;

while (iterator != NULL) {

nextNode = iterator->next;
iterator->next = previousNode;
previousNode = iterator;
iterator = nextNode;
}

/* Iterator will be NULL by the end, so the last node will be stored in
previousNode. We will set the last node to be the headNode */
*headNode = previousNode;

GoalKicker.com – C Notes for Professionals 74

}

Explanation for the Reverse List Method

We start the previousNode out as NULL, since we know on the first iteration of the loop, if we are looking for the
node before the first head node, it will be NULL. The first node will become the last node in the list, and the next
variable should naturally be NULL.

Basically, the concept of reversing the linked list here is that we actually reverse the links themselves. Each node's
next member will become the node before it, like so:

Head -> 1 -> 2 -> 3 -> 4 -> 5

Where each number represents a node. This list would become:

1 <- 2 <- 3 <- 4 <- 5 <- Head

Finally, the head should point to the 5th node instead, and each node should point to the node previous of it.

Node 1 should point to NULL since there was nothing before it. Node 2 should point to node 1, node 3 should point
to node 2, et cetera.

However, there is one small problem with this method. If we break the link to the next node and change it to the
previous node, we will not be able to traverse to the next node in the list since the link to it is gone.

The solution to this problem is to simply store the next element in a variable (nextNode) before changing the link.

Section 11.3: Inserting a node at the nth position

So far, we have looked at inserting a node at the beginning of a singly linked list. However, most of the times you
will want to be able to insert nodes elsewhere as well. The code written below shows how it is possible to write an
insert() function to insert nodes anywhere in the linked lists.

#include <stdio.h>
#include <stdlib.h>

struct Node {
int data;
struct Node* next;
};

struct Node* insert(struct Node* head, int value, size_t position);

void print_list (struct Node* head);

int main(int argc, char *argv[]) {

struct Node *head = NULL; /* Initialize the list to be empty */

/* Insert nodes at positions with values: */

head = insert(head, 1, 0);
head = insert(head, 100, 1);
head = insert(head, 21, 2);
head = insert(head, 2, 3);
head = insert(head, 5, 4);
head = insert(head, 42, 2);

print_list(head);
return 0;

GoalKicker.com – C Notes for Professionals 75

}

struct Node* insert(struct Node* head, int value, size_t position) {

size_t i = 0;
struct Node *currentNode;

/* Create our node */

currentNode = malloc(sizeof *currentNode);
/* Check for success of malloc() here! */

/* Assign data */
currentNode->data = value;

/* Holds a pointer to the 'next' field that we have to link to the new node.
By initializing it to &head we handle the case of insertion at the beginning. */
struct Node **nextForPosition = &head;
/* Iterate to get the 'next' field we are looking for.
Note: Insert at the end if position is larger than current number of elements. */
for (i = 0; i < position && *nextForPosition != NULL; i++) {
/* nextForPosition is pointing to the 'next' field of the node.
So *nextForPosition is a pointer to the next node.
Update it with a pointer to the 'next' field of the next node. */
nextForPosition = &(*nextForPosition)->next;
}

/* Here, we are taking the link to the next node (the one our newly inserted node should
point to) by dereferencing nextForPosition, which points to the 'next' field of the node
that is in the position we want to insert our node at.
We assign this link to our next value. */
currentNode->next = *nextForPosition;

/* Now, we want to correct the link of the node before the position of our
new node: it will be changed to be a pointer to our new node. */
*nextForPosition = currentNode;

return head;
}

void print_list (struct Node* head) {

/* Go through the list of nodes and print out the data in each node */
struct Node* i = head;
while (i != NULL) {
printf("%d\n", i->data);
i = i->next;
}
}

Section 11.4: Inserting a node at the beginning of a singly

linked list
The code below will prompt for numbers and continue to add them to the beginning of a linked list.

/* This program will demonstrate inserting a node at the beginning of a linked list */

#include <stdio.h>
#include <stdlib.h>

struct Node {
int data;
struct Node* next;
};

GoalKicker.com – C Notes for Professionals 76

void insert_node (struct Node **head, int nodeValue);
void print_list (struct Node *head);

int main(int argc, char *argv[]) {

struct Node* headNode;
headNode = NULL; /* Initialize our first node pointer to be NULL. */
size_t listSize, i;
do {
printf("How many numbers would you like to input?\n");
} while(1 != scanf("%zu", &listSize));

for (i = 0; i < listSize; i++) {

int numToAdd;
do {
printf("Enter a number:\n");
} while (1 != scanf("%d", &numToAdd));

insert_node (&headNode, numToAdd);

printf("Current list after your inserted node: \n");
print_list(headNode);
}

return 0;
}

void print_list (struct Node *head) {

struct node* currentNode = head;

/* Iterate through each link. */

while (currentNode != NULL) {
printf("Value: %d\n", currentNode->data);
currentNode = currentNode -> next;
}
}

void insert_node (struct Node **head, int nodeValue) {

struct Node *currentNode = malloc(sizeof *currentNode);
currentNode->data = nodeValue;
currentNode->next = (*head);

*head = currentNode;
}

Explanation for the Insertion of Nodes

In order to understand how we add nodes at the beginning, let's take a look at possible scenarios:

1. The list is empty, so we need to add a new node. In which case, our memory looks like this where HEAD is a
pointer to the first node:

| HEAD | --> NULL

The line currentNode->next = *headNode; will assign the value of currentNode->next to be NULL since headNode
originally starts out at a value of NULL.

Now, we want to set our head node pointer to point to our current node.

----- -------------
|HEAD | --> |CURRENTNODE| --> NULL /* The head node points to the current node */

GoalKicker.com – C Notes for Professionals 77

----- -------------

This is done with *headNode = currentNode;

2. The list is already populated; we need to add a new node to the beginning. For the sake of simplicity, let's
start out with 1 node:

----- -----------
HEAD --> FIRST NODE --> NULL
----- -----------

With currentNode->next = *headNode, the data structure looks like this:

--------- ----- ---------------------

currentNode --> HEAD --> POINTER TO FIRST NODE --> NULL
--------- ----- ---------------------

Which, obviously needs to be altered since *headNode should point to currentNode.

---- ----------- ---------------

HEAD -> currentNode --> NODE -> NULL
---- ----------- ---------------

This is done with *headNode = currentNode;

GoalKicker.com – C Notes for Professionals 78

Chapter 12: Enumerations
Section 12.1: Simple Enumeration
An enumeration is a user-defined data type consists of integral constants and each integral constant is given a
name. Keyword enum is used to define enumerated data type.

If you use enum instead of int or string/ char*, you increase compile-time checking and avoid errors from passing
in invalid constants, and you document which values are legal to use.

Example 1
enum color{ RED, GREEN, BLUE };

void printColor(enum color chosenColor)

{
const char *color_name = "Invalid color";
switch (chosenColor)
{
case RED:
color_name = "RED";
break;

case GREEN:
color_name = "GREEN";
break;

case BLUE:
color_name = "BLUE";
break;
}
printf("%s\n", color_name);
}

With a main function defined as follows (for example):

int main(){
enum color chosenColor;
printf("Enter a number between 0 and 2");
scanf("%d", (int*)&chosenColor);
printColor(chosenColor);
return 0;
}
Version ≥ C99
Example 2

(This example uses designated initializers which are standardized since C99.)

enum week{ MON, TUE, WED, THU, FRI, SAT, SUN };

static const char* const dow[] = {

[MON] = "Mon", [TUE] = "Tue", [WED] = "Wed",
[THU] = "Thu", [FRI] = "Fri", [SAT] = "Sat", [SUN] = "Sun" };

void printDayOfWeek(enum week day)

{
printf("%s\n", dow[day]);
}

GoalKicker.com – C Notes for Professionals 79

The same example using range checking:

enum week{ DOW_INVALID = -1,

MON, TUE, WED, THU, FRI, SAT, SUN,
DOW_MAX };

static const char* const dow[] = {

[MON] = "Mon", [TUE] = "Tue", [WED] = "Wed",
[THU] = "Thu", [FRI] = "Fri", [SAT] = "Sat", [SUN] = "Sun" };

void printDayOfWeek(enum week day)

{
assert(day > DOW_INVALID && day < DOW_MAX);
printf("%s\n", dow[day]);
}

Section 12.2: enumeration constant without typename

Enumeration types can also be declared without giving them a name:

enum { buffersize = 256, };

static unsigned char buffer [buffersize] = { 0 };

This enables us to define compile time constants of type int that can as in this example be used as array length.

Section 12.3: Enumeration with duplicate value

An enumerations value in no way needs to be unique:

#include <stdlib.h> /* for EXIT_SUCCESS */

#include <stdio.h> /* for printf() */

enum Dupes
{
Base, /* Takes 0 */
One, /* Takes Base + 1 */
Two, /* Takes One + 1 */
Negative = -1,
AnotherZero /* Takes Negative + 1 == 0, sigh */
};

int main(void)
{
printf("Base = %d\n", Base);
printf("One = %d\n", One);
printf("Two = %d\n", Two);
printf("Negative = %d\n", Negative);
printf("AnotherZero = %d\n", AnotherZero);

return EXIT_SUCCESS;
}

The sample prints:

Base = 0
One = 1
Two = 2

GoalKicker.com – C Notes for Professionals 80

Negative = -1
AnotherZero = 0

Section 12.4: Typedef enum

There are several possibilities and conventions to name an enumeration. The first is to use a tag name just after the
enum keyword.

enum color
{
RED,
GREEN,
BLUE
};

This enumeration must then always be used with the keyword and the tag like this:

enum color chosenColor = RED;

If we use typedef directly when declaring the enum, we can omit the tag name and then use the type without the
enum keyword:

typedef enum
{
RED,
GREEN,
BLUE
} color;

color chosenColor = RED;

But in this latter case we cannot use it as enum color, because we didn't use the tag name in the definition. One
common convention is to use both, such that the same name can be used with or without enum keyword. This has
the particular advantage of being compatible with C++

enum color /* as in the first example */

{
RED,
GREEN,
BLUE
};
typedef enum color color; /* also a typedef of same identifier */

color chosenColor = RED;

enum color defaultColor = BLUE;

Function:

void printColor()
{
if (chosenColor == RED)
{
printf("RED\n");
}
else if (chosenColor == GREEN)
{
printf("GREEN\n");

GoalKicker.com – C Notes for Professionals 81

 }
else if (chosenColor == BLUE)
{
printf("BLUE\n");
}
}

For more on typedef see Typedef

GoalKicker.com – C Notes for Professionals 82

Chapter 13: Structs
Structures provide a way to group a set of related variables of diverse types into a single unit of memory. The
structure as a whole can be referenced by a single name or pointer; the structure members can be accessed
individually too. Structures can be passed to functions and returned from functions. They are defined using the
keyword struct.

Section 13.1: Flexible Array Members

Version ≥ C99
Type Declaration

A structure with at least one member may additionally contain a single array member of unspecified length at the
end of the structure. This is called a flexible array member:

struct ex1
{
size_t foo;
int flex[];
};

struct ex2_header
{
int foo;
char bar;
};

struct ex2
{
struct ex2_header hdr;
int flex[];
};

/* Merged ex2_header and ex2 structures. */

struct ex3
{
int foo;
char bar;
int flex[];
};

Effects on Size and Padding

A flexible array member is treated as having no size when calculating the size of a structure, though padding
between that member and the previous member of the structure may still exist:

/* Prints "8,8" on my machine, so there is no padding. */

printf("%zu,%zu\n", sizeof(size_t), sizeof(struct ex1));

/* Also prints "8,8" on my machine, so there is no padding in the ex2 structure itself. */
printf("%zu,%zu\n", sizeof(struct ex2_header), sizeof(struct ex2));

/* Prints "5,8" on my machine, so there are 3 bytes of padding. */

printf("%zu,%zu\n", sizeof(int) + sizeof(char), sizeof(struct ex3));

The flexible array member is considered to have an incomplete array type, so its size cannot be calculated using
sizeof.

GoalKicker.com – C Notes for Professionals 83

Usage

You can declare and initialize an object with a structure type containing a flexible array member, but you must not
attempt to initialize the flexible array member since it is treated as if it does not exist. It is forbidden to try to do
this, and compile errors will result.

Similarly, you should not attempt to assign a value to any element of a flexible array member when declaring a
structure in this way since there may not be enough padding at the end of the structure to allow for any objects
required by the flexible array member. The compiler will not necessarily prevent you from doing this, however, so
this can lead to undefined behavior.

/* invalid: cannot initialize flexible array member */

struct ex1 e1 = {1, {2, 3}};
/* invalid: hdr={foo=1, bar=2} OK, but cannot initialize flexible array member */
struct ex2 e2 = {{1, 2}, {3}};
/* valid: initialize foo=1, bar=2 members */
struct ex3 e3 = {1, 2};

e1.flex[0] = 3; /* undefined behavior, in my case */

e3.flex[0] = 2; /* undefined behavior again */
e2.flex[0] = e3.flex[0]; /* undefined behavior */

You may instead choose to use malloc, calloc, or realloc to allocate the structure with extra storage and later
free it, which allows you to use the flexible array member as you wish:

/* valid: allocate an object of structure type `ex1` along with an array of 2 ints */
struct ex1 *pe1 = malloc(sizeof(*pe1) + 2 * sizeof(pe1->flex[0]));

/* valid: allocate an object of structure type ex2 along with an array of 4 ints */
struct ex2 *pe2 = malloc(sizeof(struct ex2) + sizeof(int[4]));

/* valid: allocate 5 structure type ex3 objects along with an array of 3 ints per object */
struct ex3 *pe3 = malloc(5 * (sizeof(*pe3) + sizeof(int[3])));

pe1->flex[0] = 3; /* valid */
pe3[0]->flex[0] = pe1->flex[0]; /* valid */
Version < C99
The 'struct hack'

Flexible array members did not exist prior to C99 and are treated as errors. A common workaround is to declare an
array of length 1, a technique called the 'struct hack':

struct ex1
{
size_t foo;
int flex[1];
};

This will affect the size of the structure, however, unlike a true flexible array member:

/* Prints "8,4,16" on my machine, signifying that there are 4 bytes of padding. */

printf("%d,%d,%d\n", (int)sizeof(size_t), (int)sizeof(int[1]), (int)sizeof(struct ex1));

To use the flex member as a flexible array member, you'd allocate it with malloc as shown above, except that
sizeof(*pe1) (or the equivalent sizeof(struct ex1)) would be replaced with offsetof(struct ex1, flex) or the
longer, type-agnostic expression sizeof(*pe1)-sizeof(pe1->flex). Alternatively, you might subtract 1 from the
desired length of the "flexible" array since it's already included in the structure size, assuming the desired length is

GoalKicker.com – C Notes for Professionals 84

greater than 0. The same logic may be applied to the other usage examples.

Compatibility

If compatibility with compilers that do not support flexible array members is desired, you may use a macro defined
like FLEXMEMB_SIZE below:

#if __STDC_VERSION__ < 199901L

#define FLEXMEMB_SIZE 1
#else
#define FLEXMEMB_SIZE /* nothing */
#endif

struct ex1
{
size_t foo;
int flex[FLEXMEMB_SIZE];
};

When allocating objects, you should use the offsetof(struct ex1, flex) form to refer to the structure size
(excluding the flexible array member) since it is the only expression that will remain consistent between compilers
that support flexible array members and compilers that do not:

struct ex1 *pe10 = malloc(offsetof(struct ex1, flex) + n * sizeof(pe10->flex[0]));

The alternative is to use the preprocessor to conditionally subtract 1 from the specified length. Due to the increased
potential for inconsistency and general human error in this form, I moved the logic into a separate function:

struct ex1 *ex1_alloc(size_t n)

{
struct ex1 tmp;
#if __STDC_VERSION__ < 199901L
if (n != 0)
n--;
#endif
return malloc(sizeof(tmp) + n * sizeof(tmp.flex[0]));
}
...

/* allocate an ex1 object with "flex" array of length 3 */

struct ex1 *pe1 = ex1_alloc(3);

Section 13.2: Typedef Structs

Combining typedef with struct can make code clearer. For example:

typedef struct
{
int x, y;
} Point;

as opposed to:

struct Point
{
int x, y;
};

GoalKicker.com – C Notes for Professionals 85

could be declared as:

Point point;

instead of:

struct Point point;

Even better is to use the following

typedef struct Point Point;

struct Point
{
int x, y;
};

to have advantage of both possible definitions of point. Such a declaration is most convenient if you learned C++
first, where you may omit the struct keyword if the name is not ambiguous.

typedef names for structs could be in conflict with other identifiers of other parts of the program. Some consider
this a disadvantage, but for most people having a struct and another identifier the same is quite disturbing.
Notorious is e.g POSIX' stat

int stat(const char *pathname, struct stat *buf);

where you see a function stat that has one argument that is struct stat.

typedef'd structs without a tag name always impose that the whole struct declaration is visible to code that uses
it. The entire struct declaration must then be placed in a header file.

Consider:

#include "bar.h"

struct foo
{
bar *aBar;
};

So with a typedefd struct that has no tag name, the bar.h file always has to include the whole definition of bar. If
we use

typedef struct bar bar;

in bar.h, the details of the bar structure can be hidden.

See Typedef

Section 13.3: Pointers to structs

When you have a variable containing a struct, you can access its fields using the dot operator (.). However, if you
have a pointer to a struct, this will not work. You have to use the arrow operator (->) to access its fields. Here's an
example of a terribly simple (some might say "terrible and simple") implementation of a stack that uses pointers to
structs and demonstrates the arrow operator.

GoalKicker.com – C Notes for Professionals 86

#include <stdlib.h>
#include <stdio.h>

/* structs */
struct stack
{
struct node *top;
int size;
};

struct node
{
int data;
struct node *next;
};

/* function declarations */
int push(int, struct stack*);
int pop(struct stack*);
void destroy(struct stack*);

int main(void)
{
int result = EXIT_SUCCESS;

size_t i;

/* allocate memory for a struct stack and record its pointer */

struct stack *stack = malloc(sizeof *stack);
if (NULL == stack)
{
perror("malloc() failed");
return EXIT_FAILURE;
}

/* initialize stack */
stack->top = NULL;
stack->size = 0;

/* push 10 ints */
{
int data = 0;
for(i = 0; i < 10; i++)
{
printf("Pushing: %d\n", data);
if (-1 == push(data, stack))
{
perror("push() failed");
result = EXIT_FAILURE;
break;
}

++data;
}
}

if (EXIT_SUCCESS == result)
{
/* pop 5 ints */
for(i = 0; i < 5; i++)
{
printf("Popped: %i\n", pop(stack));

GoalKicker.com – C Notes for Professionals 87

 }
}

/* destroy stack */
destroy(stack);

return result;
}

/* Push a value onto the stack. */

/* Returns 0 on success and -1 on failure. */
int push(int data, struct stack *stack)
{
int result = 0;

/* allocate memory for new node */

struct node *new_node = malloc(sizeof *new_node);
if (NULL == new_node)
{
result = -1;
}
else
{
new_node->data = data;
new_node->next = stack->top;
stack->top = new_node;
stack->size++;
}

return result;
}

/* Pop a value off of the stack. */

/* Returns the value popped off the stack */
int pop(struct stack *stack)
{
struct node *top = stack->top;
int data = top->data;
stack->top = top->next;
stack->size--;
free(top);
return data;
}

/* destroy the stack */

void destroy(struct stack *stack)
{
/* free all pointers */
while(stack->top != NULL)
{
pop(stack);
}
}

Section 13.4: Passing structs to functions

In C, all arguments are passed to functions by value, including structs. For small structs, this is a good thing as it
means there is no overhead from accessing the data through a pointer. However, it also makes it very easy to
accidentally pass a huge struct resulting in poor performance, particularly if the programmer is used to other
languages where arguments are passed by reference.

GoalKicker.com – C Notes for Professionals 88

struct coordinates
{
int x;
int y;
int z;
};

// Passing and returning a small struct by value, very fast

struct coordinates move(struct coordinates position, struct coordinates movement)
{
position.x += movement.x;
position.y += movement.y;
position.z += movement.z;
return position;
}

// A very big struct

struct lotsOfData
{
int param1;
char param2[80000];
};

// Passing and returning a large struct by value, very slow!

// Given the large size of the struct this could even cause stack overflow
struct lotsOfData doubleParam1(struct lotsOfData value)
{
value.param1 *= 2;
return value;
}

// Passing the large struct by pointer instead, fairly fast

void doubleParam1ByPtr(struct lotsOfData *value)
{
value->param1 *= 2;
}

Section 13.5: Object-based programming using structs

Structs may be used to implement code in an object oriented manner. A struct is similar to a class, but is missing
the functions which normally also form part of a class, we can add these as function pointer member variables. To
stay with our coordinates example:

/* coordinates.h */

typedef struct coordinate_s

{
/* Pointers to method functions */
void (*setx)(coordinate *this, int x);
void (*sety)(coordinate *this, int y);
void (*print)(coordinate *this);
/* Data */
int x;
int y;
} coordinate;

/* Constructor */
coordinate *coordinate_create(void);
/* Destructor */
void coordinate_destroy(coordinate *this);

GoalKicker.com – C Notes for Professionals 89

And now the implementing C file:

/* coordinates.c */

#include "coordinates.h"
#include <stdio.h>
#include <stdlib.h>

/* Constructor */
coordinate *coordinate_create(void)
{
coordinate *c = malloc(sizeof(*c));
if (c != 0)
{
c->setx = &coordinate_setx;
c->sety = &coordinate_sety;
c->print = &coordinate_print;
c->x = 0;
c->y = 0;
}
return c;
}

/* Destructor */
void coordinate_destroy(coordinate *this)
{
if (this != NULL)
{
free(this);
}
}

/* Methods */
static void coordinate_setx(coordinate *this, int x)
{
if (this != NULL)
{
this->x = x;
}
}

static void coordinate_sety(coordinate *this, int y)

{
if (this != NULL)
{
this->y = y;
}
}

static void coordinate_print(coordinate *this)

{
if (this != NULL)
{
printf("Coordinate: (%i, %i)\n", this->x, this->y);
}
else
{
printf("NULL pointer exception!\n");
}
}

GoalKicker.com – C Notes for Professionals 90

An example usage of our coordinate class would be:

/* main.c */

#include "coordinates.h"
#include <stddef.h>

int main(void)
{
/* Create and initialize pointers to coordinate objects */
coordinate *c1 = coordinate_create();
coordinate *c2 = coordinate_create();

/* Now we can use our objects using our methods and passing the object as parameter */
c1->setx(c1, 1);
c1->sety(c1, 2);

c2->setx(c2, 3);
c2->sety(c2, 4);

c1->print(c1);
c2->print(c2);

/* After using our objects we destroy them using our "destructor" function */
coordinate_destroy(c1);
c1 = NULL;
coordinate_destroy(c2);
c2 = NULL;

return 0;
}

Section 13.6: Simple data structures

Structure data types are useful way to package related data and have them behave like a single variable.

Declaring a simple struct that holds two int members:

struct point
{
int x;
int y;
};

x and y are called the members (or fields) of point struct.

Defining and using structs:

struct point p; // declare p as a point struct

p.x = 5; // assign p member variables
p.y = 3;

Structs can be initialized at definition. The above is equivalent to:

struct point p = {5, 3};

Structs may also be initialized using designated initializers.

GoalKicker.com – C Notes for Professionals 91

Accessing fields is also done using the . operator

printf("point is (x = %d, y = %d)", p.x, p.y);

GoalKicker.com – C Notes for Professionals 92

Chapter 14: Standard Math
Section 14.1: Power functions - pow(), powf(), powl()
The following example code computes the sum of 1+4(3+3^2+3^3+3^4+...+3^N) series using pow() family of standard
math library.

#include <stdio.h>
#include <math.h>
#include <errno.h>
#include <fenv.h>

int main()
{
double pwr, sum=0;
int i, n;

printf("\n1+4(3+3^2+3^3+3^4+...+3^N)=?\nEnter N:");
scanf("%d",&n);
if (n<=0) {
printf("Invalid power N=%d", n);
return -1;
}

for (i=0; i<n+1; i++) {

errno = 0;
feclearexcept(FE_ALL_EXCEPT);
pwr = powl(3,i);
if (fetestexcept(FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW |
FE_UNDERFLOW)) {
perror("Math Error");
}
sum += i ? pwr : 0;
printf("N= %d\tS= %g\n", i, 1+4*sum);
}

return 0;
}

Example Output:

1+4(3+3^2+3^3+3^4+...+3^N)=?
Enter N:10
N= 0 S= 1
N= 1 S= 13
N= 2 S= 49
N= 3 S= 157
N= 4 S= 481
N= 5 S= 1453
N= 6 S= 4369
N= 7 S= 13117
N= 8 S= 39361
N= 9 S= 118093
N= 10 S= 354289

GoalKicker.com – C Notes for Professionals 93

Section 14.2: Double precision floating-point remainder:
fmod()
This function returns the floating-point remainder of the division of x/y. The returned value has the same sign as x.

#include <math.h> /* for fmod() */

#include <stdio.h> /* for printf() */

int main(void)
{
double x = 10.0;
double y = 5.1;

double modulus = fmod(x, y);

printf("%lf\n", modulus); /* f is the same as lf. */

return 0;
}

Output:

4.90000

Important: Use this function with care, as it can return unexpected values due to the operation of floating point
values.

#include <math.h>
#include <stdio.h>

int main(void)
{
printf("%f\n", fmod(1, 0.1));
printf("%19.17f\n", fmod(1, 0.1));
return 0;
}

Output:

0.1
0.09999999999999995

Section 14.3: Single precision and long double precision

floating-point remainder: fmodf(), fmodl()
Version ≥ C99

These functions returns the floating-point remainder of the division of x/y. The returned value has the same sign as
x.

Single Precision:

#include <math.h> /* for fmodf() */

#include <stdio.h> /* for printf() */

GoalKicker.com – C Notes for Professionals 94

int main(void)
{
float x = 10.0;
float y = 5.1;

float modulus = fmodf(x, y);

printf("%f\n", modulus); /* lf would do as well as modulus gets promoted to double. */

}

Output:

4.90000

Double Double Precision:

#include <math.h> /* for fmodl() */

#include <stdio.h> /* for printf() */

int main(void)
{
long double x = 10.0;
long double y = 5.1;

long double modulus = fmodl(x, y);

printf("%Lf\n", modulus); /* Lf is for long double. */

}

Output:

4.90000

GoalKicker.com – C Notes for Professionals 95

Chapter 15: Iteration Statements/Loops:
for, while, do-while
Section 15.1: For loop
In order to execute a block of code over an over again, loops comes into the picture. The for loop is to be used
when a block of code is to executed a fixed number of times. For example, in order to fill an array of size n with the
user inputs, we need to execute scanf() for n times.

Version ≥ C99
#include <stddef.h> // for size_t

int array[10]; // array of 10 int

for (size_t i = 0; i < 10; i++) // i starts at 0 and finishes with 9

{
scanf("%d", &array[i]);
}

In this way the scanf() function call is executed n times (10 times in our example), but is written only once.

Here, the variable i is the loop index, and it is best declared as presented. The type size_t (size type) should be
used for everything that counts or loops through data objects.

This way of declaring variables inside the for is only available for compilers that have been updated to the C99
standard. If for some reason you are still stuck with an older compiler you can declare the loop index before the
for loop:

Version < C99

#include <stddef.h> /* for size_t */
size_t i;
int array[10]; /* array of 10 int */

for (i = 0; i < 10; i++) /* i starts at 0 and finishes at 9 */

{
scanf("%d", &array[i]);
}

Section 15.2: Loop Unrolling and Du's Device

Sometimes, the straight forward loop cannot be entirely contained within the loop body. This is because, the loop
needs to be primed by some statements B. Then, the iteration begins with some statements A, which are then
followed by B again before looping.

do_B();
while (condition) {
do_A();
do_B();
}

To avoid potential cut/paste problems with repeating B twice in the code, Duff's Device could be applied to start the
loop from the middle of the while body, using a switch statement and fall through behavior.

switch (true) while (condition) {

case false: do_A(); /* FALL THROUGH */

GoalKicker.com – C Notes for Professionals 96

default: do_B(); /* FALL THROUGH */
}

Duff's Device was actually invented to implement loop unrolling. Imagine applying a mask to a block of memory,
where n is a signed integral type with a positive value.

do {
*ptr++ ^= mask;
} while (--n > 0);

If n were always divisible by 4, you could unroll this easily as:

do {
*ptr++ ^= mask;
*ptr++ ^= mask;
*ptr++ ^= mask;
*ptr++ ^= mask;
} while ((n -= 4) > 0);

But, with Duff's Device, the code can follow this unrolling idiom that jumps into the right place in the middle of the
loop if n is not divisible by 4.

switch (n % 4) do {
case 0: *ptr++ ^= mask; /* FALL THROUGH */
case 3: *ptr++ ^= mask; /* FALL THROUGH */
case 2: *ptr++ ^= mask; /* FALL THROUGH */
case 1: *ptr++ ^= mask; /* FALL THROUGH */
} while ((n -= 4) > 0);

This kind of manual unrolling is rarely required with modern compilers, since the compiler's optimization engine
can unroll loops on the programmer's behalf.

Section 15.3: While loop

A while loop is used to execute a piece of code while a condition is true. The while loop is to be used when a block
of code is to be executed a variable number of times. For example the code shown gets the user input, as long as
the user inserts numbers which are not 0. If the user inserts 0, the while condition is not true anymore so execution
will exit the loop and continue on to any subsequent code:

int num = 1;

while (num != 0)
{
scanf("%d", &num);
}

Section 15.4: Do-While loop

Unlike for and while loops, do-while loops check the truth of the condition at the end of the loop, which means
the do block will execute once, and then check the condition of the while at the bottom of the block. Meaning that a
do-while loop will always run at least once.

For example this do-while loop will get numbers from user, until the sum of these values is greater than or equal to
50:

GoalKicker.com – C Notes for Professionals 97

int num, sum;
num = sum = 0;

do
{
scanf("%d", &num);
sum += num;

} while (sum < 50);

do-while loops are relatively rare in most programming styles.

Section 15.5: Structure and flow of control in a for loop

for ([declaration-or-expression]; [expression2]; [expression3])
{
/* body of the loop */
}

In a for loop, the loop condition has three expressions, all optional.

The first expression, declaration-or-expression, initializes the loop. It is executed exactly once at the
beginning of the loop.

Version ≥ C99

It can be either a declaration and initialization of a loop variable, or a general expression. If it is a declaration, the
scope of the declared variable is restricted by the for statement.

Version < C99

Historical versions of C only allowed an expression, here, and the declaration of a loop variable had to be placed
before the for.

The second expression, expression2, is the test condition. It is first executed after the initialization. If the
condition is true, then the control enters the body of the loop. If not, it shifts to outside the body of the loop
at the end of the loop. Subsequently, this conditon is checked after each execution of the body as well as the
update statement. When true, the control moves back to the beginning of the body of the loop. The
condition is usually intended to be a check on the number of times the body of the loop executes. This is the
primary way of exiting a loop, the other way being using jump statements.
The third expression, expression3, is the update statement. It is executed after each execution of the body of
the loop. It is often used to increment a variable keeping count of the number of times the loop body has
executed, and this variable is called an iterator.

Each instance of execution of the loop body is called an iteration.

Example:

Version ≥ C99
for(int i = 0; i < 10 ; i++)
{
printf("%d", i);
}

The output is:

GoalKicker.com – C Notes for Professionals 98

0123456789

In the above example, first i = 0 is executed, initializing i. Then, the condition i < 10 is checked, which evaluates
to be true. The control enters the body of the loop and the value of i is printed. Then, the control shifts to i++,
updating the value of i from 0 to 1. Then, the condition is again checked, and the process continues. This goes on
till the value of i becomes 10. Then, the condition i < 10 evaluates false, after which the control moves out of the
loop.

Section 15.6: Infinite Loops

A loop is said to be an infinite loop if the control enters but never leaves the body of the loop. This happens when
the test condition of the loop never evaluates to false.

Example:

Version ≥ C99
for (int i = 0; i >= 0; )
{
/* body of the loop where i is not changed*/
}

In the above example, the variable i, the iterator, is initialized to 0. The test condition is initially true. However, i is
not modified anywhere in the body and the update expression is empty. Hence, i will remain 0, and the test
condition will never evaluate to false, leading to an infinite loop.

Assuming that there are no jump statements, another way an infinite loop might be formed is by explicitly keeping
the condition true:

while (true)
{
/* body of the loop */
}

In a for loop, the condition statement optional. In this case, the condition is always true vacuously, leading to an
infinite loop.

for (;;)
{
/* body of the loop */
}

However, in certain cases, the condition might be kept true intentionally, with the intention of exiting the loop
using a jump statement such as break.

while (true)
{
/* statements */
if (condition)
{
/* more statements */
break;
}
}

GoalKicker.com – C Notes for Professionals 99

Chapter 16: Selection Statements
Section 16.1: if () Statements
One of the simplest ways to control program flow is by using if selection statements. Whether a block of code is to
be executed or not to be executed can be decided by this statement.

The syntax for if selection statement in C could be as follows:

if(cond)
{
statement(s); /*to be executed, on condition being true*/
}

For example,

if (a > 1) {
puts("a is larger than 1");
}

Where a > 1 is a condition that has to evaluate to true in order to execute the statements inside the if block. In
this example "a is larger than 1" is only printed if a > 1 is true.

if selection statements can omit the wrapping braces { and } if there is only one statement within the block. The
above example can be rewritten to

if (a > 1)
puts("a is larger than 1");

However for executing multiple statements within block the braces have to used.

The condition for if can include multiple expressions. if will only perform the action if the end result of expression
is true.

For example

if ((a > 1) && (b > 1)) {

puts("a is larger than 1");
a++;
}

will only execute the printf and a++ if both a and b are greater than 1.

Section 16.2: Nested if()...else VS if()..else Ladder

Nested if()...else statements take more execution time (they are slower) in comparison to an if()...else
ladder because the nested if()...else statements check all the inner conditional statements once the outer
conditional if() statement is satisfied, whereas the if()..else ladder will stop condition testing once any of the
if() or the else if() conditional statements are true.

An if()...else ladder:

#include <stdio.h>

GoalKicker.com – C Notes for Professionals 100

int main(int argc, char *argv[])
{
int a, b, c;
printf("\nEnter Three numbers = ");
scanf("%d%d%d", &a, &b, &c);
if ((a < b) && (a < c))
{
printf("\na = %d is the smallest.", a);
}
else if ((b < a) && (b < c))
{
printf("\nb = %d is the smallest.", b);
}
else if ((c < a) && (c < b))
{
printf("\nc = %d is the smallest.", c);
}
else
{
printf("\nImprove your coding logic");
}
return 0;
}

Is, in the general case, considered to be better than the equivalent nested if()...else:

#include <stdio.h>

int main(int argc, char *argv[])

{
int a, b, c;
printf("\nEnter Three numbers = ");
scanf("%d%d%d", &a, &b, &c);
if (a < b)
{
if (a < c)
{
printf("\na = %d is the smallest.", a);
}
else
{
printf("\nc = %d is the smallest.", c);
}
}
else
{
if(b < c)
{
printf("\nb = %d is the smallest.", b);
}
else
{
printf("\nc = %d is the smallest.", c);
}
}
return 0;
}

GoalKicker.com – C Notes for Professionals 101

Section 16.3: switch () Statements
switch statements are useful when you want to have your program do many different things according to the value
of a particular test variable.

An example usage of switch statement is like this:

int a = 1;

switch (a) {
case 1:
puts("a is 1");
break;
case 2:
puts("a is 2");
break;
default:
puts("a is neither 1 nor 2");
break;
}

This example is equivalent to

int a = 1;

if (a == 1) {
puts("a is 1");
} else if (a == 2) {
puts("a is 2");
} else {
puts("a is neither 1 nor 2");
}

If the value of a is 1 when the switch statement is used, a is 1 will be printed. If the value of a is 2 then, a is 2 will
be printed. Otherwise, a is neither 1 nor 2 will be printed.

case n: is used to describe where the execution flow will jump in when the value passed to switch statement is n.
n must be compile-time constant and the same n can exist at most once in one switch statement.

default: is used to describe that when the value didn't match any of the choices for case n:. It is a good practice
to include a default case in every switch statement to catch unexpected behavior.

A break; statement is required to jump out of the switch block.

Note: If you accidentally forget to add a break after the end of a case, the compiler will assume that you intend to
"fall through" and all the subsequent case statements, if any, will be executed (unless a break statement is found in
any of the subsequent cases), regardless of whether the subsequent case statement(s) match or not. This particular
property is used to implement Duff's Device. This behavior is often considered a flaw in the C language
specification.

Below is an example that shows effects of the absence of break;:

int a = 1;

switch (a) {
case 1:
case 2:

GoalKicker.com – C Notes for Professionals 102

 puts("a is 1 or 2");
case 3:
puts("a is 1, 2 or 3");
break;
default:
puts("a is neither 1, 2 nor 3");
break;
}

When the value of a is 1 or 2, a is 1 or 2 and a is 1, 2 or 3 will both be printed. When a is 3, only a is 1, 2

or 3 will be printed. Otherwise, a is neither 1, 2 nor 3 will be printed.

Note that the default case is not necessary, especially when the set of values you get in the switch is finished and
known at compile time.

The best example is using a switch on an enum.

enum msg_type { ACK, PING, ERROR };

void f(enum msg_type t)
{
switch (t) {
case ACK:
// do nothing
break;
case PING:
// do something
break;
case ERROR:
// do something else
break;
}
}

There are multiple advantages of doing this:

most compilers will report a warning if you don't handle a value (this would not be reported if a default case
were present)
for the same reason, if you add a new value to the enum, you will be notified of all the places where you forgot
to handle the new value (with a default case, you would need to manually explore your code searching for
such cases)
The reader does not need to figure out "what is hidden by the default:", whether there other enum values or
whether it is a protection for "just in case". And if there are other enum values, did the coder intentionally use
the default case for them or is there a bug that was introduced when he added the value?
handling each enum value makes the code self explanatory as you can't hide behind a wild card, you must
explicitly handle each of them.

Nevertheless, you can't prevent someone to write evil code like:

enum msg_type t = (enum msg_type)666; // I'm evil

Thus you may add an extra check before your switch to detect it, if you really need it.

void f(enum msg_type t)

{
if (!is_msg_type_valid(t)) {
// Handle this unlikely error
}

GoalKicker.com – C Notes for Professionals 103

 switch(t) {
// Same code than before
}
}

Section 16.4: if () ... else statements and syntax

While if performs an action only when its condition evaluate to true, if / else allows you to specify the different
actions when the condition true and when the condition is false.

Example:

if (a > 1)
puts("a is larger than 1");
else
puts("a is not larger than 1");

Just like the if statement, when the block within if or else is consisting of only one statement, then the braces can
be omitted (but doing so is not recommended as it can easily introduce problems involuntarily). However if there's
more than one statement within the if or else block, then the braces have to be used on that particular block.

if (a > 1)
{
puts("a is larger than 1");
a--;
}
else
{
puts("a is not larger than 1");
a++;
}

Section 16.5: if()...else Ladder Chaining two or more if () ... else

statements
While the if ()... else statement allows to define only one (default) behaviour which occurs when the condition
within the if () is not met, chaining two or more if () ... else statements allow to define a couple more
behaviours before going to the last else branch acting as a "default", if any.

Example:

int a = ... /* initialise to some value. */

if (a >= 1)
{
printf("a is greater than or equals 1.\n");
}
else if (a == 0) //we already know that a is smaller than 1
{
printf("a equals 0.\n");
}
else /* a is smaller than 1 and not equals 0, hence: */
{
printf("a is negative.\n");
}

GoalKicker.com – C Notes for Professionals 104

Chapter 17: Initialization
Section 17.1: Initialization of Variables in C
In the absence of explicit initialization, external and static variables are guaranteed to be initialized to zero;
automatic variables (including register variables) have indeterminate1 (i.e., garbage) initial values.

Scalar variables may be initialized when they are defined by following the name with an equals sign and an
expression:

int x = 1;
char squota = '\'';
long day = 1000L * 60L * 60L * 24L; /* milliseconds/day */

For external and static variables, the initializer must be a constant expression2; the initialization is done once,
conceptually before the program begins execution.

For automatic and register variables, the initializer is not restricted to being a constant: it may be any expression
involving previously defined values, even function calls.

For example, see the code snippet below

int binsearch(int x, int v[], int n)

{
int low = 0;
int high = n - 1;
int mid;
...
}

instead of

int low, high, mid;

low = 0;
high = n - 1;

In effect, initialization of automatic variables are just shorthand for assignment statements. Which form to prefer is
largely a matter of taste. We generally use explicit assignments, because initializers in declarations are harder to
see and further away from the point of use. On the other hand, variables should only be declared when they're
about to be used whenever possible.

Initializing an array:

An array may be initialized by following its declaration with a list of initializers enclosed in braces and separated by
commas.

For example, to initialize an array days with the number of days in each month:

int days_of_month[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }

(Note that January is encoded as month zero in this structure.)

When the size of the array is omitted, the compiler will compute the length by counting the initializers, of which
there are 12 in this case.

GoalKicker.com – C Notes for Professionals 105

If there are fewer initializers for an array than the specified size, the others will be zero for all types of variables.

It is an error to have too many initializers. There is no standard way to specify repetition of an initializer — but GCC
has an extension to do so.

Version < C99

In C89/C90 or earlier versions of C, there was no way to initialize an element in the middle of an array without
supplying all the preceding values as well.

Version ≥ C99

With C99 and above, designated initializers allow you to initialize arbitrary elements of an array, leaving any
uninitialized values as zeros.

Initializing Character arrays:

Character arrays are a special case of initialization; a string may be used instead of the braces and commas
notation:

char chr_array[] = "hello";

is a shorthand for the longer but equivalent:

char chr_array[] = { 'h', 'e', 'l', 'l', 'o', '\0' };

In this case, the array size is six (five characters plus the terminating '\0').

1 What happens to a declared, uninitialized variable in C? Does it have a value?

2 Note that a constant expression is defined as something that can be evaluated at compile-time. So, int
global_var = f(); is invalid. Another common misconception is thinking of a const qualified variable as a constant
expression. In C, const means "read-only", not "compile time constant". So, global definitions like const int SIZE =
10; int global_arr[SIZE]; and const int SIZE = 10; int global_var = SIZE; are not legal in C.

Section 17.2: Using designated initializers

Version ≥ C99

C99 introduced the concept of designated initializers. These allow you to specify which elements of an array,
structure or union are to be initialized by the values following.

Designated initializers for array elements

For a simple type like plain int:

int array[] = { [4] = 29, [5] = 31, [17] = 101, [18] = 103, [19] = 107, [20] = 109 };

The term in square brackets, which can be any constant integer expression, specifies which element of the array is
to be initialized by the value of the term after the = sign. Unspecified elements are default initialized, which means
zeros are defined. The example shows the designated initializers in order; they do not have to be in order. The
example shows gaps; those are legitimate. The example doesn't show two different initializations for the same
element; that too is allowed (ISO/IEC 9899:2011, §6.7.9 Initialization, ¶19 The initialization shall occur in initializer list
order, each initializer provided for a particular subobject overriding any previously listed initializer for the same
subobject).

GoalKicker.com – C Notes for Professionals 106

In this example, the size of the array is not defined explicitly, so the maximum index specified in the designated
initializers dictates the size of the array — which would be 21 elements in the example. If the size was defined,
initializing an entry beyond the end of the array would be an error, as usual.

Designated initializers for structures

You can specify which elements of a structure are initialized by using the .element notation:

struct Date
{
int year;
int month;
int day;
};

struct Date us_independence_day = { .day = 4, .month = 7, .year = 1776 };

If elements are not listed, they are default initialized (zeroed).

Designated initializer for unions

You can specify which element of a union is initialize with a designated initializer.

Version = C89

Prior to the C standard, there was no way to initialize a union. The C89/C90 standard allows you to initialize the first
member of a union — so the choice of which member is listed first matters.

struct discriminated_union
{
enum { DU_INT, DU_DOUBLE } discriminant;
union
{
int du_int;
double du_double;
} du;
};

struct discriminated_union du1 = { .discriminant = DU_INT, .du = { .du_int = 1 } };

struct discriminated_union du2 = { .discriminant = DU_DOUBLE, .du = { .du_double = 3.14159 } };
Version ≥ C11

Note that C11 allows you to use anonymous union members inside a structure, so that you don't need the du name
in the previous example:

struct discriminated_union
{
enum { DU_INT, DU_DOUBLE } discriminant;
union
{
int du_int;
double du_double;
};
};

struct discriminated_union du1 = { .discriminant = DU_INT, .du_int = 1 };

struct discriminated_union du2 = { .discriminant = DU_DOUBLE, .du_double = 3.14159 };

Designated initializers for arrays of structures, etc

GoalKicker.com – C Notes for Professionals 107

These constructs can be combined for arrays of structures containing elements that are arrays, etc. Using full sets
of braces ensures that the notation is unambiguous.

typedef struct Date Date; // See earlier in this example

struct date_range
{
Date dr_from;
Date dr_to;
char dr_what[80];
};

struct date_range ranges[] =

{
[3] = { .dr_from = { .year = 1066, .month = 10, .day = 14 },
.dr_to = { .year = 1066, .month = 12, .day = 25 },
.dr_what = "Battle of Hastings to Coronation of William the Conqueror",
},
[2] = { .dr_from = { .month = 7, .day = 4, .year = 1776 },
.dr_to = { .month = 5, .day = 14, .year = 1787 },
.dr_what = "US Declaration of Independence to Constitutional Convention",
}
};

Specifying ranges in array initializers

GCC provides an extension that allows you to specify a range of elements in an array that should be given the same
initializer:

int array[] = { [3 ... 7] = 29, 19 = 107 };

The triple dots need to be separate from the numbers lest one of the dots be interpreted as part of a floating point
number (maximimal munch rule).

Section 17.3: Initializing structures and arrays of structures

Structures and arrays of structures can be initialized by a series of values enclosed in braces, one value per
member of the structure.

struct Date
{
int year;
int month;
int day;
};

struct Date us_independence_day = { 1776, 7, 4 };

struct Date uk_battles[] =

{
{ 1066, 10, 14 }, // Battle of Hastings
{ 1815, 6, 18 }, // Battle of Waterloo
{ 1805, 10, 21 }, // Battle of Trafalgar
};

Note that the array initialization could be written without the interior braces, and in times past (before 1990, say)
often would have been written without them:

GoalKicker.com – C Notes for Professionals 108

struct Date uk_battles[] =
{
1066, 10, 14, // Battle of Hastings
1815, 6, 18, // Battle of Waterloo
1805, 10, 21, // Battle of Trafalgar
};

Although this works, it is not good modern style — you should not attempt to use this notation in new code and
should fix the compiler warnings it usually yields.

See also designated initializers.

GoalKicker.com – C Notes for Professionals 109

Chapter 18: Declaration vs Definition
Section 18.1: Understanding Declaration and Definition
A declaration introduces an identifier and describes its type, be it a type, object, or function. A declaration is what
the compiler needs to accept references to that identifier. These are declarations:

extern int bar;

extern int g(int, int);
double f(int, double); /* extern can be omitted for function declarations */
double h1(); /* declaration without prototype */
double h2(); /* ditto */

A definition actually instantiates/implements this identifier. It's what the linker needs in order to link references to
those entities. These are definitions corresponding to the above declarations:

int bar;
int g(int lhs, int rhs) {return lhs*rhs;}
double f(int i, double d) {return i+d;}
double h1(int a, int b) {return -1.5;}
double h2() {} /* prototype is implied in definition, same as double h2(void) */

A definition can be used in the place of a declaration.

However, it must be defined exactly once. If you forget to define something that's been declared and referenced
somewhere, then the linker doesn't know what to link references to and complains about a missing symbols. If you
define something more than once, then the linker doesn't know which of the definitions to link references to and
complains about duplicated symbols.

Exception:

extern int i = 0; /* defines i */

extern int j; /* declares j */

This exception can be explained using concepts of "Strong symbols vs Weak symbols" (from a linker's perspective) .
Please look here ( Slide 22 ) for more explanation.

/* All are definitions. */

struct S { int a; int b; }; /* defines S */
struct X { /* defines X */
int x; /* defines non-static data member x */
};
struct X anX; /* defines anX */

GoalKicker.com – C Notes for Professionals 110

Chapter 19: Command-line arguments
Parameter Details
argument count - initialized to the number of space-separated arguments given to the program from
argc
the command-line as well as the program name itself.
argument vector - initialized to an array of char-pointers (strings) containing the arguments (and the
argv
program name) that was given on the command-line.

Section 19.1: Print the arguments to a program and convert to

integer values
The following code will print the arguments to the program, and the code will attempt to convert each argument
into a number (to a long):

#include <stdlib.h>
#include <stdio.h>
#include <errno.h>
#include <limits.h>

int main(int argc, char* argv[]) {

for (int i = 1; i < argc; i++) {

printf("Argument %d is: %s\n", i, argv[i]);

errno = 0;
char *p;
long argument_numValue = strtol(argv[i], &p, 10);

if (p == argv[i]) {
fprintf(stderr, "Argument %d is not a number.\n", i);
}
else if ((argument_numValue == LONG_MIN || argument_numValue == LONG_MAX) && errno ==
ERANGE) {
fprintf(stderr, "Argument %d is out of range.\n", i);
}
else {
printf("Argument %d is a number, and the value is: %ld\n",
i, argument_numValue);
}
}
return 0;
}

References:

strtol() returns an incorrect value

Correct usage of strtol

Section 19.2: Printing the command line arguments

After receiving the arguments, you can print them as follows:

int main(int argc, char **argv)

{
for (int i = 1; i < argc; i++)
{
printf("Argument %d: [%s]\n", i, argv[i]);

GoalKicker.com – C Notes for Professionals 111

 }
}

Notes

1. The argv parameter can be also defined as char *argv[].

2. argv[0] may contain the program name itself (depending on how the program was executed). The first "real"
command line argument is at argv[1], and this is the reason why the loop variable i is initialized to 1.
3. In the print statement, you can use *(argv + i) instead of argv[i] - it evaluates to the same thing, but is
more verbose.
4. The square brackets around the argument value help identify the start and end. This can be invaluable if
there are trailing blanks, newlines, carriage returns, or other oddball characters in the argument. Some
variant on this program is a useful tool for debugging shell scripts where you need to understand what the
argument list actually contains (although there are simple shell alternatives that are almost equivalent).

Section 19.3: Using GNU getopt tools

Command-line options for applications are not treated any differently from command-line arguments by the C
language. They are just arguments which, in a Linux or Unix environment, traditionally begin with a dash (-).

With glibc in a Linux or Unix environment you can use the getopt tools to easily define, validate, and parse
command-line options from the rest of your arguments.

These tools expect your options to be formatted according to the GNU Coding Standards, which is an extension of
what POSIX specifies for the format of command-line options.

The example below demonstrates handling command-line options with the GNU getopt tools.

#include <stdio.h>
#include <getopt.h>
#include <string.h>

/* print a description of all supported options */

void usage (FILE *fp, const char *path)
{
/* take only the last portion of the path */
const char *basename = strrchr(path, '/');
basename = basename ? basename + 1 : path;

fprintf (fp, "usage: %s [OPTION]\n", basename);

fprintf (fp, " -h, --help\t\t"
"Print this help and exit.\n");
fprintf (fp, " -f, --file[=FILENAME]\t"
"Write all output to a file (defaults to out.txt).\n");
fprintf (fp, " -m, --msg=STRING\t"
"Output a particular message rather than 'Hello world'.\n");
}

/* parse command-line options and print message */

int main(int argc, char *argv[])
{
/* for code brevity this example just uses fixed buffer sizes for strings */
char filename[256] = { 0 };
char message[256] = "Hello world";
FILE *fp;
int help_flag = 0;
int opt;

GoalKicker.com – C Notes for Professionals 112

 /* table of all supported options in their long form.
* fields: name, has_arg, flag, val
* `has_arg` specifies whether the associated long-form option can (or, in
* some cases, must) have an argument. the valid values for `has_arg` are
* `no_argument`, `optional_argument`, and `required_argument`.
* if `flag` points to a variable, then the variable will be given a value
* of `val` when the associated long-form option is present at the command
* line.
* if `flag` is NULL, then `val` is returned by `getopt_long` (see below)
* when the associated long-form option is found amongst the command-line
* arguments.
*/
struct option longopts[] = {
{ "help", no_argument, &help_flag, 1 },
{ "file", optional_argument, NULL, 'f' },
{ "msg", required_argument, NULL, 'm' },
{ 0 }
};

/* infinite loop, to be broken when we are done parsing options */

while (1) {
/* getopt_long supports GNU-style full-word "long" options in addition
* to the single-character "short" options which are supported by
* getopt.
* the third argument is a collection of supported short-form options.
* these do not necessarily have to correlate to the long-form options.
* one colon after an option indicates that it has an argument, two
* indicates that the argument is optional. order is unimportant.
*/
opt = getopt_long (argc, argv, "hf::m:", longopts, 0);

if (opt == -1) {
/* a return value of -1 indicates that there are no more options */
break;
}

switch (opt) {
case 'h':
/* the help_flag and value are specified in the longopts table,
* which means that when the --help option is specified (in its long
* form), the help_flag variable will be automatically set.
* however, the parser for short-form options does not support the
* automatic setting of flags, so we still need this code to set the
* help_flag manually when the -h option is specified.
*/
help_flag = 1;
break;
case 'f':
/* optarg is a global variable in getopt.h. it contains the argument
* for this option. it is null if there was no argument.
*/
printf ("outarg: '%s'\n", optarg);
strncpy (filename, optarg ? optarg : "out.txt", sizeof (filename));
/* strncpy does not fully guarantee null-termination */
filename[sizeof (filename) - 1] = '\0';
break;
case 'm':
/* since the argument for this option is required, getopt guarantees
* that aptarg is non-null.
*/
strncpy (message, optarg, sizeof (message));
message[sizeof (message) - 1] = '\0';

GoalKicker.com – C Notes for Professionals 113

 break;
case '?':
/* a return value of '?' indicates that an option was malformed.
* this could mean that an unrecognized option was given, or that an
* option which requires an argument did not include an argument.
*/
usage (stderr, argv[0]);
return 1;
default:
break;
}
}

if (help_flag) {
usage (stdout, argv[0]);
return 0;
}

if (filename[0]) {
fp = fopen (filename, "w");
} else {
fp = stdout;
}

if (!fp) {
fprintf(stderr, "Failed to open file.\n");
return 1;
}

fprintf (fp, "%s\n", message);

fclose (fp);
return 0;
}

It can be compiled with gcc:

gcc example.c -o example

It supports three command-line options (--help, --file, and --msg). All have a "short form" as well (-h, -f, and -m).
The "file" and "msg" options both accept arguments. If you specify the "msg" option, its argument is required.

Arguments for options are formatted as:

--option=value (for long-form options)

-ovalue or -o"value" (for short-form options)

GoalKicker.com – C Notes for Professionals 114

Chapter 20: Files and I/O streams
Parameter Details
const char *mode A string describing the opening mode of the file-backed stream. See remarks for possible values.
Can be SEEK_SET to set from the beginning of the file, SEEK_END to set from its end, or SEEK_CUR
int whence
to set relative to the current cursor value. Note: SEEK_END is non-portable.

Section 20.1: Open and write to file

#include <stdio.h> /* for perror(), fopen(), fputs() and fclose() */
#include <stdlib.h> /* for the EXIT_* macros */

int main(int argc, char **argv)

{
int e = EXIT_SUCCESS;

/* Get path from argument to main else default to output.txt */

char *path = (argc > 1) ? argv[1] : "output.txt";

/* Open file for writing and obtain file pointer */

FILE *file = fopen(path, "w");

/* Print error message and exit if fopen() failed */

if (!file)
{
perror(path);
return EXIT_FAILURE;
}

/* Writes text to file. Unlike puts(), fputs() does not add a new-line. */
if (fputs("Output in file.\n", file) == EOF)
{
perror(path);
e = EXIT_FAILURE;
}

/* Close file */
if (fclose(file))
{
perror(path);
return EXIT_FAILURE;
}
return e;
}

This program opens the file with name given in the argument to main, defaulting to output.txt if no argument is
given. If a file with the same name already exists, its contents are discarded and the file is treated as a new empty
file. If the files does not already exist the fopen() call creates it.

If the fopen() call fails for some reason, it returns a NULL value and sets the global errno variable value. This means
that the program can test the returned value after the fopen() call and use perror() if fopen() fails.

If the fopen() call succeeds, it returns a valid FILE pointer. This pointer can then be used to reference this file until
fclose() is called on it.

The fputs() function writes the given text to the opened file, replacing any previous contents of the file. Similarly to
fopen(), the fputs() function also sets the errno value if it fails, though in this case the function returns EOF to

GoalKicker.com – C Notes for Professionals 115

indicate the fail (it otherwise returns a non-negative value).

The fclose() function flushes any buffers, closes the file and frees the memory pointed to by FILE *. The return
value indicates completion just as fputs() does (though it returns '0' if successful), again also setting the errno
value in the case of a fail.

Section 20.2: Run process

#include <stdio.h>

void print_all(FILE *stream)

{
int c;
while ((c = getc(stream)) != EOF)
putchar(c);
}
int main(void)
{
FILE *stream;

/* call netstat command. netstat is available for Windows and Linux */

if ((stream = popen("netstat", "r")) == NULL)
return 1;

print_all(stream);
pclose(stream);
return 0;
}

This program runs a process (netstat) via popen() and reads all the standard output from the process and echoes
that to standard output.

Note: popen() does not exist in the standard C library, but it is rather a part of POSIX C)

Section 20.3: fprintf

You can use fprintf on a file just like you might on a console with printf. For example to keep track of game wins,
losses and ties you might write

/* saves wins, losses and, ties */

void savewlt(FILE *fout, int wins, int losses, int ties)
{
fprintf(fout, "Wins: %d\nTies: %d\nLosses: %d\n", wins, ties, losses);
}

A side note: Some systems (infamously, Windows) do not use what most programmers would call "normal" line
endings. While UNIX-like systems use \n to terminate lines, Windows uses a pair of characters: \r (carriage return)
and \n (line feed). This sequence is commonly called CRLF. However, whenever using C, you do not need to worry
about these highly platform-dependent details. A C compiler is required to convert every instance of \n to the
correct platform line ending. So a Windows compiler would convert \n to \r\n, but a UNIX compiler would keep it as-
is.

Section 20.4: Get lines from a file using getline()

The POSIX C library defines the getline() function. This function allocates a buffer to hold the line contents and
returns the new line, the number of characters in the line, and the size of the buffer.

GoalKicker.com – C Notes for Professionals 116

Example program that gets each line from example.txt:

#include <stdlib.h>
#include <stdio.h>

#define FILENAME "example.txt"

int main(void)
{
/* Open the file for reading */
char *line_buf = NULL;
size_t line_buf_size = 0;
int line_count = 0;
ssize_t line_size;
FILE *fp = fopen(FILENAME, "r");
if (!fp)
{
fprintf(stderr, "Error opening file '%s'\n", FILENAME);
return EXIT_FAILURE;
}

/* Get the first line of the file. */

line_size = getline(&line_buf, &line_buf_size, fp);

/* Loop through until we are done with the file. */

while (line_size >= 0)
{
/* Increment our line count */
line_count++;

/* Show the line details */

printf("line[%06d]: chars=%06zd, buf size=%06zu, contents: %s", line_count,
line_size, line_buf_size, line_buf);

/* Get the next line */

line_size = getline(&line_buf, &line_buf_size, fp);
}

/* Free the allocated line buffer */

free(line_buf);
line_buf = NULL;

/* Close the file now that we are done with it */

fclose(fp);

return EXIT_SUCCESS;
}

Input file example.txt

This is a file
which has
multiple lines
with various indentation,
blank lines

a really long line to show that getline() will reallocate the line buffer if the length of a line
is too long to fit in the buffer it has been given,
and punctuation at the end of the lines.

GoalKicker.com – C Notes for Professionals 117

Output

line[000001]: chars=000015, buf size=000016, contents: This is a file

line[000002]: chars=000012, buf size=000016, contents: which has
line[000003]: chars=000015, buf size=000016, contents: multiple lines
line[000004]: chars=000030, buf size=000032, contents: with various indentation,
line[000005]: chars=000012, buf size=000032, contents: blank lines
line[000006]: chars=000001, buf size=000032, contents:
line[000007]: chars=000001, buf size=000032, contents:
line[000008]: chars=000001, buf size=000032, contents:
line[000009]: chars=000150, buf size=000160, contents: a really long line to show that getline()
will reallocate the line buffer if the length of a line is too long to fit in the buffer it has
been given,
line[000010]: chars=000042, buf size=000160, contents: and punctuation at the end of the lines.
line[000011]: chars=000001, buf size=000160, contents:

In the example, getline() is initially called with no buffer allocated. During this first call, getline() allocates a
buffer, reads the first line and places the line's contents in the new buffer. On subsequent calls, getline() updates
the same buffer and only reallocates the buffer when it is no longer large enough to fit the whole line. The
temporary buffer is then freed when we are done with the file.

Another option is getdelim(). This is the same as getline() except you specify the line ending character. This is
only necessary if the last character of the line for your file type is not '\n'. getline() works even with Windows text
files because with the multibyte line ending ("\r\n")'\n'` is still the last character on the line.

Example implementation of getline()

#include <stdlib.h>
#include <stdio.h>
#include <errno.h>
#include <stdint.h>

#if !(defined _POSIX_C_SOURCE)

typedef long int ssize_t;
#endif

/* Only include our version of getline() if the POSIX version isn't available. */

#if !(defined _POSIX_C_SOURCE) || _POSIX_C_SOURCE < 200809L

#if !(defined SSIZE_MAX)

#define SSIZE_MAX (SIZE_MAX >> 1)
#endif

ssize_t getline(char **pline_buf, size_t *pn, FILE *fin)

{
const size_t INITALLOC = 16;
const size_t ALLOCSTEP = 16;
size_t num_read = 0;

/* First check that none of our input pointers are NULL. */

if ((NULL == pline_buf) || (NULL == pn) || (NULL == fin))
{
errno = EINVAL;
return -1;
}

/* If output buffer is NULL, then allocate a buffer. */

if (NULL == *pline_buf)

GoalKicker.com – C Notes for Professionals 118

 {
*pline_buf = malloc(INITALLOC);
if (NULL == *pline_buf)
{
/* Can't allocate memory. */
return -1;
}
else
{
/* Note how big the buffer is at this time. */
*pn = INITALLOC;
}
}

/* Step through the file, pulling characters until either a newline or EOF. */

{
int c;
while (EOF != (c = getc(fin)))
{
/* Note we read a character. */
num_read++;

/* Reallocate the buffer if we need more room */

if (num_read >= *pn)
{
size_t n_realloc = *pn + ALLOCSTEP;
char * tmp = realloc(*pline_buf, n_realloc + 1); /* +1 for the trailing NUL. */
if (NULL != tmp)
{
/* Use the new buffer and note the new buffer size. */
*pline_buf = tmp;
*pn = n_realloc;
}
else
{
/* Exit with error and let the caller free the buffer. */
return -1;
}

/* Test for overflow. */

if (SSIZE_MAX < *pn)
{
errno = ERANGE;
return -1;
}
}

/* Add the character to the buffer. */

(*pline_buf)[num_read - 1] = (char) c;

/* Break from the loop if we hit the ending character. */

if (c == '\n')
{
break;
}
}

/* Note if we hit EOF. */

if (EOF == c)
{
errno = 0;

GoalKicker.com – C Notes for Professionals 119

 return -1;
}
}

/* Terminate the string by suffixing NUL. */

(*pline_buf)[num_read] = '\0';

return (ssize_t) num_read;

}

#endif

Section 20.5: fscanf()

Let's say we have a text file and we want to read all words in that file, in order to do some requirements.

file.txt:

This is just
a test file
to be used by fscanf()

This is the main function:

#include <stdlib.h>
#include <stdio.h>

void printAllWords(FILE *);

int main(void)
{
FILE *fp;

if ((fp = fopen("file.txt", "r")) == NULL) {

perror("Error opening file");
exit(EXIT_FAILURE);
}

printAllWords(fp);

fclose(fp);

return EXIT_SUCCESS;
}

void printAllWords(FILE * fp)

{
char tmp[20];
int i = 1;

while (fscanf(fp, "%19s", tmp) != EOF) {

printf("Word %d: %s\n", i, tmp);
i++;
}
}

The output will be:

GoalKicker.com – C Notes for Professionals 120

Word 1: This
Word 2: is
Word 3: just
Word 4: a
Word 5: test
Word 6: file
Word 7: to
Word 8: be
Word 9: used
Word 10: by
Word 11: fscanf()

Section 20.6: Read lines from a file

The stdio.h header defines the fgets() function. This function reads a line from a stream and stores it in a
specified string. The function stops reading text from the stream when either n - 1 characters are read, the
newline character ('\n') is read or the end of file (EOF) is reached.

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#define MAX_LINE_LENGTH 80

int main(int argc, char **argv)

{
char *path;
char line[MAX_LINE_LENGTH] = {0};
unsigned int line_count = 0;

if (argc < 1)
return EXIT_FAILURE;
path = argv[1];

/* Open file */
FILE *file = fopen(path, "r");

if (!file)
{
perror(path);
return EXIT_FAILURE;
}

/* Get each line until there are none left */

while (fgets(line, MAX_LINE_LENGTH, file))
{
/* Print each line */
printf("line[%06d]: %s", ++line_count, line);

/* Add a trailing newline to lines that don't already have one */

if (line[strlen(line) - 1] != '\n')
printf("\n");
}

/* Close file */
if (fclose(file))
{
return EXIT_FAILURE;
perror(path);

GoalKicker.com – C Notes for Professionals 121

 }
}

Calling the program with an argument that is a path to a file containing the following text:

This is a file
which has
multiple lines
with various indentation,
blank lines

a really long line to show that the line will be counted as two lines if the length of a line is
too long to fit in the buffer it has been given,
and punctuation at the end of the lines.

Will result in the following output:

line[000001]: This is a file

line[000002]: which has
line[000003]: multiple lines
line[000004]: with various indentation,
line[000005]: blank lines
line[000006]:
line[000007]:
line[000008]:
line[000009]: a really long line to show that the line will be counted as two lines if the le
line[000010]: ngth of a line is too long to fit in the buffer it has been given,
line[000011]: and punctuation at the end of the lines.
line[000012]:

This very simple example allows a fixed maximum line length, such that longer lines will effectively be counted as
two lines. The fgets() function requires that the calling code provide the memory to be used as the destination for
the line that is read.

POSIX makes the getline() function available which instead internally allocates memory to enlarge the buffer as
necessary for a line of any length (as long as there is sufficient memory).

Section 20.7: Open and write to a binary file

#include <stdlib.h>
#include <stdio.h>

int main(void)
{
result = EXIT_SUCCESS;

char file_name[] = "outbut.bin";

char str[] = "This is a binary file example";
FILE * fp = fopen(file_name, "wb");

if (fp == NULL) /* If an error occurs during the file creation */

{
result = EXIT_FAILURE;
fprintf(stderr, "fopen() failed for '%s'\n", file_name);

GoalKicker.com – C Notes for Professionals 122

 }
else
{
size_t element_size = sizeof *str;
size_t elements_to_write = sizeof str;

/* Writes str (_including_ the NUL-terminator) to the binary file. */

size_t elements_written = fwrite(str, element_size, elements_to_write, fp);
if (elements_written != elements_to_write)
{
result = EXIT_FAILURE;
/* This works for >=c99 only, else the z length modifier is unknown. */
fprintf(stderr, "fwrite() failed: wrote only %zu out of %zu elements.\n",
elements_written, elements_to_write);
/* Use this for <c99: *
fprintf(stderr, "fwrite() failed: wrote only %lu out of %lu elements.\n",
(unsigned long) elements_written, (unsigned long) elements_to_write);
*/
}

fclose(fp);
}

return result;
}

This program creates and writes text in the binary form through the fwrite function to the file output.bin.

If a file with the same name already exists, its contents are discarded and the file is treated as a new empty file.

A binary stream is an ordered sequence of characters that can transparently record internal data. In this mode,
bytes are written between the program and the file without any interpretation.

To write integers portably, it must be known whether the file format expects them in big or little-endian format, and
the size (usually 16, 32 or 64 bits). Bit shifting and masking may then be used to write out the bytes in the correct
order. Integers in C are not guaranteed to have two's complement representation (though almost all
implementations do). Fortunately a conversion to unsigned is guaranteed to use twos complement. The code for
writing a signed integer to a binary file is therefore a little surprising.

/* write a 16-bit little endian integer */

int fput16le(int x, FILE *fp)
{
unsigned int rep = x;
int e1, e2;

e1 = fputc(rep & 0xFF, fp);

e2 = fputc((rep >> 8) & 0xFF, fp);

if(e1 == EOF || e2 == EOF)

return EOF;
return 0;
}

The other functions follow the same pattern with minor modifications for size and byte order.

GoalKicker.com – C Notes for Professionals 123

Chapter 21: Formatted Input/Output
Section 21.1: Conversion Specifiers for printing
Conversion
Type of Argument Description
Specifier
i, d int prints decimal
u unsigned int prints decimal
o unsigned int prints octal
x unsigned int prints hexadecimal, lower-case
X unsigned int prints hexadecimal, upper-case
prints float with a default precision of 6, if no precision is given (lower-case used
f double
for special numbers nan and inf or infinity)
prints float with a default precision of 6, if no precision is given (upper-case
F double
used for special numbers NAN and INF or INFINITY)
prints float with a default precision of 6, if no precision is given, using scientific
e double
notation using mantissa/exponent; lower-case exponent and special numbers
prints float with a default precision of 6, if no precision is given, using scientific
E double
notation using mantissa/exponent; upper-case exponent and special numbers
g double uses either f or e [see below]
G double uses either F or E [see below]
a double prints hexadecimal, lower-case
A double prints hexadecimal, upper-case
c char prints single character
prints string of characters up to a NUL terminator, or truncated to length given
s char*
by precision, if specified
prints void-pointer value; a nonvoid-pointer should be explicitly converted
p void*
("cast") to void*; pointer to object only, not a function-pointer
% n/a prints the % character
n int * write the number of bytes printed so far into the int pointed at.

Note that length modifiers can be applied to %n (e.g. %hhn indicates that a following n conversion specifier applies to a
pointer to a signed char argument, according to the ISO/IEC 9899:2011 §7.21.6.1 ¶7).

Note that the floating point conversions apply to types float and double because of default promotion rules —
§6.5.2.2 Function calls, ¶7 The ellipsis notation in a function prototype declarator causes argument type conversion to
stop after the last declared parameter. The default argument promotions are performed on trailing arguments.) Thus,
functions such as printf() are only ever passed double values, even if the variable referenced is of type float.

With the g and G formats, the choice between e and f (or E and F) notation is documented in the C standard and in
the POSIX specification for printf():

The double argument representing a floating-point number shall be converted in the style f or e (or in the
style F or E in the case of a G conversion specifier), depending on the value converted and the precision.
Let P equal the precision if non-zero, 6 if the precision is omitted, or 1 if the precision is zero. Then, if a
conversion with style E would have an exponent of X:

If P > X >= -4, the conversion shall be with style f (or F) and precision P - (X+1).
Otherwise, the conversion shall be with style e (or E) and precision P - 1.

GoalKicker.com – C Notes for Professionals 124

 Finally, unless the '#' flag is used, any trailing zeros shall be removed from the fractional portion of the
result and the decimal-point character shall be removed if there is no fractional portion remaining.

Section 21.2: The printf() Function

Accessed through including <stdio.h>, the function printf() is the primary tool used for printing text to the
console in C.

printf("Hello world!");
// Hello world!

Normal, unformatted character arrays can be printed by themselves by placing them directly in between the
parentheses.

printf("%d is the answer to life, the universe, and everything.", 42);

// 42 is the answer to life, the universe, and everything.

int x = 3;
char y = 'Z';
char* z = "Example";
printf("Int: %d, Char: %c, String: %s", x, y, z);
// Int: 3, Char: Z, String: Example

Alternatively, integers, floating-point numbers, characters, and more can be printed using the escape character %,
followed by a character or sequence of characters denoting the format, known as the format specifier.

All additional arguments to the function printf() are separated by commas, and these arguments should be in the
same order as the format specifiers. Additional arguments are ignored, while incorrectly typed arguments or a lack
of arguments will cause errors or undefined behavior. Each argument can be either a literal value or a variable.

After successful execution, the number of characters printed is returned with type int. Otherwise, a failure returns
a negative value.

Section 21.3: Printing format flags

The C standard (C11, and C99 too) defines the following flags for printf():

Flag Conversions Meaning

The result of the conversion shall be left-justified within the field. The conversion is right-
- all
justified if this flag is not specified.
signed The result of a signed conversion shall always begin with a sign ( '+' or '-' ). The conversion
+
numeric shall begin with a sign only when a negative value is converted if this flag is not specified.
If the first character of a signed conversion is not a sign or if a signed conversion results in
signed
<space> no characters, a <space> shall be prefixed to the result. This means that if the <space>
numeric
and '+' flags both appear, the <space> flag shall be ignored.
Specifies that the value is to be converted to an alternative form. For o conversion, it shall
increase the precision, if and only if necessary, to force the first digit of the result to be a
zero (if the value and precision are both 0, a single 0 is printed). For x or X conversion
specifiers, a non-zero result shall have 0x (or 0X) prefixed to it. For a, A, e, E, f, F, g, and G
# all conversion specifiers, the result shall always contain a radix character, even if no digits
follow the radix character. Without this flag, a radix character appears in the result of
these conversions only if a digit follows it. For g and G conversion specifiers, trailing zeros
shall not be removed from the result as they normally are. For other conversion
specifiers, the behavior is undefined.

GoalKicker.com – C Notes for Professionals 125

 For d, i, o, u, x, X, a, A, e, E, f, F, g, and G conversion specifiers, leading zeros (following any
indication of sign or base) are used to pad to the field width rather than performing space
padding, except when converting an infinity or NaN. If the '0' and '-' flags both appear, the
0 numeric '0' flag is ignored. For d, i, o, u, x, and X conversion specifiers, if a precision is specified, the
'0' flag shall be ignored. ⌦ If the '0' and <apostrophe> flags both appear, the grouping
characters are inserted before zero padding. For other conversions, the behavior is
undefined. ⌫

These flags are also supported by Microsoft with the same meanings.

The POSIX specification for printf() adds:

Flag Conversions Meaning

The integer portion of the result of a decimal conversion shall be formatted with thousands'
' i, d, u, f, F, g, G grouping characters. For other conversions the behavior is undefined. The non-monetary
grouping character is used.

Section 21.4: Printing the Value of a Pointer to an Object

To print the value of a pointer to an object (as opposed to a function pointer) use the p conversion specifier. It is
defined to print void-pointers only, so to print out the value of a non void-pointer it needs to be explicitly
converted ("casted*") to void*.

#include <stdlib.h> /* for EXIT_SUCCESS */

#include <stdio.h> /* for printf() */

int main(void)
{
int i;
int * p = &i;

printf("The address of i is %p.\n", (void*) p);

return EXIT_SUCCESS;
}
Version ≥ C99
Using <inttypes.h> and uintptr_t

Another way to print pointers in C99 or later uses the uintptr_t type and the macros from <inttypes.h>:

#include <inttypes.h> /* for uintptr_t and PRIXPTR */

#include <stdio.h> /* for printf() */

int main(void)
{
int i;
int *p = &i;

printf("The address of i is 0x%" PRIXPTR ".\n", (uintptr_t)p);

return 0;
}

In theory, there might not be an integer type that can hold any pointer converted to an integer (so the type
uintptr_t might not exist). In practice, it does exist. Pointers to functions need not be convertible to the uintptr_t
type — though again they most often are convertible.

If the uintptr_t type exists, so does the intptr_t type. It is not clear why you'd ever want to treat addresses as

GoalKicker.com – C Notes for Professionals 126

signed integers, though.

Version = K&R Version < C89

Pre-Standard History:

Prior to C89 during K&R-C times there was no type void* (nor header <stdlib.h>, nor prototypes, and hence no
int main(void) notation), so the pointer was cast to long unsigned int and printed using the lx length
modifier/conversion specifier.

The example below is just for informational purpose. Nowadays this is invalid code, which very well might
provoke the infamous Undefined Behaviour.

#include <stdio.h> /* optional in pre-standard C - for printf() */

int main()
{
int i;
int *p = &i;

printf("The address of i is 0x%lx.\n", (long unsigned) p);

return 0;
}

Section 21.5: Printing the Dierence of the Values of two

Pointers to an Object
Subtracting the values of two pointers to an object results in a signed integer *1. So it would be printed using at
least the d conversion specifier.

To make sure there is a type being wide enough to hold such a "pointer-difference", since C99 <stddef.h> defines
the type ptrdiff_t. To print a ptrdiff_t use the t length modifier.

Version ≥ C99
#include <stdlib.h> /* for EXIT_SUCCESS */
#include <stdio.h> /* for printf() */
#include <stddef.h> /* for ptrdiff_t */

int main(void)
{
int a[2];
int * p1 = &a[0], * p2 = &a[1];
ptrdiff_t pd = p2 - p1;

printf("p1 = %p\n", (void*) p1);

printf("p2 = %p\n", (void*) p2);
printf("p2 - p1 = %td\n", pd);

return EXIT_SUCCESS;
}

The result might look like this:

p1 = 0x7fff6679f430
p2 = 0x7fff6679f434
p2 - p1 = 1

GoalKicker.com – C Notes for Professionals 127

Please note that the resulting value of the difference is scaled by the size of the type the pointers subtracted point
to, an int here. The size of an int for this example is 4.

*1If the two pointers to be subtracted do not point to the same object the behaviour is undefined.

Section 21.6: Length modifiers

The C99 and C11 standards specify the following length modifiers for printf(); their meanings are:

Modifier Modifies Applies to

hh d, i, o, u, x, or X char, signed char or unsigned char
h d, i, o, u, x, or X short int or unsigned short int
l d, i, o, u, x, or X long int or unsigned long int
l a, A, e, E, f, F, g, or G double (for compatibility with scanf(); undefined in C90)
ll d, i, o, u, x, or X long long int or unsigned long long int
j d, i, o, u, x, or X intmax_t or uintmax_t
z d, i, o, u, x, or X size_t or the corresponding signed type (ssize_t in POSIX)
t d, i, o, u, x, or X ptrdiff_t or the corresponding unsigned integer type
L a, A, e, E, f, F, g, or G long double

If a length modifier appears with any conversion specifier other than as specified above, the behavior is undefined.

Microsoft specifies some different length modifiers, and explicitly does not support hh, j, z, or t.

Modifier Modifies Applies to

I32 d, i, o, x, or X __int32
I32 o, u, x, or X unsigned __int32
I64 d, i, o, x, or X __int64
I64 o, u, x, or X unsigned __int64
I d, i, o, x, or X ptrdiff_t (that is, __int32 on 32-bit platforms, __int64 on 64-bit platforms)
size_t (that is, unsigned __int32 on 32-bit platforms, unsigned __int64 on 64-bit
I o, u, x, or X
platforms)
long double (In Visual C++, although long double is a distinct type, it has the same
l or L a, A, e, E, f, g, or G
internal representation as double.)
Wide character with printf and wprintf functions. (An lc, lC, wc or wC type specifier is
l or w c or C
synonymous with C in printf functions and with c in wprintf functions.)
Wide-character string with printf and wprintf functions. (An ls, lS, ws or wS type
l or w s, S, or Z
specifier is synonymous with S in printf functions and with s in wprintf functions.)

Note that the C, S, and Z conversion specifiers and the I, I32, I64, and w length modifiers are Microsoft extensions.
Treating l as a modifier for long double rather than double is different from the standard, though you'll be hard-
pressed to spot the difference unless long double has a different representation from double.

GoalKicker.com – C Notes for Professionals 128

Chapter 22: Pointers
A pointer is a type of variable which can store the address of another object or a function.

Section 22.1: Introduction

A pointer is declared much like any other variable, except an asterisk (*) is placed between the type and the name
of the variable to denote it is a pointer.

int *pointer; /* inside a function, pointer is uninitialized and doesn't point to any valid object
yet */

To declare two pointer variables of the same type, in the same declaration, use the asterisk symbol before each
identifier. For example,

int *iptr1, *iptr2;

int *iptr3, iptr4; /* iptr3 is a pointer variable, whereas iptr4 is misnamed and is an int */

The address-of or reference operator denoted by an ampersand (&) gives the address of a given variable which can
be placed in a pointer of appropriate type.

int value = 1;
pointer = &value;

The indirection or dereference operator denoted by an asterisk (*) gets the contents of an object pointed to by a
pointer.

printf("Value of pointed to integer: %d\n", *pointer);

/* Value of pointed to integer: 1 */

If the pointer points to a structure or union type then you can dereference it and access its members directly using
the -> operator:

SomeStruct *s = &someObject;
s->someMember = 5; /* Equivalent to (*s).someMember = 5 */

In C, a pointer is a distinct value type which can be reassigned and otherwise is treated as a variable in its own right.
For example the following example prints the value of the pointer (variable) itself.

printf("Value of the pointer itself: %p\n", (void *)pointer);

/* Value of the pointer itself: 0x7ffcd41b06e4 */
/* This address will be different each time the program is executed */

Because a pointer is a mutable variable, it is possible for it to not point to a valid object, either by being set to null

pointer = 0; /* or alternatively */
pointer = NULL;

or simply by containing an arbitrary bit pattern that isn't a valid address. The latter is a very bad situation, because
it cannot be tested before the pointer is being dereferenced, there is only a test for the case a pointer is null:

if (!pointer) exit(EXIT_FAILURE);

GoalKicker.com – C Notes for Professionals 129

A pointer may only be dereferenced if it points to a valid object, otherwise the behavior is undefined. Many modern
implementations may help you by raising some kind of error such as a segmentation fault and terminate execution,
but others may just leave your program in an invalid state.

The value returned by the dereference operator is a mutable alias to the original variable, so it can be changed,
modifying the original variable.

*pointer += 1;
printf("Value of pointed to variable after change: %d\n", *pointer);
/* Value of pointed to variable after change: 2 */

Pointers are also re-assignable. This means that a pointer pointing to an object can later be used to point to
another object of the same type.

int value2 = 10;

pointer = &value2;
printf("Value from pointer: %d\n", *pointer);
/* Value from pointer: 10 */

Like any other variable, pointers have a specific type. You can't assign the address of a short int to a pointer to a
long int, for instance. Such behavior is referred to as type punning and is forbidden in C, though there are a few
exceptions.

Although pointer must be of a specific type, the memory allocated for each type of pointer is equal to the memory
used by the environment to store addresses, rather than the size of the type that is pointed to.

#include <stdio.h>

int main(void) {
printf("Size of int pointer: %zu\n", sizeof (int*)); /* size 4 bytes */
printf("Size of int variable: %zu\n", sizeof (int)); /* size 4 bytes */
printf("Size of char pointer: %zu\n", sizeof (char*)); /* size 4 bytes */
printf("Size of char variable: %zu\n", sizeof (char)); /* size 1 bytes */
printf("Size of short pointer: %zu\n", sizeof (short*)); /* size 4 bytes */
printf("Size of short variable: %zu\n", sizeof (short)); /* size 2 bytes */
return 0;
}

(NB: if you are using Microsoft Visual Studio, which does not support the C99 or C11 standards, you must use %Iu1 instead
of %zu in the above sample.)

Note that the results above can vary from environment to environment in numbers but all environments would
show equal sizes for different types of pointer.

Extract based on information from Cardiff University C Pointers Introduction

Pointers and Arrays

Pointers and arrays are intimately connected in C. Arrays in C are always held in contiguous locations in memory.
Pointer arithmetic is always scaled by the size of the item pointed to. So if we have an array of three doubles, and a
pointer to the base, *ptr refers to the first double, *(ptr + 1) to the second, *(ptr + 2) to the third. A more
convenient notation is to use array notation [].

double point[3] = {0.0, 1.0, 2.0};

double *ptr = point;

GoalKicker.com – C Notes for Professionals 130

/* prints x 0.0, y 1.0 z 2.0 */
printf("x %f y %f z %f\n", ptr[0], ptr[1], ptr[2]);

So essentially ptr and the array name are interchangeable. This rule also means that an array decays to a pointer
when passed to a subroutine.

double point[3] = {0.0, 1.0, 2.0};

printf("length of point is %s\n", length(point));

/* get the distance of a 3D point from the origin */

double length(double *pt)
{
return sqrt(pt[0] * pt[0] + pt[1] * pt[1] + pt[2] * pt[2])
}

A pointer may point to any element in an array, or to the element beyond the last element. It is however an error to
set a pointer to any other value, including the element before the array. (The reason is that on segmented
architectures the address before the first element may cross a segment boundary, the compiler ensures that does
not happen for the last element plus one).

Footnote 1: Microsoft format information can be found via printf() and format specification syntax.

Section 22.2: Common errors

Improper use of pointers are frequently a source of bugs that can include security bugs or program crashes, most
often due to segmentation faults.

Not checking for allocation failures

Memory allocation is not guaranteed to succeed, and may instead return a NULL pointer. Using the returned value,
without checking if the allocation is successful, invokes undefined behavior. This usually leads to a crash, but there
is no guarantee that a crash will happen so relying on that can also lead to problems.

For example, unsafe way:

struct SomeStruct *s = malloc(sizeof *s);

s->someValue = 0; /* UNSAFE, because s might be a null pointer */

Safe way:

struct SomeStruct *s = malloc(sizeof *s);

if (s)
{
s->someValue = 0; /* This is safe, we have checked that s is valid */
}

Using literal numbers instead of sizeof when requesting memory

For a given compiler/machine configuration, types have a known size; however, there isn't any standard which
defines that the size of a given type (other than char) will be the same for all compiler/machine configurations. If
the code uses 4 instead of sizeof(int) for memory allocation, it may work on the original machine, but the code
isn't necessarily portable to other machines or compilers. Fixed sizes for types should be replaced by

GoalKicker.com – C Notes for Professionals 131

sizeof(that_type) or sizeof(*var_ptr_to_that_type).

Non-portable allocation:

int *intPtr = malloc(4*1000); /* allocating storage for 1000 int */

long *longPtr = malloc(8*1000); /* allocating storage for 1000 long */

Portable allocation:

int *intPtr = malloc(sizeof(int)*1000); /* allocating storage for 1000 int */

long *longPtr = malloc(sizeof(long)*1000); /* allocating storage for 1000 long */

Or, better still:

int *intPtr = malloc(sizeof(*intPtr)*1000); /* allocating storage for 1000 int */

long *longPtr = malloc(sizeof(*longPtr)*1000); /* allocating storage for 1000 long */

Memory leaks

Failure to de-allocate memory using free leads to a buildup of non-reusable memory, which is no longer used by
the program; this is called a memory leak. Memory leaks waste memory resources and can lead to allocation
failures.

Logical errors

All allocations must follow the same pattern:

1. Allocation using malloc (or calloc)

2. Usage to store data
3. De-allocation using free

Failure to adhere to this pattern, such as using memory after a call to free (dangling pointer) or before a call to
malloc (wild pointer), calling free twice ("double free"), etc., usually causes a segmentation fault and results in a
crash of the program.

These errors can be transient and hard to debug – for example, freed memory is usually not immediately reclaimed
by the OS, and thus dangling pointers may persist for a while and appear to work.

On systems where it works, Valgrind is an invaluable tool for identifying what memory is leaked and where it was
originally allocated.

Creating pointers to stack variables

Creating a pointer does not extend the life of the variable being pointed to. For example:

int* myFunction()
{
int x = 10;
return &x;
}

Here, x has automatic storage duration (commonly known as stack allocation). Because it is allocated on the stack, its
lifetime is only as long as myFunction is executing; after myFunction has exited, the variable x is destroyed. This
function gets the address of x (using &x), and returns it to the caller, leaving the caller with a pointer to a non-
existent variable. Attempting to access this variable will then invoke undefined behavior.

GoalKicker.com – C Notes for Professionals 132

Most compilers don't actually clear a stack frame after the function exits, thus dereferencing the returned pointer
often gives you the expected data. When another function is called however, the memory being pointed to may be
overwritten, and it appears that the data being pointed to has been corrupted.

To resolve this, either malloc the storage for the variable to be returned, and return a pointer to the newly created
storage, or require that a valid pointer is passed in to the function instead of returning one, for example:

#include <stdlib.h>
#include <stdio.h>

int *solution1(void)
{
int *x = malloc(sizeof *x);
if (x == NULL)
{
/* Something went wrong */
return NULL;
}

*x = 10;

return x;
}

void solution2(int *x)

{
/* NB: calling this function with an invalid or null pointer
causes undefined behaviour. */

*x = 10;
}

int main(void)
{
{
/* Use solution1() */

int *foo = solution1();

if (foo == NULL)
{
/* Something went wrong */
return 1;
}

printf("The value set by solution1() is %i\n", *foo);

/* Will output: "The value set by solution1() is 10" */

free(foo); /* Tidy up */
}

{
/* Use solution2() */

int bar;
solution2(&bar);

printf("The value set by solution2() is %i\n", bar);

/* Will output: "The value set by solution2() is 10" */
}

return 0;

GoalKicker.com – C Notes for Professionals 133

}

Incrementing / decrementing and dereferencing

If you write *p++ to increment what is pointed by p, you are wrong.

Post incrementing / decrementing is executed before dereferencing. Therefore, this expression will increment the
pointer p itself and return what was pointed by p before incrementing without changing it.

You should write (*p)++ to increment what is pointed by p.

This rule also applies to post decrementing: *p-- will decrement the pointer p itself, not what is pointed by p.

Section 22.3: Dereferencing a Pointer

int a = 1;
int *a_pointer = &a;

To dereference a_pointer and change the value of a, we use the following operation

*a_pointer = 2;

This can be verified using the following print statements.

printf("%d\n", a); /* Prints 2 */

printf("%d\n", *a_pointer); /* Also prints 2 */

However, one would be mistaken to dereference a NULL or otherwise invalid pointer. This

int *p1, *p2;

p1 = (int *) 0xbad;
p2 = NULL;

*p1 = 42;
*p2 = *p1 + 1;

is usually undefined behavior. p1 may not be dereferenced because it points to an address 0xbad which may not be
a valid address. Who knows what's there? It might be operating system memory, or another program's memory.
The only time code like this is used, is in embedded development, which stores particular information at hard-
coded addresses. p2 cannot be dereferenced because it is NULL, which is invalid.

Section 22.4: Dereferencing a Pointer to a struct

Let's say we have the following structure:

struct MY_STRUCT
{
int my_int;
float my_float;
};

We can define MY_STRUCT to omit the struct keyword so we don't have to type struct MY_STRUCT each time we use
it. This, however, is optional.

GoalKicker.com – C Notes for Professionals 134

typedef struct MY_STRUCT MY_STRUCT;

If we then have a pointer to an instance of this struct

MY_STRUCT *instance;

If this statement appears at file scope, instance will be initialized with a null pointer when the program starts. If
this statement appears inside a function, its value is undefined. The variable must be initialized to point to a valid
MY_STRUCT variable, or to dynamically allocated space, before it can be dereferenced. For example:

MY_STRUCT info = { 1, 3.141593F };

MY_STRUCT *instance = &info;

When the pointer is valid, we can dereference it to access its members using one of two different notations:

int a = (*instance).my_int;
float b = instance->my_float;

While both these methods work, it is better practice to use the arrow -> operator rather than the combination of
parentheses, the dereference * operator and the dot . operator because it is easier to read and understand,
especially with nested uses.

Another important difference is shown below:

MY_STRUCT copy = *instance;

copy.my_int = 2;

In this case, copy contains a copy of the contents of instance. Changing my_int of copy will not change it in
instance.

MY_STRUCT *ref = instance;

ref->my_int = 2;

In this case, ref is a reference to instance. Changing my_int using the reference will change it in instance.

It is common practice to use pointers to structs as parameters in functions, rather than the structs themselves.
Using the structs as function parameters could cause the stack to overflow if the struct is large. Using a pointer to a
struct only uses enough stack space for the pointer, but can cause side effects if the function changes the struct
which is passed into the function.

Section 22.5: Const Pointers

Single Pointers

Pointer to an int

The pointer can point to different integers and the int's can be changed through the pointer. This sample of
code assigns b to point to int b then changes b's value to 100.

int b;
int* p;
p = &b; /* OK */
*p = 100; /* OK */

GoalKicker.com – C Notes for Professionals 135

 Pointer to a const int

The pointer can point to different integers but the int's value can't be changed through the pointer.

int b;
const int* p;
p = &b; /* OK */
*p = 100; /* Compiler Error */

const pointer to int

The pointer can only point to one int but the int's value can be changed through the pointer.

int a, b;
int* const p = &b; /* OK as initialisation, no assignment */
*p = 100; /* OK */
p = &a; /* Compiler Error */

const pointer to const int

The pointer can only point to one int and the int can not be changed through the pointer.

int a, b;
const int* const p = &b; /* OK as initialisation, no assignment */
p = &a; /* Compiler Error */
*p = 100; /* Compiler Error */

Pointer to Pointer

Pointer to a pointer to an int

This code assigns the address of p1 to the to double pointer p (which then points to int* p1 (which points to
int)).

Then changes p1 to point to int a. Then changes the value of a to be 100.

void f1(void)
{
int a, b;
int *p1;
int **p;
p1 = &b; /* OK */
p = &p1; /* OK */
*p = &a; /* OK */
**p = 100; /* OK */
}

Pointer to pointer to a const int

void f2(void)
{
int b;
const int *p1;
const int **p;

GoalKicker.com – C Notes for Professionals 136

 p = &p1; /* OK */
*p = &b; /* OK */
**p = 100; /* error: assignment of read-only location ‘**p’ */
}

Pointer to const pointer to an int

void f3(void)
{
int b;
int *p1;
int * const *p;
p = &p1; /* OK */
*p = &b; /* error: assignment of read-only location ‘*p’ */
**p = 100; /* OK */
}

const pointer to pointer to int

void f4(void)
{
int b;
int *p1;
int ** const p = &p1; /* OK as initialisation, not assignment */
p = &p1; /* error: assignment of read-only variable ‘p’ */
*p = &b; /* OK */
**p = 100; /* OK */
}

Pointer to const pointer to const int

void f5(void)
{
int b;
const int *p1;
const int * const *p;
p = &p1; /* OK */
*p = &b; /* error: assignment of read-only location ‘*p’ */
**p = 100; /* error: assignment of read-only location ‘**p’ */
}

const pointer to pointer to const int

void f6(void)
{
int b;
const int *p1;
const int ** const p = &p1; /* OK as initialisation, not assignment */
p = &p1; /* error: assignment of read-only variable ‘p’ */
*p = &b; /* OK */
**p = 100; /* error: assignment of read-only location ‘**p’ */
}

const pointer to const pointer to int

void f7(void)
{

GoalKicker.com – C Notes for Professionals 137

 int b;
int *p1;
int * const * const p = &p1; /* OK as initialisation, not assignment */
p = &p1; /* error: assignment of read-only variable ‘p’ */
*p = &b; /* error: assignment of read-only location ‘*p’ */
**p = 100; /* OK */
}

Section 22.6: Function pointers

Pointers can also be used to point at functions.

Let's take a basic function:

int my_function(int a, int b)

{
return 2 * a + 3 * b;
}

Now, let's define a pointer of that function's type:

int (*my_pointer)(int, int);

To create one, just use this template:

return_type_of_func (*my_func_pointer)(type_arg1, type_arg2, ...)

We then must assign this pointer to the function:

my_pointer = &my_function;

This pointer can now be used to call the function:

/* Calling the pointed function */

int result = (*my_pointer)(4, 2);

...

/* Using the function pointer as an argument to another function */

void another_function(int (*another_pointer)(int, int))
{
int a = 4;
int b = 2;
int result = (*another_pointer)(a, b);

printf("%d\n", result);
}

Although this syntax seems more natural and coherent with basic types, attributing and dereferencing function
pointers don't require the usage of & and * operators. So the following snippet is equally valid:

/* Attribution without the & operator */

my_pointer = my_function;

/* Dereferencing without the * operator */

int result = my_pointer(4, 2);

GoalKicker.com – C Notes for Professionals 138

To increase the readability of function pointers, typedefs may be used.

typedef void (*Callback)(int a);

void some_function(Callback callback)

{
int a = 4;
callback(a);
}

Another readability trick is that the C standard allows one to simplify a function pointer in arguments like above
(but not in variable declaration) to something that looks like a function prototype; thus the following can be
equivalently used for function definitions and declarations:

void some_function(void callback(int))

{
int a = 4;
callback(a);
}

See also

Function Pointers

Section 22.7: Polymorphic behaviour with void pointers

The qsort() standard library function is a good example of how one can use void pointers to make a single
function operate on a large variety of different types.

void qsort (
void *base, /* Array to be sorted */
size_t num, /* Number of elements in array */
size_t size, /* Size in bytes of each element */
int (*compar)(const void *, const void *)); /* Comparison function for two elements */

The array to be sorted is passed as a void pointer, so an array of any type of element can be operated on. The next
two arguments tell qsort() how many elements it should expect in the array, and how large, in bytes, each element
is.

The last argument is a function pointer to a comparison function which itself takes two void pointers. By making the
caller provide this function, qsort() can effectively sort elements of any type.

Here's an example of such a comparison function, for comparing floats. Note that any comparison function passed
to qsort() needs to have this type signature. The way it is made polymorphic is by casting the void pointer
arguments to pointers of the type of element we wish to compare.

int compare_floats(const void *a, const void *b)

{
float fa = *((float *)a);
float fb = *((float *)b);
if (fa < fb)
return -1;
if (fa > fb)
return 1;
return 0;
}

GoalKicker.com – C Notes for Professionals 139

Since we know that qsort will use this function to compare floats, we cast the void pointer arguments back to float
pointers before dereferencing them.

Now, the usage of the polymorphic function qsort on an array "array" with length "len" is very simple:

qsort(array, len, sizeof(array[0]), compare_floats);

Section 22.8: Address-of Operator ( & )

For any object (i.e, variable, array, union, struct, pointer or function) the unary address operator can be used to
access the address of that object.

Suppose that

int i = 1;
int *p = NULL;

So then a statement p = &i;, copies the address of the variable i to the pointer p.

It's expressed as p points to i.

printf("%d\n", *p); prints 1, which is the value of i.

Section 22.9: Initializing Pointers

Pointer initialization is a good way to avoid wild pointers. The initialization is simple and is no different from
initialization of a variable.

#include <stddef.h>

int main()
{
int *p1 = NULL;
char *p2 = NULL;
float *p3 = NULL;

/* NULL is a macro defined in stddef.h, stdio.h, stdlib.h, and string.h */

...
}

In most operating systems, inadvertently using a pointer that has been initialized to NULL will often result in the
program crashing immediately, making it easy to identify the cause of the problem. Using an uninitialized pointer
can often cause hard-to-diagnose bugs.

Caution:

The result of dereferencing a NULL pointer is undefined, so it will not necessarily cause a crash even if that is the
natural behaviour of the operating system the program is running on. Compiler optimizations may mask the crash,
cause the crash to occur before or after the point in the source code at which the null pointer dereference
occurred, or cause parts of the code that contains the null pointer dereference to be unexpectedly removed from
the program. Debug builds will not usually exhibit these behaviours, but this is not guaranteed by the language
standard. Other unexpected and/or undesirable behaviour is also allowed.

Because NULL never points to a variable, to allocated memory, or to a function, it is safe to use as a guard value.

GoalKicker.com – C Notes for Professionals 140

Caution:

Usually NULL is defined as (void *)0. But this does not imply that the assigned memory address is 0x0. For more
clarification refer to C-faq for NULL pointers

Note that you can also initialize pointers to contain values other than NULL.

int i1;

int main()
{
int *p1 = &i1;
const char *p2 = "A constant string to point to";
float *p3 = malloc(10 * sizeof(float));
}

Section 22.10: Pointer to Pointer

In C, a pointer can refer to another pointer.

#include <stdio.h>
#include <stdlib.h>

int main(void) {
int A = 42;
int* pA = &A;
int** ppA = &pA;
int*** pppA = &ppA;

printf("%d", ***pppA); /* prints 42 */

return EXIT_SUCCESS;
}

But, reference-and-reference directly is not allowed.

#include <stdio.h>
#include <stdlib.h>

int main(void) {
int A = 42;
int* pA = &A;
int** ppA = &&A; /* Compilation error here! */
int*** pppA = &&&A; /* Compilation error here! */

...

Section 22.11: void* pointers as arguments and return values

to standard functions
Version > K&R

void* is a catch all type for pointers to object types. An example of this in use is with the malloc function, which is
declared as

void* malloc(size_t);

GoalKicker.com – C Notes for Professionals 141

The pointer-to-void return type means that it is possible to assign the return value from malloc to a pointer to any
other type of object:

int* vector = malloc(10 * sizeof *vector);

It is generally considered good practice to not explicitly cast the values into and out of void pointers. In specific case
of malloc() this is because with an explicit cast, the compiler may otherwise assume, but not warn about, an
incorrect return type for malloc(), if you forget to include stdlib.h. It is also a case of using the correct behavior of
void pointers to better conform to the DRY (don't repeat yourself) principle; compare the above to the following,
wherein the following code contains several needless additional places where a typo could cause issues:

int* vector = (int*)malloc(10 * sizeof int*);

Similarly, functions such as

void* memcpy(void *restrict target, void const *restrict source, size_t size);

have their arguments specified as void * because the address of any object, regardless of the type, can be passed
in. Here also, a call should not use a cast

unsigned char buffer[sizeof(int)];

int b = 67;
memcpy(buffer, &b, sizeof buffer);

Section 22.12: Same Asterisk, Dierent Meanings

Premise

The most confusing thing surrounding pointer syntax in C and C++ is that there are actually two different meanings
that apply when the pointer symbol, the asterisk (*), is used with a variable.

Example

Firstly, you use * to declare a pointer variable.

int i = 5;
/* 'p' is a pointer to an integer, initialized as NULL */
int *p = NULL;
/* '&i' evaluates into address of 'i', which then assigned to 'p' */
p = &i;
/* 'p' is now holding the address of 'i' */

When you're not declaring (or multiplying), * is used to dereference a pointer variable:

*p = 123;
/* 'p' was pointing to 'i', so this changes value of 'i' to 123 */

When you want an existing pointer variable to hold address of other variable, you don't use *, but do it like this:

p = &another_variable;

A common confusion among C-programming newbies arises when they declare and initialize a pointer variable at
the same time.

GoalKicker.com – C Notes for Professionals 142

int *p = &i;

Since int i = 5; and int i; i = 5; give the same result, some of them might thought int *p = &i; and int *p;
*p = &i; give the same result too. The fact is, no, int *p; *p = &i; will attempt to deference an uninitialized
pointer which will result in UB. Never use * when you're not declaring nor dereferencing a pointer.

Conclusion

The asterisk (*) has two distinct meanings within C in relation to pointers, depending on where it's used. When used
within a variable declaration, the value on the right hand side of the equals side should be a pointer value to an
address in memory. When used with an already declared variable, the asterisk will dereference the pointer value,
following it to the pointed-to place in memory, and allowing the value stored there to be assigned or retrieved.

Takeaway

It is important to mind your P's and Q's, so to speak, when dealing with pointers. Be mindful of when you're using
the asterisk, and what it means when you use it there. Overlooking this tiny detail could result in buggy and/or
undefined behavior that you really don't want to have to deal with.

GoalKicker.com – C Notes for Professionals 143

Chapter 23: Sequence points
Section 23.1: Unsequenced expressions
Version ≥ C11

The following expressions are unsequenced:

a + b;
a - b;
a * b;
a / b;
a % b;
a & b;
a | b;

In the above examples, the expression a may be evaluated before or after the expression b, b may be evaluated
before a, or they may even be intermixed if they correspond to several instructions.

A similar rule holds for function calls:

f(a, b);

Here not only a and b are unsequenced (i.e. the , operator in a function call does not produce a sequence point) but
also f, the expression that determines the function that is to be called.

Side effects may be applied immediately after evaluation or deferred until a later point.

Expressions like

x++ & x++;

f(x++, x++); /* the ',' in a function call is *not* the same as the comma operator */
x++ * x++;
a[i] = i++;

or

x++ & x;
f(x++, x);
x++ * x;
a[i++] = i;

will yield undefined behavior because

a modification of an object and any other access to it must be sequenced

the order of evaluation and the order in which side effects1 are applied is not specified.

1 Any changes in the state of the execution environment.

Section 23.2: Sequenced expressions

The following expressions are sequenced:

a && b
a || b

GoalKicker.com – C Notes for Professionals 144

a , b
a ? b : c
for ( a ; b ; c ) { ... }

In all cases, the expression a is fully evaluated and all side effects are applied before either b or c are evaluated. In
the fourth case, only one of b or c will be evaluated. In the last case, b is fully evaluated and all side effects are
applied before c is evaluated.

In all cases, the evaluation of expression a is sequenced before the evaluations of b or c (alternately, the evaluations
of b and c are sequenced after the evaluation of a).

Thus, expressions like

x++ && x++

x++ ? x++ : y++
(x = f()) && x != 0
for ( x = 0; x < 10; x++ ) { ... }
y = (x++, x++);

have well defined behavior.

Section 23.3: Indeterminately sequenced expressions

Function calls as f(a) always imply a sequence point between the evaluation of the arguments and the designator
(here f and a) and the actual call. If two such calls are unsequenced, the two function calls are indeterminately
sequenced, that is, one is executed before the other, and order is unspecified.

unsigned counter = 0;

unsingned account(void) {
return counter++;
}

int main(void) {
printf("the order is %u %u\n", account(), account());
}

This implicit twofold modification of counter during the evaluation of the printf arguments is valid, we just don't
know which of the calls comes first. As the order is unspecified, it may vary and cannot be depended on. So the
printout could be:

the order is 0 1

or

the order is 1 0

The analogous statement to the above without intermediate function call

printf("the order is %u %u\n", counter++, counter++); // undefined behavior

has undefined behavior because there is no sequence point between the two modifications of counter.

GoalKicker.com – C Notes for Professionals 145

Chapter 24: Function Pointers
Function pointers are pointers that point to functions instead of data types. They can be used to allow variability in
the function that is to be called, at run-time.

Section 24.1: Introduction

Just like char and int, a function is a fundamental feature of C. As such, you can declare a pointer to one: which
means that you can pass which function to call to another function to help it do its job. For example, if you had a
graph() function that displayed a graph, you could pass which function to graph into graph().

// A couple of external definitions to make the example clearer

extern unsigned int screenWidth;
extern void plotXY(double x, double y);

// The graph() function.

// Pass in the bounds: the minimum and maximum X and Y that should be plotted.
// Also pass in the actual function to plot.
void graph(double minX, double minY,
double maxX, double maxY,
???? *fn) { // See below for syntax

double stepX = (maxX - minX) / screenWidth;

for (double x=minX; x<maxX; x+=stepX) {

double y = fn(x); // Get y for this x by calling passed-in fn()

if (minY<=y && y<maxY) {

plotXY(x, y); // Plot calculated point
} // if
} for
} // graph(minX, minY, maxX, maxY, fn)

Usage

So the above code will graph whatever function you passed into it - as long as that function meets certain criteria:
namely, that you pass a double in and get a double out. There are many functions like that - sin(), cos(), tan(),
exp() etc. - but there are many that aren't, such as graph() itself!

Syntax

So how do you specify which functions you can pass into graph() and which ones you can't? The conventional way
is by using a syntax that may not be easy to read or understand:

double (*fn)(double); // fn is a pointer-to-function that takes a double and returns one

The problem above is that there are two things trying to be defined at the same time: the structure of the function,
and the fact that it's a pointer. So, split the two definitions! But by using typedef, a better syntax (easier to read &
understand) can be achieved.

Section 24.2: Returning Function Pointers from a Function

#include <stdio.h>

enum Op
{

GoalKicker.com – C Notes for Professionals 146

 ADD = '+',
SUB = '-',
};

/* add: add a and b, return result */

int add(int a, int b)
{
return a + b;
}

/* sub: subtract b from a, return result */

int sub(int a, int b)
{
return a - b;
}

/* getmath: return the appropriate math function */

int (*getmath(enum Op op))(int,int)
{
switch (op)
{
case ADD:
return &add;
case SUB:
return &sub;
default:
return NULL;
}
}

int main(void)
{
int a, b, c;
int (*fp)(int,int);

fp = getmath(ADD);

a = 1, b = 2;
c = (*fp)(a, b);
printf("%d + %d = %d\n", a, b, c);
return 0;
}

Section 24.3: Best Practices

Using typedef

It might be handy to use a typedef instead of declaring the function pointer each time by hand.

The syntax for declaring a typedef for a function pointer is:

typedef returnType (*name)(parameters);

Example:

Posit that we have a function, sort, that expects a function pointer to a function compare such that:

compare - A compare function for two elements which is to be supplied to a sort function.

GoalKicker.com – C Notes for Professionals 147

 "compare" is expected to return 0 if the two elements are deemed equal, a positive value if the first
element passed is "larger" in some sense than the latter element and otherwise the function returns a
negative value (meaning that the first element is "lesser" than the latter).

Without a typedef we would pass a function pointer as an argument to a function in the following manner:

void sort(int (*compare)(const void *elem1, const void *elem2)) {

/* inside of this block, the function is named "compare" */
}

With a typedef, we'd write:

typedef int (*compare_func)(const void *, const void *);

and then we could change the function signature of sort to:

void sort(compare_func func) {

/* In this block the function is named "func" */
}

both definitions of sort would accept any function of the form

int compare(const void *arg1, const void *arg2) {

/* Note that the variable names do not have to be "elem1" and "elem2" */
}

Function pointers are the only place where you should include the pointer property of the type, e.g. do not try to
define types like typedef struct something_struct *something_type. This applies even for a structure with
members which are not supposed to accessed directly by API callers, for example the stdio.h FILE type (which as
you now will notice is not a pointer).

Taking context pointers.

A function pointer should almost always take a user-supplied void * as a context pointer.

Example
/* function minimiser, details unimportant */
double findminimum( double (*fptr)(double x, double y, void *ctx), void *ctx)
{
...
/* repeatedly make calls like this */
temp = (*fptr)(testx, testy, ctx);
}

/* the function we are minimising, sums two cubics */

double *cubics(double x, double y, void *ctx)
{
double *coeffsx = ctx;
double *coeffsy = coeffx + 4;

return coeffsx[0] * x * x * x + coeffsx[1] * x * x + coeffsx[2] * x + coeffsx[3] +

coeffsy[0] * y * y * y + coeffsy[1] * y * y + coeffsy[2] * y + coeffsy[3];

void caller()

GoalKicker.com – C Notes for Professionals 148

{
/* context, the coefficients of the cubics */
double coeffs[8] = {1, 2, 3, 4, 5, 6, 7, 8};
double min;

min = findminimum(cubics, coeffs);

}

Using the context pointer means that the extra parameters do not need to be hard-coded into the function pointed
to, or require the use globals.

The library function qsort() does not follow this rule, and one can often get away without context for trivial
comparison functions. But for anything more complicated, the context pointer becomes essential.

See also

Functions pointers

Section 24.4: Assigning a Function Pointer

#include <stdio.h>

/* increment: take number, increment it by one, and return it */

int increment(int i)
{
printf("increment %d by 1\n", i);
return i + 1;
}

/* decrement: take number, decrement it by one, and return it */

int decrement(int i)
{
printf("decrement %d by 1\n", i);
return i - 1;
}

int main(void)
{
int num = 0; /* declare number to increment */
int (*fp)(int); /* declare a function pointer */

fp = &increment; /* set function pointer to increment function */

num = (*fp)(num); /* increment num */
num = (*fp)(num); /* increment num a second time */

fp = &decrement; /* set function pointer to decrement function */

num = (*fp)(num); /* decrement num */
printf("num is now: %d\n", num);
return 0;
}

Section 24.5: Mnemonic for writing function pointers

All C functions are in actuality pointers to a spot in the program memory where some code exists. The main use of
a function pointer is to provide a "callback" to other functions (or to simulate classes and objects).

The syntax of a function, as defined further down on this page is:

GoalKicker.com – C Notes for Professionals 149

returnType (*name)(parameters)

A mnemonic for writing a function pointer definition is the following procedure:

1. Begin by writing a normal function declaration: returnType name(parameters)

2. Wrap the function name with pointer syntax: returnType (*name)(parameters)

Section 24.6: Basics

Just like you can have a pointer to an int, char, float, array/string, struct, etc. - you can have a pointer to a
function.

Declaring the pointer takes the return value of the function, the name of the function, and the type of
arguments/parameters it receives.

Say you have the following function declared and initialized:

int addInt(int n, int m){

return n+m;
}

You can declare and initialize a pointer to this function:

int (*functionPtrAdd)(int, int) = addInt; // or &addInt - the & is optional

If you have a void function it could look like this:

void Print(void){
printf("look ma' - no hands, only pointers!\n");
}

Then declaring the pointer to it would be:

void (*functionPtrPrint)(void) = Print;

Accessing the function itself would require dereferencing the pointer:

sum = (*functionPtrAdd)(2, 3); //will assign 5 to sum

(*functionPtrPrint)(); //will print the text in Print function

As seen in more advanced examples in this document, declaring a pointer to a function could get messy if the
function is passed more than a few parameters. If you have a few pointers to functions that have identical
"structure" (same type of return value, and same type of parameters) it's best to use the typedef command to save
you some typing, and to make the code more clear:

typedef int (*ptrInt)(int, int);

int Add(int i, int j){

return i+j;
}

int Multiply(int i, int j){

return i*j;
}

int main()

GoalKicker.com – C Notes for Professionals 150

{
ptrInt ptr1 = Add;
ptrInt ptr2 = Multiply;

printf("%d\n", (*ptr1)(2,3)); //will print 5

printf("%d\n", (*ptr2)(2,3)); //will print 6
return 0;
}

You can also create an Array of function-pointers. If all the pointers are of the same "structure":

int (*array[2]) (int x, int y); // can hold 2 function pointers

array[0] = Add;
array[1] = Multiply;

You can learn more here and here.

It is also possible to define an array of function-pointers of different types, though that would require casting when-
ever you want to access the specific function. You can learn more here.

GoalKicker.com – C Notes for Professionals 151

Chapter 25: Function Parameters
Section 25.1: Parameters are passed by value
In C, all function parameters are passed by value, so modifying what is passed in callee functions won't affect caller
functions' local variables.

#include <stdio.h>

void modify(int v) {
printf("modify 1: %d\n", v); /* 0 is printed */
v = 42;
printf("modify 2: %d\n", v); /* 42 is printed */
}

int main(void) {
int v = 0;
printf("main 1: %d\n", v); /* 0 is printed */
modify(v);
printf("main 2: %d\n", v); /* 0 is printed, not 42 */
return 0;
}

You can use pointers to let callee functions modify caller functions' local variables. Note that this is not pass by
reference but the pointer values pointing at the local variables are passed.

#include <stdio.h>

void modify(int* v) {
printf("modify 1: %d\n", *v); /* 0 is printed */
*v = 42;
printf("modify 2: %d\n", *v); /* 42 is printed */
}

int main(void) {
int v = 0;
printf("main 1: %d\n", v); /* 0 is printed */
modify(&v);
printf("main 2: %d\n", v); /* 42 is printed */
return 0;
}

However returning the address of a local variable to the callee results in undefined behaviour. See Dereferencing a
pointer to variable beyond its lifetime.

Section 25.2: Passing in Arrays to Functions

int getListOfFriends(size_t size, int friend_indexes[]) {
size_t i = 0;
for (; i < size; i++) {
friend_indexes[i] = i;
}
}

Version ≥ C99 Version < C11

/* Type "void" and VLAs ("int friend_indexes[static size]") require C99 at least.
In C11 VLAs are optional. */
void getListOfFriends(size_t size, int friend_indexes[static size]) {

GoalKicker.com – C Notes for Professionals 152

 size_t i = 0;
for (; i < size; i++) {
friend_indexes[i] = 1;
}
}

Here the static inside the [] of the function parameter, request that the argument array must have at least as
many elements as are specified (i.e. size elements). To be able to use that feature we have to ensure that the size
parameter comes before the array parameter in the list.

Use getListOfFriends() like this:

#define LIST_SIZE (50)

int main(void) {
size_t size_of_list = LIST_SIZE;
int friends_indexes[size_of_list];

getListOfFriends(size_of_list, friend_indexes); /* friend_indexes decays to a pointer to the

address of its 1st element:
&friend_indexes[0] */

/* Here friend_indexes carries: {0, 1, 2, ..., 49}; */

return 0;
}

See also

Passing multidimensional arrays to a function

Section 25.3: Order of function parameter execution

The order of execution of parameters is undefined in C programming. Here it may execute from left to right or from
right to left. The order depends on the implementation.

#include <stdio.h>

void function(int a, int b)

{
printf("%d %d\n", a, b);
}

int main(void)
{
int a = 1;
function(a++, ++a);
return 0;
}

Section 25.4: Using pointer parameters to return multiple

values
A common pattern in C, to easily imitate returning multiple values from a function, is to use pointers.

#include <stdio.h>

GoalKicker.com – C Notes for Professionals 153

void Get( int* c , double* d )
{
*c = 72;
*d = 175.0;
}

int main(void)
{
int a = 0;
double b = 0.0;

Get( &a , &b );

printf("a: %d, b: %f\n", a , b );

return 0;
}

Section 25.5: Example of function returning struct containing

values with error codes
Most examples of a function returning a value involve providing a pointer as one of the arguments to allow the
function to modify the value pointed to, similar to the following. The actual return value of the function is usually
some type such as an int to indicate the status of the result, whether it worked or not.

int func (int *pIvalue)

{
int iRetStatus = 0; /* Default status is no change */

if (*pIvalue > 10) {

*pIvalue = *pIvalue * 45; /* Modify the value pointed to */
iRetStatus = 1; /* indicate value was changed */
}

return iRetStatus; /* Return an error code */

}

However you can also use a struct as a return value which allows you to return both an error status along with
other values as well. For instance.

typedef struct {
int iStat; /* Return status */
int iValue; /* Return value */
} RetValue;

RetValue func (int iValue)

{
RetValue iRetStatus = {0, iValue};

if (iValue > 10) {

iRetStatus.iValue = iValue * 45;
iRetStatus.iStat = 1;
}

return iRetStatus;
}

This function could then be used like the following sample.

GoalKicker.com – C Notes for Professionals 154

int usingFunc (int iValue)
{
RetValue iRet = func (iValue);

if (iRet.iStat == 1) {
/* do things with iRet.iValue, the returned value */
}
return 0;
}

Or it could be used like the following.

int usingFunc (int iValue)

{
RetValue iRet;

if ( (iRet = func (iValue)).iStat == 1 ) {

/* do things with iRet.iValue, the returned value */
}
return 0;
}

GoalKicker.com – C Notes for Professionals 155

Chapter 26: Pass 2D-arrays to functions
Section 26.1: Pass a 2D-array to a function
Passing a 2d array to a functions seems simple and obvious and we happily write:

#include <stdio.h>
#include <stdlib.h>

#define ROWS 3
#define COLS 2

void fun1(int **, int, int);

int main()
{
int array_2D[ROWS][COLS] = { {1, 2}, {3, 4}, {5, 6} };
int n = ROWS;
int m = COLS;

fun1(array_2D, n, m);

return EXIT_SUCCESS;
}

void fun1(int **a, int n, int m)

{
int i, j;
for (i = 0; i < n; i++) {
for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);
}
}
}

But the compiler, here GCC in version 4.9.4 , does not appreciate it well.

$ gcc-4.9 -O3 -g3 -W -Wall -Wextra -std=c11 passarr.c -o passarr

passarr.c: In function ‘main’:
passarr.c:16:8: warning: passing argument 1 of ‘fun1’ from incompatible pointer type
fun1(array_2D, n, m);
^
passarr.c:8:6: note: expected ‘int **’ but argument is of type ‘int (*)[2]’
void fun1(int **, int, int);

The reasons for this are twofold: the main problem is that arrays are not pointers and the second inconvenience is
the so called pointer decay. Passing an array to a function will decay the array to a pointer to the first element of the
array--in the case of a 2d array it decays to a pointer to the first row because in C arrays are sorted row-first.

#include <stdio.h>
#include <stdlib.h>

#define ROWS 3
#define COLS 2

void fun1(int (*)[COLS], int, int);

GoalKicker.com – C Notes for Professionals 156

int main()
{
int array_2D[ROWS][COLS] = { {1, 2}, {3, 4}, {5, 6} };
int n = ROWS;
int m = COLS;

fun1(array_2D, n, m);

return EXIT_SUCCESS;
}

void fun1(int (*a)[COLS], int n, int m)

{
int i, j;
for (i = 0; i < n; i++) {
for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);
}
}
}

It is necessary to pass the number of rows, they cannot be computed.

#include <stdio.h>
#include <stdlib.h>

#define ROWS 3
#define COLS 2

void fun1(int (*)[COLS], int);

int main()
{
int array_2D[ROWS][COLS] = { {1, 2}, {3, 4}, {5, 6} };
int rows = ROWS;

/* works here because array_2d is still in scope and still an array */

printf("MAIN: %zu\n",sizeof(array_2D)/sizeof(array_2D[0]));

fun1(array_2D, rows);

return EXIT_SUCCESS;
}

void fun1(int (*a)[COLS], int rows)

{
int i, j;
int n, m;

n = rows;
/* Works, because that information is passed (as "COLS").
It is also redundant because that value is known at compile time (in "COLS"). */
m = (int) (sizeof(a[0])/sizeof(a[0][0]));

/* Does not work here because the "decay" in "pointer decay" is meant
literally--information is lost. */
printf("FUN1: %zu\n",sizeof(a)/sizeof(a[0]));

for (i = 0; i < n; i++) {

for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);

GoalKicker.com – C Notes for Professionals 157

 }
}
}
Version = C99

The number of columns is predefined and hence fixed at compile time, but the predecessor to the current C-
standard (that was ISO/IEC 9899:1999, current is ISO/IEC 9899:2011) implemented VLAs (TODO: link it) and although
the current standard made it optional, almost all modern C-compilers support it (TODO: check if MS Visual Studio
supports it now).

#include <stdio.h>
#include <stdlib.h>

/* ALL CHECKS OMMITTED!*/

void fun1(int (*)[], int rows, int cols);

int main(int argc, char **argv)

{
int rows, cols, i, j;

if(argc != 3){
fprintf(stderr,"Usage: %s rows cols\n",argv[0]);
exit(EXIT_FAILURE);
}

rows = atoi(argv[1]);
cols = atoi(argv[2]);

int array_2D[rows][cols];

for (i = 0; i < rows; i++) {

for (j = 0; j < cols; j++) {
array_2D[i][j] = (i + 1) * (j + 1);
printf("array[%d][%d]=%d\n", i, j, array_2D[i][j]);
}
}

fun1(array_2D, rows, cols);

exit(EXIT_SUCCESS);
}

void fun1(int (*a)[], int rows, int cols)

{
int i, j;
int n, m;

n = rows;
/* Does not work anymore, no sizes are specified anymore
m = (int) (sizeof(a[0])/sizeof(a[0][0])); */
m = cols;

for (i = 0; i < n; i++) {

for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);
}
}
}

GoalKicker.com – C Notes for Professionals 158

This does not work, the compiler complains:

$ gcc-4.9 -O3 -g3 -W -Wall -Wextra -std=c99 passarr.c -o passarr

passarr.c: In function ‘fun1’:
passarr.c:168:7: error: invalid use of array with unspecified bounds
printf("array[%d][%d]=%d\n", i, j, a[i][j]);

It becomes a bit clearer if we intentionally make an error in the call of the function by changing the declaration to
void fun1(int **a, int rows, int cols). That causes the compiler to complain in a different, but equally
nebulous way

$ gcc-4.9 -O3 -g3 -W -Wall -Wextra -std=c99 passarr.c -o passarr

passarr.c: In function ‘main’:
passarr.c:208:8: warning: passing argument 1 of ‘fun1’ from incompatible pointer type
fun1(array_2D, rows, cols);
^
passarr.c:185:6: note: expected ‘int **’ but argument is of type ‘int (*)[(sizetype)(cols)]’
void fun1(int **, int rows, int cols);

We can react in several ways, one of it is to ignore all of it and do some illegible pointer juggling:

#include <stdio.h>
#include <stdlib.h>

/* ALL CHECKS OMMITTED!*/

void fun1(int (*)[], int rows, int cols);

int main(int argc, char **argv)

{
int rows, cols, i, j;

if(argc != 3){
fprintf(stderr,"Usage: %s rows cols\n",argv[0]);
exit(EXIT_FAILURE);
}

rows = atoi(argv[1]);
cols = atoi(argv[2]);

int array_2D[rows][cols];
printf("Make array with %d rows and %d columns\n", rows, cols);
for (i = 0; i < rows; i++) {
for (j = 0; j < cols; j++) {
array_2D[i][j] = i * cols + j;
printf("array[%d][%d]=%d\n", i, j, array_2D[i][j]);
}
}

fun1(array_2D, rows, cols);

exit(EXIT_SUCCESS);
}

void fun1(int (*a)[], int rows, int cols)

{
int i, j;
int n, m;

GoalKicker.com – C Notes for Professionals 159

 n = rows;
m = cols;

printf("\nPrint array with %d rows and %d columns\n", rows, cols);

for (i = 0; i < n; i++) {
for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, *( (*a) + (i * cols + j)));
}
}
}

Or we do it right and pass the needed information to fun1. To do so wee need to rearrange the arguments to fun1:
the size of the column must come before the declaration of the array. To keep it more readable the variable holding
the number of rows has changed its place, too, and is first now.

#include <stdio.h>
#include <stdlib.h>

/* ALL CHECKS OMMITTED!*/

void fun1(int rows, int cols, int (*)[]);

int main(int argc, char **argv)

{
int rows, cols, i, j;

if(argc != 3){
fprintf(stderr,"Usage: %s rows cols\n",argv[0]);
exit(EXIT_FAILURE);
}

rows = atoi(argv[1]);
cols = atoi(argv[2]);

int array_2D[rows][cols];
printf("Make array with %d rows and %d columns\n", rows, cols);
for (i = 0; i < rows; i++) {
for (j = 0; j < cols; j++) {
array_2D[i][j] = i * cols + j;
printf("array[%d][%d]=%d\n", i, j, array_2D[i][j]);
}
}

fun1(rows, cols, array_2D);

exit(EXIT_SUCCESS);
}

void fun1(int rows, int cols, int (*a)[cols])

{
int i, j;
int n, m;

n = rows;
m = cols;

printf("\nPrint array with %d rows and %d columns\n", rows, cols);

for (i = 0; i < n; i++) {
for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);
}

GoalKicker.com – C Notes for Professionals 160

 }
}

This looks awkward to some people, who hold the opinion that the order of variables should not matter. That is not
much of a problem, just declare a pointer and let it point to the array.

#include <stdio.h>
#include <stdlib.h>

/* ALL CHECKS OMMITTED!*/

void fun1(int rows, int cols, int **);

int main(int argc, char **argv)

{
int rows, cols, i, j;

if(argc != 3){
fprintf(stderr,"Usage: %s rows cols\n",argv[0]);
exit(EXIT_FAILURE);
}

rows = atoi(argv[1]);
cols = atoi(argv[2]);

int array_2D[rows][cols];
printf("Make array with %d rows and %d columns\n", rows, cols);
for (i = 0; i < rows; i++) {
for (j = 0; j < cols; j++) {
array_2D[i][j] = i * cols + j;
printf("array[%d][%d]=%d\n", i, j, array_2D[i][j]);
}
}
// a "rows" number of pointers to "int". Again a VLA
int *a[rows];
// initialize them to point to the individual rows
for (i = 0; i < rows; i++) {
a[i] = array_2D[i];
}

fun1(rows, cols, a);

exit(EXIT_SUCCESS);
}

void fun1(int rows, int cols, int **a)

{
int i, j;
int n, m;

n = rows;
m = cols;

printf("\nPrint array with %d rows and %d columns\n", rows, cols);

for (i = 0; i < n; i++) {
for (j = 0; j < m; j++) {
printf("array[%d][%d]=%d\n", i, j, a[i][j]);
}
}
}

GoalKicker.com – C Notes for Professionals 161

Section 26.2: Using flat arrays as 2D arrays
Often the easiest solution is simply to pass 2D and higher arrays around as flat memory.

/* create 2D array with dimensions determined at runtime */

double *matrix = malloc(width * height * sizeof(double));

/* initialise it (for the sake of illustration we want 1.0 on the diagonal) */

int x, y;
for (y = 0; y < height; y++)
{
for (x = 0; x < width; x++)
{
if (x == y)
matrix[y * width + x] = 1.0;
else
matrix[y * width + x] = 0.0;
}
}

/* pass it to a subroutine */
manipulate_matrix(matrix, width, height);

/* do something with the matrix, e.g. scale by 2 */

void manipulate_matrix(double *matrix, int width, int height)
{
int x, y;

for (y = 0; y < height; y++)

{
for (x = 0; x < width; x++)
{
matrix[y * width + x] *= 2.0;
}
}
}

GoalKicker.com – C Notes for Professionals 162

Chapter 27: Error handling
Section 27.1: errno
When a standard library function fails, it often sets errno to the appropriate error code. The C standard requires at
least 3 values for errno be set:

Value Meaning
EDOM Domain error
ERANGE Range error
EILSEQ Illegal multi-byte character sequence

Section 27.2: strerror

If perror is not flexible enough, you may obtain a user-readable error description by calling strerror from
<string.h>.

int main(int argc, char *argv[])

{
FILE *fout;
int last_error = 0;

if ((fout = fopen(argv[1], "w")) == NULL) {

last_error = errno;
/* reset errno and continue */
errno = 0;
}

/* do some processing and try opening the file differently, then */

if (last_error) {
fprintf(stderr, "fopen: Could not open %s for writing: %s",
argv[1], strerror(last_error));
fputs("Cross fingers and continue", stderr);
}

/* do some other processing */

return EXIT_SUCCESS;
}

Section 27.3: perror

To print a user-readable error message to stderr, call perror from <stdio.h>.

int main(int argc, char *argv[])

{
FILE *fout;

if ((fout = fopen(argv[1], "w")) == NULL) {

perror("fopen: Could not open file for writing");
return EXIT_FAILURE;
}
return EXIT_SUCCESS;
}

GoalKicker.com – C Notes for Professionals 163

This will print an error message concerning the current value of errno.

GoalKicker.com – C Notes for Professionals 164

Chapter 28: Undefined behavior
In C, some expressions yield undefined behavior. The standard explicitly chooses to not define how a compiler
should behave if it encounters such an expression. As a result, a compiler is free to do whatever it sees fit and may
produce useful results, unexpected results, or even crash.

Code that invokes UB may work as intended on a specific system with a specific compiler, but will likely not work on
another system, or with a different compiler, compiler version or compiler settings.

Section 28.1: Dereferencing a pointer to variable beyond its

lifetime
int* foo(int bar)
{
int baz = 6;
baz += bar;
return &baz; /* (&baz) copied to new memory location outside of foo. */
} /* (1) The lifetime of baz and bar end here as they have automatic storage
* duration (local variables), thus the returned pointer is not valid! */

int main (void)

{
int* p;

p = foo(5); /* (2) this expression's behavior is undefined */

*p = *p - 6; /* (3) Undefined behaviour here */

return 0;
}

Some compilers helpfully point this out. For example, gcc warns with:

warning: function returns address of local variable [-Wreturn-local-addr]

and clang warns with:

warning: address of stack memory associated with local variable 'baz' returned
[-Wreturn-stack-address]

for the above code. But compilers may not be able to help in complex code.

(1) Returning reference to variable declared static is defined behaviour, as the variable is not destroyed after
leaving current scope.

(2) According to ISO/IEC 9899:2011 6.2.4 §2, "The value of a pointer becomes indeterminate when the object it
points to reaches the end of its lifetime."

(3) Dereferencing the pointer returned by the function foo is undefined behaviour as the memory it references
holds an indeterminate value.

Section 28.2: Copying overlapping memory

A wide variety of standard library functions have among their effects copying byte sequences from one memory

GoalKicker.com – C Notes for Professionals 165

region to another. Most of these functions have undefined behavior when the source and destination regions
overlap.

For example, this ...

#include <string.h> /* for memcpy() */

char str[19] = "This is an example";

memcpy(str + 7, str, 10);

... attempts to copy 10 bytes where the source and destination memory areas overlap by three bytes. To visualize:

overlapping area
|
_ _
| |
v v
T h i s i s a n e x a m p l e \0
^ ^
| |
| destination
|
source

Because of the overlap, the resulting behavior is undefined.

Among the standard library functions with a limitation of this kind are memcpy(), strcpy(), strcat(), sprintf(),
and sscanf(). The standard says of these and several other functions:

If copying takes place between objects that overlap, the behavior is undefined.

The memmove() function is the principal exception to this rule. Its definition specifies that the function behaves as if
the source data were first copied into a temporary buffer and then written to the destination address. There is no
exception for overlapping source and destination regions, nor any need for one, so memmove() has well-defined
behavior in such cases.

The distinction reflects an efficiency vs. generality tradeoff. Copying such as these functions perform usually occurs
between disjoint regions of memory, and often it is possible to know at development time whether a particular
instance of memory copying will be in that category. Assuming non-overlap affords comparatively more efficient
implementations that do not reliably produce correct results when the assumption does not hold. Most C library
functions are allowed the more efficient implementations, and memmove() fills in the gaps, serving the cases where
the source and destination may or do overlap. To produce the correct effect in all cases, however, it must perform
additional tests and / or employ a comparatively less efficient implementation.

Section 28.3: Signed integer overflow

Per paragraph 6.5/5 of both C99 and C11, evaluation of an expression produces undefined behavior if the result is
not a representable value of the expression's type. For arithmetic types, that's called an overflow. Unsigned integer
arithmetic does not overflow because paragraph 6.2.5/9 applies, causing any unsigned result that otherwise would
be out of range to be reduced to an in-range value. There is no analogous provision for signed integer types,
however; these can and do overflow, producing undefined behavior. For example,

#include <limits.h> /* to get INT_MAX */

GoalKicker.com – C Notes for Professionals 166

int main(void) {
int i = INT_MAX + 1; /* Overflow happens here */
return 0;
}

Most instances of this type of undefined behavior are more difficult to recognize or predict. Overflow can in
principle arise from any addition, subtraction, or multiplication operation on signed integers (subject to the usual
arithmetic conversions) where there are not effective bounds on or a relationship between the operands to prevent
it. For example, this function:

int square(int x) {
return x * x; /* overflows for some values of x */
}

is reasonable, and it does the right thing for small enough argument values, but its behavior is undefined for larger
argument values. You cannot judge from the function alone whether programs that call it exhibit undefined
behavior as a result. It depends on what arguments they pass to it.

On the other hand, consider this trivial example of overflow-safe signed integer arithmetic:

int zero(int x) {
return x - x; /* Cannot overflow */
}

The relationship between the operands of the subtraction operator ensures that the subtraction never overflows.
Or consider this somewhat more practical example:

int sizeDelta(FILE *f1, FILE *f2) {

int count1 = 0;
int count2 = 0;
while (fgetc(f1) != EOF) count1++; /* might overflow */
while (fgetc(f2) != EOF) count2++; /* might overflow */

return count1 - count2; /* provided no UB to this point, will not overflow */

}

As long as that the counters do not overflow individually, the operands of the final subtraction will both be non-
negative. All differences between any two such values are representable as int.

Section 28.4: Use of an uninitialized variable

int a;
printf("%d", a);

The variable a is an int with automatic storage duration. The example code above is trying to print the value of an
uninitialized variable (a was never initialized). Automatic variables which are not initialized have indeterminate
values; accessing these can lead to undefined behavior.

Note: Variables with static or thread local storage, including global variables without the static keyword, are
initialized to either zero, or their initialized value. Hence the following is legal.

static int b;
printf("%d", b);

A very common mistake is to not initialize the variables that serve as counters to 0. You add values to them, but

GoalKicker.com – C Notes for Professionals 167

since the initial value is garbage, you will invoke Undefined Behavior, such as in the question Compilation on
terminal gives off pointer warning and strange symbols.

Example:

#include <stdio.h>

int main(void) {
int i, counter;
for(i = 0; i < 10; ++i)
counter += i;
printf("%d\n", counter);
return 0;
}

Output:

C02QT2UBFVH6-lm:~ gsamaras$ gcc main.c -Wall -o main

main.c:6:9: warning: variable 'counter' is uninitialized when used here [-Wuninitialized]
counter += i;
^~~~~~~
main.c:4:19: note: initialize the variable 'counter' to silence this warning
int i, counter;
^
= 0
1 warning generated.
C02QT2UBFVH6-lm:~ gsamaras$ ./main
32812

The above rules are applicable for pointers as well. For example, the following results in undefined behavior

int main(void)
{
int *p;
p++; // Trying to increment an uninitialized pointer.
}

Note that the above code on its own might not cause an error or segmentation fault, but trying to dereference this
pointer later would cause the undefined behavior.

Section 28.5: Data race

Version ≥ C11

C11 introduced support for multiple threads of execution, which affords the possibility of data races. A program
contains a data race if an object in it is accessed1 by two different threads, where at least one of the accesses is
non-atomic, at least one modifies the object, and program semantics fail to ensure that the two accesses cannot
overlap temporally.2 Note well that actual concurrency of the accesses involved is not a condition for a data race;
data races cover a broader class of issues arising from (allowed) inconsistencies in different threads' views of
memory.

Consider this example:

#include <threads.h>

int a = 0;

GoalKicker.com – C Notes for Professionals 168

int Function( void* ignore )
{
a = 1;

return 0;
}

int main( void )

{
thrd_t id;
thrd_create( &id , Function , NULL );

int b = a;

thrd_join( id , NULL );
}

The main thread calls thrd_create to start a new thread running function Function. The second thread modifies a,
and the main thread reads a. Neither of those access is atomic, and the two threads do nothing either individually
or jointly to ensure that they do not overlap, so there is a data race.

Among the ways this program could avoid the data race are

the main thread could perform its read of a before starting the other thread;
the main thread could perform its read of a after ensuring via thrd_join that the other has terminated;
the threads could synchronize their accesses via a mutex, each one locking that mutex before accessing a
and unlocking it afterward.

As the mutex option demonstrates, avoiding a data race does not require ensuring a specific order of operations,
such as the child thread modifying a before the main thread reads it; it is sufficient (for avoiding a data race) to
ensure that for a given execution, one access will happen before the other.

1 Modifying or reading an object.

2 (Quoted from ISO:IEC 9889:201x, section 5.1.2.4 "Multi-threaded executions and data races")
The execution of a program contains a data race if it contains two conflicting actions in different threads, at least
one of which is not atomic, and neither happens before the other. Any such data race results in undefined
behavior.

Section 28.6: Read value of pointer that was freed

Even just reading the value of a pointer that was freed (i.e. without trying to dereference the pointer) is undefined
behavior(UB), e.g.

char *p = malloc(5);
free(p);
if (p == NULL) /* NOTE: even without dereferencing, this may have UB */
{

Quoting ISO/IEC 9899:2011, section 6.2.4 §2:

[…] The value of a pointer becomes indeterminate when the object it points to (or just past) reaches the
end of its lifetime.

GoalKicker.com – C Notes for Professionals 169

The use of indeterminate memory for anything, including apparently harmless comparison or arithmetic, can have
undefined behavior if the value can be a trap representation for the type.

Section 28.7: Using incorrect format specifier in printf

Using an incorrect format specifier in the first argument to printf invokes undefined behavior. For example, the
code below invokes undefined behavior:

long z = 'B';
printf("%c\n", z);

Here is another example

printf("%f\n",0);

Above line of code is undefined behavior. %f expects double. However 0 is of type int.

Note that your compiler usually can help you avoid cases like these, if you turn on the proper flags during compiling
(-Wformat in clang and gcc). From the last example:

warning: format specifies type 'double' but the argument has type
'int' [-Wformat]
printf("%f\n",0);
~~ ^
%d

Section 28.8: Modify string literal

In this code example, the char pointer p is initialized to the address of a string literal. Attempting to modify the
string literal has undefined behavior.

char *p = "hello world";

p[0] = 'H'; // Undefined behavior

However, modifying a mutable array of char directly, or through a pointer is naturally not undefined behavior, even
if its initializer is a literal string. The following is fine:

char a[] = "hello, world";

char *p = a;

a[0] = 'H';
p[7] = 'W';

That's because the string literal is effectively copied to the array each time the array is initialized (once for variables
with static duration, each time the array is created for variables with automatic or thread duration — variables with
allocated duration aren't initialized), and it is fine to modify array contents.

Section 28.9: Passing a null pointer to printf %s conversion

The %s conversion of printf states that the corresponding argument a pointer to the initial element of an array of
character type. A null pointer does not point to the initial element of any array of character type, and thus the
behavior of the following is undefined:

char *foo = NULL;

GoalKicker.com – C Notes for Professionals 170

printf("%s", foo); /* undefined behavior */

However, the undefined behavior does not always mean that the program crashes — some systems take steps to
avoid the crash that normally happens when a null pointer is dereferenced. For example Glibc is known to print

(null)

for the code above. However, add (just) a newline to the format string and you will get a crash:

char *foo = 0;
printf("%s\n", foo); /* undefined behavior */

In this case, it happens because GCC has an optimization that turns printf("%s\n", argument); into a call to puts
with puts(argument), and puts in Glibc does not handle null pointers. All this behavior is standard conforming.

Note that null pointer is different from an empty string. So, the following is valid and has no undefined behaviour. It'll
just print a newline:

char *foo = "";

printf("%s\n", foo);

Section 28.10: Modifying any object more than once between

two sequence points
int i = 42;
i = i++; /* Assignment changes variable, post-increment as well */
int a = i++ + i--;

Code like this often leads to speculations about the "resulting value" of i. Rather than specifying an outcome,
however, the C standards specify that evaluating such an expression produces undefined behavior. Prior to C2011,
the standard formalized these rules in terms of so-called sequence points:

Between the previous and next sequence point a scalar object shall have its stored value modified at
most once by the evaluation of an expression. Furthermore, the prior value shall be read only to
determine the value to be stored.

(C99 standard, section 6.5, paragraph 2)

That scheme proved to be a little too coarse, resulting in some expressions exhibiting undefined behavior with
respect to C99 that plausibly should not do. C2011 retains sequence points, but introduces a more nuanced
approach to this area based on sequencing and a relationship it calls "sequenced before":

If a side effect on a scalar object is unsequenced relative to either a different side effect on the same
scalar object or a value computation using the value of the same scalar object, the behavior is undefined.
If there are multiple allowable orderings of the subexpressions of an expression, the behavior is
undefined if such an unsequenced side effect occurs in any of the orderings.

(C2011 standard, section 6.5, paragraph 2)

The full details of the "sequenced before" relation are too long to describe here, but they supplement sequence
points rather than supplanting them, so they have the effect of defining behavior for some evaluations whose

GoalKicker.com – C Notes for Professionals 171

behavior previously was undefined. In particular, if there is a sequence point between two evaluations, then the
one before the sequence point is "sequenced before" the one after.

The following example has well-defined behaviour:

int i = 42;
i = (i++, i+42); /* The comma-operator creates a sequence point */

The following example has undefined behaviour:

int i = 42;
printf("%d %d\n", i++, i++); /* commas as separator of function arguments are not comma-operators */

As with any form of undefined behavior, observing the actual behavior of evaluating expressions that violate the
sequencing rules is not informative, except in a retrospective sense. The language standard provides no basis for
expecting such observations to be predictive even of the future behavior of the same program.

Section 28.11: Freeing memory twice

Freeing memory twice is undefined behavior, e.g.

int * x = malloc(sizeof(int));
*x = 9;
free(x);
free(x);

Quote from standard(7.20.3.2. The free function of C99 ):

Otherwise, if the argument does not match a pointer earlier returned by the calloc, malloc, or realloc
function, or if the space has been deallocated by a call to free or realloc, the behavior is undefined.

Section 28.12: Bit shifting using negative counts or beyond the

width of the type
If the shift count value is a negative value then both left shift and right shift operations are undefined1:

int x = 5 << -3; /* undefined */

int x = 5 >> -3; /* undefined */

If left shift is performed on a negative value, it's undefined:

int x = -5 << 3; /* undefined */

If left shift is performed on a positive value and result of the mathematical value is not representable in the type,
it's undefined1:

/* Assuming an int is 32-bits wide, the value '5 * 2^72' doesn't fit
* in an int. So, this is undefined. */

int x = 5 << 72;

Note that right shift on a negative value (.e.g -5 >> 3) is not undefined but implementation-defined.

GoalKicker.com – C Notes for Professionals 172

1 Quoting ISO/IEC 9899:201x, section 6.5.7:

If the value of the right operand is negative or is greater than or equal to the width of the promoted left
operand, the behavior is undefined.

Section 28.13: Returning from a function that's declared with

`_Noreturn` or `noreturn` function specifier
Version ≥ C11

The function specifier _Noreturn was introduced in C11. The header <stdnoreturn.h> provides a macro noreturn
which expands to _Noreturn. So using _Noreturn or noreturn from <stdnoreturn.h> is fine and equivalent.

A function that's declared with _Noreturn (or noreturn) is not allowed to return to its caller. If such a function does
return to its caller, the behavior is undefined.

In the following example, func() is declared with noreturn specifier but it returns to its caller.

#include <stdio.h>
#include <stdlib.h>
#include <stdnoreturn.h>

noreturn void func(void);

void func(void)
{
printf("In func()...\n");
} /* Undefined behavior as func() returns */

int main(void)
{
func();
return 0;
}

gcc and clang produce warnings for the above program:

$ gcc test.c
test.c: In function ‘func’:
test.c:9:1: warning: ‘noreturn’ function does return
}
^
$ clang test.c
test.c:9:1: warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
^

An example using noreturn that has well-defined behavior:

#include <stdio.h>
#include <stdlib.h>
#include <stdnoreturn.h>

noreturn void my_exit(void);

GoalKicker.com – C Notes for Professionals 173

/* calls exit() and doesn't return to its caller. */
void my_exit(void)
{
printf("Exiting...\n");
exit(0);
}

int main(void)
{
my_exit();
return 0;
}

Section 28.14: Accessing memory beyond allocated chunk

A a pointer to a piece of memory containing n elements may only be dereferenced if it is in the range memory and
memory + (n - 1). Dereferencing a pointer outside of that range results in undefined behavior. As an example,
consider the following code:

int array[3];
int *beyond_array = array + 3;
*beyond_array = 0; /* Accesses memory that has not been allocated. */

The third line accesses the 4th element in an array that is only 3 elements long, leading to undefined behavior.
Similarly, the behavior of the second line in the following code fragment is also not well defined:

int array[3];
array[3] = 0;

Note that pointing past the last element of an array is not undefined behavior (beyond_array = array + 3 is well
defined here), but dereferencing it is (*beyond_array is undefined behavior). This rule also holds for dynamically
allocated memory (such as buffers created through malloc).

Section 28.15: Modifying a const variable using a pointer

int main (void)
{
const int foo_readonly = 10;
int *foo_ptr;

foo_ptr = (int *)&foo_readonly; /* (1) This casts away the const qualifier */
*foo_ptr = 20; /* This is undefined behavior */

return 0;
}

Quoting ISO/IEC 9899:201x, section 6.7.3 §2:

If an attempt is made to modify an object defined with a const-qualified type through use of an lvalue
with non-const-qualified type, the behavior is undefined. [...]

(1) In GCC this can throw the following warning: warning: assignment discards ‘const’ qualifier from pointer
target type [-Wdiscarded-qualifiers]

GoalKicker.com – C Notes for Professionals 174

Section 28.16: Reading an uninitialized object that is not
backed by memory
Version ≥ C11

Reading an object will cause undefined behavior, if the object is1:

uninitialized
defined with automatic storage duration
it's address is never taken

The variable a in the below example satisfies all those conditions:

void Function( void )

{
int a;
int b = a;
}

1 (Quoted from: ISO:IEC 9899:201X 6.3.2.1 Lvalues, arrays, and function designators 2)
If the lvalue designates an object of automatic storage duration that could have been declared with the register
storage class (never had its address taken), and that object is uninitialized (not declared with an initializer and no
assignment to it has been performed prior to use), the behavior is undefined.

Section 28.17: Addition or subtraction of pointer not properly

bounded
The following code has undefined behavior:

char buffer[6] = "hello";

char *ptr1 = buffer - 1; /* undefined behavior */
char *ptr2 = buffer + 5; /* OK, pointing to the '\0' inside the array */
char *ptr3 = buffer + 6; /* OK, pointing to just beyond */
char *ptr4 = buffer + 7; /* undefined behavior */

According to C11, if addition or subtraction of a pointer into, or just beyond, an array object and an integer type
produces a result that does not point into, or just beyond, the same array object, the behavior is undefined (6.5.6).

Additionally it is naturally undefined behavior to dereference a pointer that points to just beyond the array:

char buffer[6] = "hello";

char *ptr3 = buffer + 6; /* OK, pointing to just beyond */
char value = *ptr3; /* undefined behavior */

Section 28.18: Dereferencing a null pointer

This is an example of dereferencing a NULL pointer, causing undefined behavior.

int * pointer = NULL;

int value = *pointer; /* Dereferencing happens here */

A NULL pointer is guaranteed by the C standard to compare unequal to any pointer to a valid object, and
dereferencing it invokes undefined behavior.

GoalKicker.com – C Notes for Professionals 175

Section 28.19: Using ush on an input stream
The POSIX and C standards explicitly state that using fflush on an input stream is undefined behavior. The fflush
is defined only for output streams.

#include <stdio.h>

int main()
{
int i;
char input[4096];

scanf("%i", &i);
fflush(stdin); // <-- undefined behavior
gets(input);

return 0;
}

There is no standard way to discard unread characters from an input stream. On the other hand, some
implementations uses fflush to clear stdin buffer. Microsoft defines the behavior of fflush on an input stream: If
the stream is open for input, fflush clears the contents of the buffer. According to POSIX.1-2008, the behavior of
fflush is undefined unless the input file is seekable.

See Using fflush(stdin) for many more details.

Section 28.20: Inconsistent linkage of identifiers

extern int var;
static int var; /* Undefined behaviour */

C11, §6.2.2, 7 says:

If, within a translation unit, the same identifier appears with both internal and external linkage, the
behavior is undefined.

Note that if an prior declaration of an identifier is visible then it'll have the prior declaration's linkage. C11, §6.2.2, 4
allows it:

For an identifier declared with the storage-class specifier extern in a scope in which a prior declaration of
that identifier is visible,31) if the prior declaration specifies internal or external linkage, the linkage of the
identifier at the later declaration is the same as the linkage specified at the prior declaration. If no prior
declaration is visible, or if the prior declaration specifies no linkage, then the identifier has external
linkage.

/* 1. This is NOT undefined */

static int var;
extern int var;

/* 2. This is NOT undefined */

static int var;
static int var;

GoalKicker.com – C Notes for Professionals 176

/* 3. This is NOT undefined */
extern int var;
extern int var;

Section 28.21: Missing return statement in value returning

function
int foo(void) {
/* do stuff */
/* no return here */
}

int main(void) {
/* Trying to use the (not) returned value causes UB */
int value = foo();
return 0;
}

When a function is declared to return a value then it has to do so on every possible code path through it. Undefined
behavior occurs as soon as the caller (which is expecting a return value) tries to use the return value1.

Note that the undefined behaviour happens only if the caller attempts to use/access the value from the function.
For example,

int foo(void) {
/* do stuff */
/* no return here */
}

int main(void) {
/* The value (not) returned from foo() is unused. So, this program
* doesn't cause *undefined behaviour*. */
foo();
return 0;
}
Version ≥ C99

The main() function is an exception to this rule in that it is possible for it to be terminated without a return
statement because an assumed return value of 0 will automatically be used in this case2.

1 (ISO/IEC 9899:201x, 6.9.1/12)

If the } that terminates a function is reached, and the value of the function call is used by the caller, the
behavior is undefined.

2 (ISO/IEC 9899:201x, 5.1.2.2.3/1)

reaching the } that terminates the main function returns a value of 0.

Section 28.22: Division by zero

int x = 0;

GoalKicker.com – C Notes for Professionals 177

int y = 5 / x; /* integer division */

or

double x = 0.0;
double y = 5.0 / x; /* floating point division */

or

int x = 0;
int y = 5 % x; /* modulo operation */

For the second line in each example, where the value of the second operand (x) is zero, the behaviour is undefined.

Note that most implementations of floating point math will follow a standard (e.g. IEEE 754), in which case
operations like divide-by-zero will have consistent results (e.g., INFINITY) even though the C standard says the
operation is undefined.

Section 28.23: Conversion between pointer types produces

incorrectly aligned result
The following might have undefined behavior due to incorrect pointer alignment:

char *memory_block = calloc(sizeof(uint32_t) + 1, 1);

uint32_t *intptr = (uint32_t*)(memory_block + 1); /* possible undefined behavior */
uint32_t mvalue = *intptr;

The undefined behavior happens as the pointer is converted. According to C11, if a conversion between two pointer
types produces a result that is incorrectly aligned (6.3.2.3), the behavior is undefined. Here an uint32_t could require
alignment of 2 or 4 for example.

calloc on the other hand is required to return a pointer that is suitably aligned for any object type; thus
memory_block is properly aligned to contain an uint32_t in its initial part. Then, on a system where uint32_t has
required alignment of 2 or 4, memory_block + 1 will be an odd address and thus not properly aligned.

Observe that the C standard requests that already the cast operation is undefined. This is imposed because on
platforms where addresses are segmented, the byte address memory_block + 1 may not even have a proper
representation as an integer pointer.

Casting char * to pointers to other types without any concern to alignment requirements is sometimes incorrectly
used for decoding packed structures such as file headers or network packets.

You can avoid the undefined behavior arising from misaligned pointer conversion by using memcpy:

memcpy(&mvalue, memory_block + 1, sizeof mvalue);

Here no pointer conversion to uint32_t* takes place and the bytes are copied one by one.

This copy operation for our example only leads to valid value of mvalue because:

We used calloc, so the bytes are properly initialized. In our case all bytes have value 0, but any other proper
initialization would do.
uint32_t is an exact width type and has no padding bits
Any arbitrary bit pattern is a valid representation for any unsigned type.

GoalKicker.com – C Notes for Professionals 178

Section 28.24: Modifying the string returned by getenv,
strerror, and setlocale functions
Modifying the strings returned by the standard functions getenv(), strerror() and setlocale() is undefined. So,
implementations may use static storage for these strings.

The getenv() function, C11, §7.22.4.7, 4, says:

The getenv function returns a pointer to a string associated with the matched list member. The string
pointed to shall not be modified by the program, but may be overwritten by a subsequent call to the
getenv function.

The strerror() function, C11, §7.23.6.3, 4 says:

The strerror function returns a pointer to the string, the contents of which are localespecific. The array
pointed to shall not be modified by the program, but may be overwritten by a subsequent call to the
strerror function.

The setlocale() function, C11, §7.11.1.1, 8 says:

The pointer to string returned by the setlocale function is such that a subsequent call with that string
value and its associated category will restore that part of the program’s locale. The string pointed to shall
not be modified by the program, but may be overwritten by a subsequent call to the setlocale function.

Similarly the localeconv() function returns a pointer to struct lconv which shall not be modified.

The localeconv() function, C11, §7.11.2.1, 8 says:

The localeconv function returns a pointer to the filled-in object. The structure pointed to by the return
value shall not be modified by the program, but may be overwritten by a subsequent call to the
localeconv function.

GoalKicker.com – C Notes for Professionals 179

Chapter 29: Random Number Generation
Section 29.1: Basic Random Number Generation
The function rand() can be used to generate a pseudo-random integer value between 0 and RAND_MAX (0 and
RAND_MAX included).

srand(int) is used to seed the pseudo-random number generator. Each time rand() is seeded wih the same seed,
it must produce the same sequence of values. It should only be seeded once before calling rand(). It should not be
repeatedly seeded, or reseeded every time you wish to generate a new batch of pseudo-random numbers.

Standard practice is to use the result of time(NULL) as a seed. If your random number generator requires to have a
deterministic sequence, you can seed the generator with the same value on each program start. This is generally
not required for release code, but is useful in debug runs to make bugs reproducible.

It is advised to always seed the generator, if not seeded, it behaves as if it was seeded with srand(1).

#include <stdio.h>
#include <stdlib.h>
#include <time.h>

int main(void) {
int i;
srand(time(NULL));
i = rand();

printf("Random value between [0, %d]: %d\n", RAND_MAX, i);

return 0;
}

Possible output:

Random value between [0, 2147483647]: 823321433

Notes:

The C Standard does not guarantee the quality of the random sequence produced. In the past, some
implementations of rand() had serious issues in distribution and randomness of the generated numbers. The
usage of rand() is not recommended for serious random number generation needs, like cryptography.

Section 29.2: Permuted Congruential Generator

Here's a standalone random number generator that doesn't rely on rand() or similar library functions.

Why would you want such a thing? Maybe you don't trust your platform's builtin random number generator, or
maybe you want a reproducible source of randomness independent of any particular library implementation.

This code is PCG32 from pcg-random.org, a modern, fast, general-purpose RNG with excellent statistical properties.
It's not cryptographically secure, so don't use it for cryptography.

#include <stdint.h>

/* *Really* minimal PCG32 code / (c) 2014 M.E. O'Neill / pcg-random.org

* Licensed under Apache License 2.0 (NO WARRANTY, etc. see website) */

GoalKicker.com – C Notes for Professionals 180

typedef struct { uint64_t state; uint64_t inc; } pcg32_random_t;

uint32_t pcg32_random_r(pcg32_random_t* rng) {

uint64_t oldstate = rng->state;
/* Advance internal state */
rng->state = oldstate * 6364136223846793005ULL + (rng->inc | 1);
/* Calculate output function (XSH RR), uses old state for max ILP */
uint32_t xorshifted = ((oldstate >> 18u) ^ oldstate) >> 27u;
uint32_t rot = oldstate >> 59u;
return (xorshifted >> rot) | (xorshifted << ((-rot) & 31));
}

void pcg32_srandom_r(pcg32_random_t* rng, uint64_t initstate, uint64_t initseq) {

rng->state = 0U;
rng->inc = (initseq << 1u) | 1u;
pcg32_random_r(rng);
rng->state += initstate;
pcg32_random_r(rng);
}

And here's how to call it:

#include <stdio.h>
int main(void) {
pcg32_random_t rng; /* RNG state */
int i;

/* Seed the RNG */

pcg32_srandom_r(&rng, 42u, 54u);

/* Print some random 32-bit integers */

for (i = 0; i < 6; i++)
printf("0x%08x\n", pcg32_random_r(&rng));

return 0;
}

Section 29.3: Xorshift Generation

A good and easy alternative to the flawed rand() procedures, is xorshift, a class of pseudo-random number
generators discovered by George Marsaglia. The xorshift generator is among the fastest non-cryptographically-
secure random number generators. More information and other example implementaions are available on the
xorshift Wikipedia page

Example implementation

#include <stdint.h>

/* These state variables must be initialised so that they are not all zero. */
uint32_t w, x, y, z;

uint32_t xorshift128(void)
{
uint32_t t = x;
t ^= t << 11U;
t ^= t >> 8U;
x = y; y = z; z = w;
w ^= w >> 19U;
w ^= t;

GoalKicker.com – C Notes for Professionals 181

 return w;
}

Section 29.4: Restrict generation to a given range

Usually when generating random numbers it is useful to generate integers within a range, or a p value between 0.0
and 1.0. Whilst modulus operation can be used to reduce the seed to a low integer this uses the low bits, which
often go through a short cycle, resulting in a slight skewing of distribution if N is large in proportion to RAND_MAX.

The macro

#define uniform() (rand() / (RAND_MAX + 1.0))

produces a p value on 0.0 to 1.0 - epsilon, so

i = (int)(uniform() * N)

will set i to a uniform random number within the range 0 to N - 1.

Unfortunately there is a technical flaw, in that RAND_MAX is permitted to be larger than a variable of type double
can accurately represent. This means that RAND_MAX + 1.0 evaluates to RAND_MAX and the function occasionally
returns unity. This is unlikely however.

GoalKicker.com – C Notes for Professionals 182

Chapter 30: Preprocessor and Macros
All preprocessor commands begins with the hash (pound) symbol #. A C macro is just a preprocessor command
that is defined using the #define preprocessor directive. During the preprocessing stage, the C preprocessor (a part
of the C compiler) simply substitutes the body of the macro wherever its name appears.

Section 30.1: Header Include Guards

Pretty much every header file should follow the include guard idiom:

my-header-file.h

#ifndef MY_HEADER_FILE_H
#define MY_HEADER_FILE_H

// Code body for header file

#endif

This ensures that when you #include "my-header-file.h" in multiple places, you don't get duplicate declarations
of functions, variables, etc. Imagine the following hierarchy of files:

header-1.h

typedef struct {
…
} MyStruct;

int myFunction(MyStruct *value);

header-2.h

#include "header-1.h"

int myFunction2(MyStruct *value);

main.c

#include "header-1.h"
#include "header-2.h"

int main() {
// do something
}

This code has a serious problem: the detailed contents of MyStruct is defined twice, which is not allowed. This
would result in a compilation error that can be difficult to track down, since one header file includes another. If you
instead did it with header guards:

header-1.h

#ifndef HEADER_1_H
#define HEADER_1_H

typedef struct {

GoalKicker.com – C Notes for Professionals 183

 …
} MyStruct;

int myFunction(MyStruct *value);

#endif

header-2.h

#ifndef HEADER_2_H
#define HEADER_2_H

#include "header-1.h"

int myFunction2(MyStruct *value);

#endif

main.c

#include "header-1.h"
#include "header-2.h"

int main() {
// do something
}

This would then expand to:

#ifndef HEADER_1_H
#define HEADER_1_H

typedef struct {
…
} MyStruct;

int myFunction(MyStruct *value);

#endif

#ifndef HEADER_2_H
#define HEADER_2_H

#ifndef HEADER_1_H // Safe, since HEADER_1_H was #define'd before.

#define HEADER_1_H

typedef struct {
…
} MyStruct;

int myFunction(MyStruct *value);

#endif

int myFunction2(MyStruct *value);

#endif

int main() {
// do something

GoalKicker.com – C Notes for Professionals 184

}

When the compiler reaches the second inclusion of header-1.h, HEADER_1_H was already defined by the previous
inclusion. Ergo, it boils down to the following:

#define HEADER_1_H

typedef struct {
…
} MyStruct;

int myFunction(MyStruct *value);

#define HEADER_2_H

int myFunction2(MyStruct *value);

int main() {
// do something
}

And thus there is no compilation error.

Note: There are multiple different conventions for naming the header guards. Some people like to name it
HEADER_2_H_, some include the project name like MY_PROJECT_HEADER_2_H. The important thing is to ensure that the
convention you follow makes it so that each file in your project has a unique header guard.

If the structure details were not included in the header, the type declared would be incomplete or an opaque type.
Such types can be useful, hiding implementation details from users of the functions. For many purposes, the FILE
type in the standard C library can be regarded as an opaque type (though it usually isn't opaque so that macro
implementations of the standard I/O functions can make use of the internals of the structure). In that case, the
header-1.h could contain:

#ifndef HEADER_1_H
#define HEADER_1_H

typedef struct MyStruct MyStruct;

int myFunction(MyStruct *value);

#endif

Note that the structure must have a tag name (here MyStruct — that's in the tags namespace, separate from the
ordinary identifiers namespace of the typedef name MyStruct), and that the { … } is omitted. This says "there is a
structure type struct MyStruct and there is an alias for it MyStruct".

In the implementation file, the details of the structure can be defined to make the type complete:

struct MyStruct {
…
};

If you are using C11, you could repeat the typedef struct MyStruct MyStruct; declaration without causing a
compilation error, but earlier versions of C would complain. Consequently, it is still best to use the include guard

GoalKicker.com – C Notes for Professionals 185

idiom, even though in this example, it would be optional if the code was only ever compiled with compilers that
supported C11.

Many compilers support the #pragma once directive, which has the same results:

my-header-file.h

#pragma once

// Code for header file

However, #pragma once is not part of the C standard, so the code is less portable if you use it.

A few headers do not use the include guard idiom. One specific example is the standard <assert.h> header. It may
be included multiple times in a single translation unit, and the effect of doing so depends on whether the macro
NDEBUG is defined each time the header is included. You may occasionally have an analogous requirement; such
cases will be few and far between. Ordinarily, your headers should be protected by the include guard idiom.

Section 30.2: #if 0 to block out code sections

If there are sections of code that you are considering removing or want to temporarily disable, you can comment it
out with a block comment.

/* Block comment around whole function to keep it from getting used.

* What's even the purpose of this function?
int myUnusedFunction(void)
{
int i = 5;
return i;
}
*/

However, if the source code you have surrounded with a block comment has block style comments in the source,
the ending */ of the existing block comments can cause your new block comment to be invalid and cause
compilation problems.

/* Block comment around whole function to keep it from getting used.

* What's even the purpose of this function?
int myUnusedFunction(void)
{
int i = 5;

/* Return 5 */
return i;
}
*/

In the previous example, the last two lines of the function and the last '*/' are seen by the compiler, so it would
compile with errors. A safer method is to use an #if 0 directive around the code you want to block out.

#if 0
/* #if 0 evaluates to false, so everything between here and the #endif are
* removed by the preprocessor. */
int myUnusedFunction(void)
{

GoalKicker.com – C Notes for Professionals 186

 int i = 5;
return i;
}
#endif

A benefit with this is that when you want to go back and find the code, it's much easier to do a search for "#if 0"
than searching all your comments.

Another very important benefit is that you can nest commenting out code with #if 0. This cannot be done with
comments.

An alternative to using #if 0 is to use a name that will not be #defined but is more descriptive of why the code is
being blocked out. For instance if there is a function that seems to be useless dead code you might use #if
defined(POSSIBLE_DEAD_CODE) or #if defined(FUTURE_CODE_REL_020201) for code needed once other
functionality is in place or something similar. Then when going back through to remove or enable that source, those
sections of source are easy to find.

Section 30.3: Function-like macros

Function-like macros are similar to inline functions, these are useful in some cases, such as temporary debug log:

#ifdef DEBUG
# define LOGFILENAME "/tmp/logfile.log"

# define LOG(str) do { \
FILE *fp = fopen(LOGFILENAME, "a"); \
if (fp) { \
fprintf(fp, "%s:%d %s\n", __FILE__, __LINE__, \
/* don't print null pointer */ \
str ?str :"<null>"); \
fclose(fp); \
} \
else { \
perror("Opening '" LOGFILENAME "' failed"); \
} \
} while (0)
#else
/* Make it a NOOP if DEBUG is not defined. */
# define LOG(LINE) (void)0
#endif

#include <stdio.h>

int main(int argc, char* argv[])

{
if (argc > 1)
LOG("There are command line arguments");
else
LOG("No command line arguments");
return 0;
}

Here in both cases (with DEBUG or not) the call behaves the same way as a function with void return type. This
ensures that the if/else conditionals are interpreted as expected.

In the DEBUG case this is implemented through a do { ... } while(0) construct. In the other case, (void)0 is a
statement with no side effect that is just ignored.

GoalKicker.com – C Notes for Professionals 187

An alternative for the latter would be

#define LOG(LINE) do { /* empty */ } while (0)

such that it is in all cases syntactically equivalent to the first.

If you use GCC, you can also implement a function-like macro that returns result using a non-standard GNU
extension — statement expressions. For example:

#include <stdio.h>

#define POW(X, Y) \
({ \
int i, r = 1; \
for (i = 0; i < Y; ++i) \
r *= X; \
r; \ // returned value is result of last operation
})

int main(void)
{
int result;

result = POW(2, 3);

printf("Result: %d\n", result);
}

Section 30.4: Source file inclusion

The most common uses of #include preprocessing directives are as in the following:

#include <stdio.h>
#include "myheader.h"

#include replaces the statement with the contents of the file referred to. Angle brackets (<>) refer to header files
installed on the system, while quotation marks ("") are for user-supplied files.

Macros themselves can expand other macros once, as this example illustrates:

#if VERSION == 1
#define INCFILE "vers1.h"
#elif VERSION == 2
#define INCFILE "vers2.h"
/* and so on */
#else
#define INCFILE "versN.h"
#endif
/* ... */
#include INCFILE

Section 30.5: Conditional inclusion and conditional function

signature modification
To conditionally include a block of code, the preprocessor has several directives (e.g #if, #ifdef, #else, #endif,
etc).

/* Defines a conditional `printf` macro, which only prints if `DEBUG`

GoalKicker.com – C Notes for Professionals 188

 * has been defined
*/
#ifdef DEBUG
#define DLOG(x) (printf(x))
#else
#define DLOG(x)
#endif

Normal C relational operators may be used for the #if condition

#if __STDC_VERSION__ >= 201112L

/* Do stuff for C11 or higher */
#elif __STDC_VERSION__ >= 199901L
/* Do stuff for C99 */
#else
/* Do stuff for pre C99 */
#endif

The #if directives behaves similar to the C if statement, it shall only contain integral constant expressions, and no
casts. It supports one additional unary operator, defined( identifier ), which returns 1 if the identifier is
defined, and 0 otherwise.

#if defined(DEBUG) && !defined(QUIET)

#define DLOG(x) (printf(x))
#else
#define DLOG(x)
#endif

Conditional Function Signature Modification

In most cases a release build of an application is expected to have as little overhead as possible. However during
testing of an interim build, additional logs and information about problems found can be helpful.

For example assume there is some function SHORT SerOpPluAllRead(PLUIF *pPif, USHORT usLockHnd) which
when doing a test build it is desired will generate a log about its use. However this function is used in multiple
places and it is desired that when generating the log, part of the information is to know where is the function being
called from.

So using conditional compilation you can have something like the following in the include file declaring the function.
This replaces the standard version of the function with a debug version of the function. The preprocessor is used to
replace calls to the function SerOpPluAllRead() with calls to the function SerOpPluAllRead_Debug() with two
additional arguments, the name of the file and the line number of where the function is used.

Conditional compilation is used to choose whether to override the standard function with a debug version or not.

#if 0
// function declaration and prototype for our debug version of the function.
SHORT SerOpPluAllRead_Debug(PLUIF *pPif, USHORT usLockHnd, char *aszFilePath, int nLineNo);

// macro definition to replace function call using old name with debug function with additional
arguments.
#define SerOpPluAllRead(pPif,usLock) SerOpPluAllRead_Debug(pPif,usLock,__FILE__,__LINE__)
#else
// standard function declaration that is normally used with builds.
SHORT SerOpPluAllRead(PLUIF *pPif, USHORT usLockHnd);
#endif

GoalKicker.com – C Notes for Professionals 189

This allows you to override the standard version of the function SerOpPluAllRead() with a version that will provide
the name of the file and line number in the file of where the function is called.

There is one important consideration: any file using this function must include the header file where this approach is
used in order for the preprocessor to modify the function. Otherwise you will see a linker error.

The definition of the function would look something like the following. What this source does is to request that the
preprocessor rename the function SerOpPluAllRead() to be SerOpPluAllRead_Debug() and to modify the
argument list to include two additional arguments, a pointer to the name of the file where the function was called
and the line number in the file at which the function is used.

#if defined(SerOpPluAllRead)
// forward declare the replacement function which we will call once we create our log.
SHORT SerOpPluAllRead_Special(PLUIF *pPif, USHORT usLockHnd);

SHORT SerOpPluAllRead_Debug(PLUIF *pPif, USHORT usLockHnd, char *aszFilePath, int nLineNo)

{
int iLen = 0;
char xBuffer[256];

// only print the last 30 characters of the file name to shorten the logs.
iLen = strlen (aszFilePath);
if (iLen > 30) {
iLen = iLen - 30;
}
else {
iLen = 0;
}

sprintf (xBuffer, "SerOpPluAllRead_Debug(): husHandle = %d, File %s, lineno = %d",

pPif->husHandle, aszFilePath + iLen, nLineNo);
IssueDebugLog(xBuffer);

// now that we have issued the log, continue with standard processing.
return SerOpPluAllRead_Special(pPif, usLockHnd);
}

// our special replacement function name for when we are generating logs.
SHORT SerOpPluAllRead_Special(PLUIF *pPif, USHORT usLockHnd)
#else
// standard, normal function name (signature) that is replaced with our debug version.
SHORT SerOpPluAllRead(PLUIF *pPif, USHORT usLockHnd)
#endif
{
if (STUB_SELF == SstReadAsMaster()) {
return OpPluAllRead(pPif, usLockHnd);
}
return OP_NOT_MASTER;
}

Section 30.6: __cplusplus for using C externals in C++ code

compiled with C++ - name mangling
There are times when an include file has to generate different output from the preprocessor depending on whether
the compiler is a C compiler or a C++ compiler due to language differences.

For example a function or other external is defined in a C source file but is used in a C++ source file. Since C++ uses
name mangling (or name decoration) in order to generate unique function names based on function argument

GoalKicker.com – C Notes for Professionals 190

types, a C function declaration used in a C++ source file will cause link errors. The C++ compiler will modify the
specified external name for the compiler output using the name mangling rules for C++. The result is link errors due
to externals not found when the C++ compiler output is linked with the C compiler output.

Since C compilers do not do name mangling but C++ compilers do for all external labels (function names or variable
names) generated by the C++ compiler, a predefined preprocessor macro, __cplusplus, was introduced to allow for
compiler detection.

In order to work around this problem of incompatible compiler output for external names between C and C++, the
macro __cplusplus is defined in the C++ Preprocessor and is not defined in the C Preprocessor. This macro name
can be used with the conditional preprocessor #ifdef directive or #if with the defined() operator to tell whether
a source code or include file is being compiled as C++ or C.

#ifdef __cplusplus
printf("C++\n");
#else
printf("C\n");
#endif

Or you could use

#if defined(__cplusplus)
printf("C++\n");
#else
printf("C\n");
#endif

In order to specify the correct function name of a function from a C source file compiled with the C compiler that is
being used in a C++ source file you could check for the __cplusplus defined constant in order to cause the extern
"C" { /* ... */ }; to be used to declare C externals when the header file is included in a C++ source file.
However when compiled with a C compiler, the extern "C" { */ ... */ }; is not used. This conditional
compilation is needed because extern "C" { /* ... */ }; is valid in C++ but not in C.

#ifdef __cplusplus
// if we are being compiled with a C++ compiler then declare the
// following functions as C functions to prevent name mangling.
extern "C" {
#endif

// exported C function list.

int foo (void);

#ifdef __cplusplus
// if this is a C++ compiler, we need to close off the extern declaration.
};
#endif

Section 30.7: Token pasting

Token pasting allows one to glue together two macro arguments. For example, front##back yields frontback. A
famous example is Win32's <TCHAR.H> header. In standard C, one can write L"string" to declare a wide character
string. However, Windows API allows one to convert between wide character strings and narrow character strings
simply by #defineing UNICODE. In order to implement the string literals, TCHAR.H uses this

#ifdef UNICODE
#define TEXT(x) L##x

GoalKicker.com – C Notes for Professionals 191

#endif

Whenever a user writes TEXT("hello, world"), and UNICODE is defined, the C preprocessor concatenates L and
the macro argument. L concatenated with "hello, world" gives L"hello, world".

Section 30.8: Predefined Macros

A predefined macro is a macro that is already understood by the C pre processor without the program needing to
define it. Examples include

Mandatory Pre-Defined Macros

__FILE__, which gives the file name of the current source file (a string literal),
__LINE__ for the current line number (an integer constant),
__DATE__ for the compilation date (a string literal),
__TIME__ for the compilation time (a string literal).

There's also a related predefined identifier, __func__ (ISO/IEC 9899:2011 §6.4.2.2), which is not a macro:

The identifier __func__ shall be implicitly declared by the translator as if, immediately following the
opening brace of each function definition, the declaration:

static const char __func__[] = "function-name";

appeared, where function-name is the name of the lexically-enclosing function.

__FILE__, __LINE__ and __func__ are especially useful for debugging purposes. For example:

fprintf(stderr, "%s: %s: %d: Denominator is 0", __FILE__, __func__, __LINE__);

Pre-C99 compilers, may or may not support __func__ or may have a macro that acts the same that is named
differently. For example, gcc used __FUNCTION__ in C89 mode.

The below macros allow to ask for detail on the implementation:

__STDC_VERSION__ The version of the C Standard implemented. This is a constant integer using the format
yyyymmL (the value 201112L for C11, the value 199901L for C99; it wasn't defined for C89/C90)
__STDC_HOSTED__ 1 if it's a hosted implementation, else 0.
__STDC__ If 1, the implementation conforms to the C Standard.

Other Pre-Defined Macros (non mandatory)

ISO/IEC 9899:2011 §6.10.9.2 Environment macros:

__STDC_ISO_10646__ An integer constant of the form yyyymmL (for example, 199712L). If this
symbol is defined, then every character in the Unicode required set, when stored in an object of
type wchar_t, has the same value as the short identifier of that character. The Unicode required set
consists of all the characters that are defined by ISO/IEC 10646, along with all amendments and
technical corrigenda, as of the specified year and month. If some other encoding is used, the macro
shall not be defined and the actual encoding used is implementation-defined.

__STDC_MB_MIGHT_NEQ_WC__ The integer constant 1, intended to indicate that, in the encoding for
wchar_t, a member of the basic character set need not have a code value equal to its value when

GoalKicker.com – C Notes for Professionals 192

 used as the lone character in an integer character constant.

__STDC_UTF_16__ The integer constant 1, intended to indicate that values of type char16_t are
UTF−16 encoded. If some other encoding is used, the macro shall not be defined and the actual
encoding used is implementation-defined.

__STDC_UTF_32__ The integer constant 1, intended to indicate that values of type char32_t are
UTF−32 encoded. If some other encoding is used, the macro shall not be defined and the actual
encoding used is implementation-defined.

ISO/IEC 9899:2011 §6.10.8.3 Conditional feature macros

__STDC_ANALYZABLE__ The integer constant 1, intended to indicate conformance to the

specifications in annex L (Analyzability).
__STDC_IEC_559__ The integer constant 1, intended to indicate conformance to the specifications
in annex F (IEC 60559 floating-point arithmetic).
__STDC_IEC_559_COMPLEX__ The integer constant 1, intended to indicate adherence to the
specifications in annex G (IEC 60559 compatible complex arithmetic).
__STDC_LIB_EXT1__ The integer constant 201112L, intended to indicate support for the extensions
defined in annex K (Bounds-checking interfaces).
__STDC_NO_ATOMICS__ The integer constant 1, intended to indicate that the implementation does
not support atomic types (including the _Atomic type qualifier) and the <stdatomic.h> header.
__STDC_NO_COMPLEX__ The integer constant 1, intended to indicate that the implementation does
not support complex types or the <complex.h> header.
__STDC_NO_THREADS__ The integer constant 1, intended to indicate that the implementation does
not support the <threads.h> header.
__STDC_NO_VLA__ The integer constant 1, intended to indicate that the implementation does not
support variable length arrays or variably modified types.

Section 30.9: Variadic arguments macro

Version ≥ C99

Macros with variadic args:

Let's say you want to create some print-macro for debugging your code, let's take this macro as an example:

#define debug_print(msg) printf("%s:%d %s", __FILE__, __LINE__, msg)

Some examples of usage:

The function somefunc() returns -1 if failed and 0 if succeeded, and it is called from plenty different places within
the code:

int retVal = somefunc();

if(retVal == -1)
{
debug_printf("somefunc() has failed");
}

/* some other code */

GoalKicker.com – C Notes for Professionals 193

 retVal = somefunc();

if(retVal == -1)
{
debug_printf("somefunc() has failed");
}

What happens if the implementation of somefunc() changes, and it now returns different values matching different
possible error types? You still want use the debug macro and print the error value.

debug_printf(retVal); /* this would obviously fail */

debug_printf("%d",retVal); /* this would also fail */

To solve this problem the __VA_ARGS__ macro was introduced. This macro allows multiple parameters X-macro's:

Example:

#define debug_print(msg, ...) printf(msg, __VA_ARGS__) \

printf("\nError occurred in file:line (%s:%d)\n", __FILE__, __LINE)

Usage:

int retVal = somefunc();

debug_print("retVal of somefunc() is-> %d", retVal);

This macro allows you to pass multiple parameters and print them, but now it forbids you from sending any
parameters at all.

debug_print("Hey");

This would raise some syntax error as the macro expects at least one more argument and the pre-processor would
not ignore the lack of comma in the debug_print() macro. Also debug_print("Hey",); would raise a syntax error
as you cant keep the argument passed to macro empty.

To solve this, ##__VA_ARGS__ macro was introduced, this macro states that if no variable arguments exist, the
comma is deleted by the pre-processor from code.

Example:

#define debug_print(msg, ...) printf(msg, ##__VA_ARGS__) \

printf("\nError occurred in file:line (%s:%d)\n", __FILE__, __LINE)

Usage:

debug_print("Ret val of somefunc()?");

debug_print("%d",somefunc());

Section 30.10: Macro Replacement

The simplest form of macro replacement is to define a manifest constant, as in

#define ARRSIZE 100

int array[ARRSIZE];

GoalKicker.com – C Notes for Professionals 194

This defines a function-like macro that multiplies a variable by 10 and stores the new value:

#define TIMES10(A) ((A) *= 10)

double b = 34;
int c = 23;

TIMES10(b); // good: ((b) *= 10);

TIMES10(c); // good: ((c) *= 10);
TIMES10(5); // bad: ((5) *= 10);

The replacement is done before any other interpretation of the program text. In the first call to TIMES10 the name A
from the definition is replaced by b and the so expanded text is then put in place of the call. Note that this
definition of TIMES10 is not equivalent to

#define TIMES10(A) ((A) = (A) * 10)

because this could evaluate the replacement of A, twice, which can have unwanted side effects.

The following defines a function-like macro which value is the maximum of its arguments. It has the advantages of
working for any compatible types of the arguments and of generating in-line code without the overhead of function
calling. It has the disadvantages of evaluating one or the other of its arguments a second time (including side
effects) and of generating more code than a function if invoked several times.

#define max(a, b) ((a) > (b) ? (a) : (b))

int maxVal = max(11, 43); /* 43 */

int maxValExpr = max(11 + 36, 51 - 7); /* 47 */

/* Should not be done, due to expression being evaluated twice */

int j = 0, i = 0;
int sideEffect = max(++i, ++j); /* i == 4 */

Because of this, such macros that evaluate their arguments multiple times are usually avoided in production code.
Since C11 there is the _Generic feature that allows to avoid such multiple invocations.

The abundant parentheses in the macro expansions (right hand side of the definition) ensure that the arguments
and the resulting expression are bound properly and fit well into the context in which the macro is called.

Section 30.11: Error directive

If the preprocessor encounters an #error directive, compilation is halted and the diagnostic message included is
printed.

#define DEBUG

#ifdef DEBUG
#error "Debug Builds Not Supported"
#endif

int main(void) {
return 0;
}

Possible output:

GoalKicker.com – C Notes for Professionals 195

$ gcc error.c
error.c: error: #error "Debug Builds Not Supported"

Section 30.12: FOREACH implementation

We can also use macros for making code easier to read and write. For example we can implement macros for
implementing the foreach construct in C for some data structures like singly- and doubly-linked lists, queues, etc.

Here is a small example.

#include <stdio.h>
#include <stdlib.h>

struct LinkedListNode
{
int data;
struct LinkedListNode *next;
};

#define FOREACH_LIST(node, list) \

for (node=list; node; node=node->next)

/* Usage */
int main(void)
{
struct LinkedListNode *list, **plist = &list, *node;
int i;

for (i=0; i<10; i++)

{
*plist = malloc(sizeof(struct LinkedListNode));
(*plist)->data = i;
(*plist)->next = NULL;
plist = &(*plist)->next;
}

/* printing the elements here */

FOREACH_LIST(node, list)
{
printf("%d\n", node->data);
}
}

You can make a standard interface for such data-structures and write a generic implementation of FOREACH as:

#include <stdio.h>
#include <stdlib.h>

typedef struct CollectionItem_

{
int data;
struct CollectionItem_ *next;
} CollectionItem;

typedef struct Collection_

{
/* interface functions */
void* (*first)(void *coll);
void* (*last) (void *coll);
void* (*next) (void *coll, CollectionItem *currItem);

GoalKicker.com – C Notes for Professionals 196

 CollectionItem *collectionHead;
/* Other fields */
} Collection;

/* must implement */
void *first(void *coll)
{
return ((Collection*)coll)->collectionHead;
}

/* must implement */
void *last(void *coll)
{
return NULL;
}

/* must implement */
void *next(void *coll, CollectionItem *curr)
{
return curr->next;
}

CollectionItem *new_CollectionItem(int data)

{
CollectionItem *item = malloc(sizeof(CollectionItem));
item->data = data;
item->next = NULL;
return item;
}

void Add_Collection(Collection *coll, int data)

{
CollectionItem **item = &coll->collectionHead;
while(*item)
item = &(*item)->next;
(*item) = new_CollectionItem(data);
}

Collection *new_Collection()
{
Collection *nc = malloc(sizeof(Collection));
nc->first = first;
nc->last = last;
nc->next = next;
return nc;
}

/* generic implementation */
#define FOREACH(node, collection) \
for (node = (collection)->first(collection); \
node != (collection)->last(collection); \
node = (collection)->next(collection, node))

int main(void)
{
Collection *coll = new_Collection();
CollectionItem *node;
int i;

for(i=0; i<10; i++)

{
Add_Collection(coll, i);

GoalKicker.com – C Notes for Professionals 197

 }

/* printing the elements here */

FOREACH(node, coll)
{
printf("%d\n", node->data);
}
}

To use this generic implementation just implement these functions for your data structure.

1. void* (*first)(void *coll);

2. void* (*last) (void *coll);
3. void* (*next) (void *coll, CollectionItem *currItem);

GoalKicker.com – C Notes for Professionals 198

Chapter 31: Signal handling
Parameter Details
The signal to set the signal handler to, one of SIGABRT, SIGFPE, SIGILL, SIGTERM, SIGINT, SIGSEGV or
sig
some implementation defined value
The signal handler, which is either of the following: SIG_DFL, for the default handler, SIG_IGN to ignore
func
the signal, or a function pointer with the signature void foo(int sig);.

Section 31.1: Signal Handling with “signal()”

Signal numbers can be synchronous (like SIGSEGV – segmentation fault) when they are triggered by a
malfunctioning of the program itself or asynchronous (like SIGINT - interactive attention) when they are initiated
from outside the program, e.g by a keypress as Cntrl-C.

The signal() function is part of the ISO C standard and can be used to assign a function to handle a specific signal

#include <stdio.h> /* printf() */

#include <stdlib.h> /* abort() */
#include <signal.h> /* signal() */

void handler_nonportable(int sig)

{
/* undefined behavior, maybe fine on specific platform */
printf("Catched: %d\n", sig);

/* abort is safe to call */

abort();
}

sig_atomic_t volatile finished = 0;

void handler(int sig)

{
switch (sig) {
/* hardware interrupts should not return */
case SIGSEGV:
case SIGFPE:
case SIGILL:
Version ≥ C11
/* quick_exit is safe to call */
quick_exit(EXIT_FAILURE);
Version < C11
/* use _Exit in pre-C11 */
_Exit(EXIT_FAILURE);

default:
/* Reset the signal to the default handler,
so we will not be called again if things go
wrong on return. */
signal(sig, SIG_DFL);
/* let everybody know that we are finished */
finished = sig;
return;
}
}

int main(void)
{

GoalKicker.com – C Notes for Professionals 199

 /* Catch the SIGSEGV signal, raised on segmentation faults (i.e NULL ptr access */
if (signal(SIGSEGV, &handler) == SIG_ERR) {
perror("could not establish handler for SIGSEGV");
return EXIT_FAILURE;
}

/* Catch the SIGTERM signal, termination request */

if (signal(SIGTERM, &handler) == SIG_ERR) {
perror("could not establish handler for SIGTERM");
return EXIT_FAILURE;
}

/* Ignore the SIGINT signal, by setting the handler to `SIG_IGN`. */

signal(SIGINT, SIG_IGN);

/* Do something that takes some time here, and leaves

the time to terminate the program from the keyboard. */

/* Then: */

if (finished) {
fprintf(stderr, "we have been terminated by signal %d\n", (int)finished);
return EXIT_FAILURE;
}

/* Try to force a segmentation fault, and raise a SIGSEGV */

{
char* ptr = 0;
*ptr = 0;
}

/* This should never be executed */

return EXIT_SUCCESS;
}

Using signal() imposes important limitations what you are allowed to do inside the signal handlers, see the
remarks for further information.

POSIX recommends the usage of sigaction() instead of signal(), due to its underspecified behavior and
significant implementation variations. POSIX also defines many more signals than ISO C standard, including
SIGUSR1 and SIGUSR2, which can be used freely by the programmer for any purpose.

GoalKicker.com – C Notes for Professionals 200

Chapter 32: Variable arguments
Parameter Details
va_list ap argument pointer, current position in the list of variadic arguments
name of last non-variadic function argument, so the compiler finds the correct place to start
last processing variadic arguments; may not be declared as a register variable, a function, or an array
type
type promoted type of the variadic argument to read (e.g. int for a short int argument)
va_list src current argument pointer to copy
va_list dst new argument list to be filled in

Variable arguments are used by functions in the printf family (printf, fprintf, etc) and others to allow a function
to be called with a different number of arguments each time, hence the name varargs.

To implement functions using the variable arguments feature, use #include <stdarg.h>.

To call functions which take a variable number of arguments, ensure there is a full prototype with the trailing
ellipsis in scope: void err_exit(const char *format, ...); for example.

Section 32.1: Using an explicit count argument to determine

the length of the va_list
With any variadic function, the function must know how to interpret the variable arguments list. With the printf()
or scanf() functions, the format string tells the function what to expect.

The simplest technique is to pass an explicit count of the other arguments (which are normally all the same type).
This is demonstrated in the variadic function in the code below which calculates the sum of a series of integers,
where there may be any number of integers but that count is specified as an argument prior to the variable
argument list.

#include <stdio.h>
#include <stdarg.h>

/* first arg is the number of following int args to sum. */

int sum(int n, ...) {
int sum = 0;
va_list it; /* hold information about the variadic argument list. */

va_start(it, n); /* start variadic argument processing */

while (n--)
sum += va_arg(it, int); /* get and sum the next variadic argument */
va_end(it); /* end variadic argument processing */

return sum;
}

int main(void)
{
printf("%d\n", sum(5, 1, 2, 3, 4, 5)); /* prints 15 */
printf("%d\n", sum(10, 5, 9, 2, 5, 111, 6666, 42, 1, 43, -6218)); /* prints 666 */
return 0;
}

GoalKicker.com – C Notes for Professionals 201

Section 32.2: Using terminator values to determine the end of
va_list
With any variadic function, the function must know how to interpret the variable arguments list. The “traditional”
approach (exemplified by printf) is to specify number of arguments up front. However, this is not always a good
idea:

/* First argument specifies the number of parameters; the remainder are also int */
extern int sum(int n, ...);

/* But it's far from obvious from the code. */

sum(5, 2, 1, 4, 3, 6)

/* What happens if i.e. one argument is removed later on? */

sum(5, 2, 1, 3, 6) /* Disaster */

Sometimes it's more robust to add an explicit terminator, exemplified by the POSIX execlp() function. Here's
another function to calculate the sum of a series of double numbers:

#include <stdarg.h>
#include <stdio.h>
#include <math.h>

/* Sums args up until the terminator NAN */

double sum (double x, ...) {
double sum = 0;
va_list va;

va_start(va, x);
for (; !isnan(x); x = va_arg(va, double)) {
sum += x;
}
va_end(va);

return sum;
}

int main (void) {

printf("%g\n", sum(5., 2., 1., 4., 3., 6., NAN));
printf("%g\n", sum(1, 0.5, 0.25, 0.125, 0.0625, 0.03125, NAN));
}

Good terminator values:

integer (supposed to be all positive or non-negative) — 0 or -1

floating point types — NAN
pointer types — NULL
enumerator types — some special value

Section 32.3: Implementing functions with a `printf()`-like

interface
One common use of variable-length argument lists is to implement functions that are a thin wrapper around the
printf() family of functions. One such example is a set of error reporting functions.

errmsg.h

GoalKicker.com – C Notes for Professionals 202

#ifndef ERRMSG_H_INCLUDED
#define ERRMSG_H_INCLUDED

#include <stdarg.h>
#include <stdnoreturn.h> // C11

void verrmsg(int errnum, const char *fmt, va_list ap);

noreturn void errmsg(int exitcode, int errnum, const char *fmt, ...);
void warnmsg(int errnum, const char *fmt, ...);

#endif

This is a bare-bones example; such packages can be much elaborate. Normally, programmers will use either
errmsg() or warnmsg(), which themselves use verrmsg() internally. If someone comes up with a need to do more,
though, then the exposed verrmsg() function will be useful. You could avoid exposing it until you have a need for it
(YAGNI — you aren't gonna need it), but the need will arise eventually (you are gonna need it — YAGNI).

errmsg.c

This code only needs to forward the variadic arguments to the vfprintf() function for outputting to standard
error. It also reports the system error message corresponding to the system error number (errno) passed to the
functions.

#include "errmsg.h"
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

void
verrmsg(int errnum, const char *fmt, va_list ap)
{
if (fmt)
vfprintf(stderr, fmt, ap);
if (errnum != 0)
fprintf(stderr, ": %s", strerror(errnum));
putc('\n', stderr);
}

void
errmsg(int exitcode, int errnum, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
verrmsg(errnum, fmt, ap);
va_end(ap);
exit(exitcode);
}

void
warnmsg(int errnum, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
verrmsg(errnum, fmt, ap);
va_end(ap);
}

Using errmsg.h

GoalKicker.com – C Notes for Professionals 203

Now you can use those functions as follows:

#include "errmsg.h"
#include <errno.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>

int main(int argc, char **argv)

{
char buffer[BUFSIZ];
int fd;
if (argc != 2)
{
fprintf(stderr, "Usage: %s filename\n", argv[0]);
exit(EXIT_FAILURE);
}
const char *filename = argv[1];

if ((fd = open(filename, O_RDONLY)) == -1)

errmsg(EXIT_FAILURE, errno, "cannot open %s", filename);
if (read(fd, buffer, sizeof(buffer)) != sizeof(buffer))
errmsg(EXIT_FAILURE, errno, "cannot read %zu bytes from %s", sizeof(buffer), filename);
if (close(fd) == -1)
warnmsg(errno, "cannot close %s", filename);
/* continue the program */
return 0;
}

If either the open() or read() system calls fails, the error is written to standard error and the program exits with
exit code 1. If the close() system call fails, the error is merely printed as a warning message, and the program
continues.

Checking the correct use of printf() formats

If you are using GCC (the GNU C Compiler, which is part of the GNU Compiler Collection), or using Clang, then you
can have the compiler check that the arguments you pass to the error message functions match what printf()
expects. Since not all compilers support the extension, it needs to be compiled conditionally, which is a little bit
fiddly. However, the protection it gives is worth the effort.

First, we need to know how to detect that the compiler is GCC or Clang emulating GCC. The answer is that GCC
defines __GNUC__ to indicate that.

See common function attributes for information about the attributes — specifically the format attribute.

Rewritten errmsg.h

#ifndef ERRMSG_H_INCLUDED
#define ERRMSG_H_INCLUDED

#include <stdarg.h>
#include <stdnoreturn.h> // C11

#if !defined(PRINTFLIKE)
#if defined(__GNUC__)
#define PRINTFLIKE(n,m) __attribute__((format(printf,n,m)))
#else
#define PRINTFLIKE(n,m) /* If only */

GoalKicker.com – C Notes for Professionals 204

#endif /* __GNUC__ */
#endif /* PRINTFLIKE */

void verrmsg(int errnum, const char *fmt, va_list ap);

void noreturn errmsg(int exitcode, int errnum, const char *fmt, ...)
PRINTFLIKE(3, 4);
void warnmsg(int errnum, const char *fmt, ...)
PRINTFLIKE(2, 3);

#endif

Now, if you make a mistake like:

errmsg(EXIT_FAILURE, errno, "Failed to open file '%d' for reading", filename);

(where the %d should be %s), then the compiler will complain:

$ gcc -O3 -g -std=c11 -Wall -Wextra -Werror -Wmissing-prototypes -Wstrict-prototypes \

> -Wold-style-definition -c erruse.c
erruse.c: In function ‘main’:
erruse.c:20:64: error: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘const char
*’ [-Werror=format=]
errmsg(EXIT_FAILURE, errno, "Failed to open file '%d' for reading", filename);
~^
%s
cc1: all warnings being treated as errors
$

Section 32.4: Using a format string

Using a format string provides information about the expected number and type of the subsequent variadic
arguments in such a way as to avoid the need for an explicit count argument or a terminator value.

The example below shows a a function that wraps the standard printf() function, only allowing for the use of
variadic arguments of the type char, int and double (in decimal floating point format). Here, like with printf(), the
first argument to the wrapping function is the format string. As the format string is parsed the function is able to
determine if there is another variadic argument expected and what it's type should be.

#include <stdio.h>
#include <stdarg.h>

int simple_printf(const char *format, ...)

{
va_list ap; /* hold information about the variadic argument list. */
int printed = 0; /* count of printed characters */

va_start(ap, format); /* start variadic argument processing */

while (*format != '\0') /* read format string until string terminator */

{
int f = 0;

if (*format == '%')
{
++format;
switch(*format)
{
case 'c' :

GoalKicker.com – C Notes for Professionals 205

 f = printf("%d", va_arg(ap, int)); /* print next variadic argument, note type
promotion from char to int */
break;
case 'd' :
f = printf("%d", va_arg(ap, int)); /* print next variadic argument */
break;

case 'f' :
f = printf("%f", va_arg(ap, double)); /* print next variadic argument */
break;
default :
f = -1; /* invalid format specifier */
break;
}
}
else
{
f = printf("%c", *format); /* print any other characters */
}

if (f < 0) /* check for errors */

{
printed = f;
break;
}
else
{
printed += f;
}
++format; /* move on to next character in string */
}

va_end(ap); /* end variadic argument processing */

return printed;
}

int main (int argc, char *argv[])

{
int x = 40;
int y = 0;

y = simple_printf("There are %d characters in this sentence", x);

simple_printf("\n%d were printed\n", y);
}

GoalKicker.com – C Notes for Professionals 206

Chapter 33: Assertion
Parameter Details
expression expression of scalar type.
message string literal to be included in the diagnostic message.

An assertion is a predicate that the presented condition must be true at the moment the assertion is encountered
by the software. Most common are simple assertions, which are validated at execution time. However, static
assertions are checked at compile time.

Section 33.1: Simple Assertion

An assertion is a statement used to assert that a fact must be true when that line of code is reached. Assertions are
useful for ensuring that expected conditions are met. When the condition passed to an assertion is true, there is no
action. The behavior on false conditions depends on compiler flags. When assertions are enabled, a false input
causes an immediate program halt. When they are disabled, no action is taken. It is common practice to enable
assertions in internal and debug builds, and disable them in release builds, though assertions are often enabled in
release. (Whether termination is better or worse than errors depends on the program.) Assertions should be used
only to catch internal programming errors, which usually means being passed bad parameters.

#include <stdio.h>
/* Uncomment to disable `assert()` */
/* #define NDEBUG */
#include <assert.h>

int main(void)
{
int x = -1;
assert(x >= 0);

printf("x = %d\n", x);

return 0;
}

Possible output with NDEBUG undefined:

a.out: main.c:9: main: Assertion `x >= 0' failed.

Possible output with NDEBUG defined:

x = -1

It's good practice to define NDEBUG globally, so that you can easily compile your code with all assertions either on or
off. An easy way to do this is define NDEBUG as an option to the compiler, or define it in a shared configuration
header (e.g. config.h).

Section 33.2: Static Assertion

Version ≥ C11

Static assertions are used to check if a condition is true when the code is compiled. If it isn't, the compiler is
required to issue an error message and stop the compiling process.

GoalKicker.com – C Notes for Professionals 207

A static assertion is one that is checked at compile time, not run time. The condition must be a constant expression,
and if false will result in a compiler error. The first argument, the condition that is checked, must be a constant
expression, and the second a string literal.

Unlike assert, _Static_assert is a keyword. A convenience macro static_assert is defined in <assert.h>.

#include <assert.h>

enum {N = 5};
_Static_assert(N == 5, "N does not equal 5");
static_assert(N > 10, "N is not greater than 10"); /* compiler error */
Version = C99

Prior to C11, there was no direct support for static assertions. However, in C99, static assertions could be emulated
with macros that would trigger a compilation failure if the compile time condition was false. Unlike _Static_assert,
the second parameter needs to be a proper token name so that a variable name can be created with it. If the
assertion fails, the variable name is seen in the compiler error, since that variable was used in a syntactically
incorrect array declaration.

#define STATIC_MSG(msg, l) STATIC_MSG2(msg, l)

#define STATIC_MSG2(msg,l) on_line_##l##__##msg
#define STATIC_ASSERT(x, msg) extern char STATIC_MSG(msg, __LINE__) [(x)?1:-1]

enum { N = 5 };
STATIC_ASSERT(N == 5, N_must_equal_5);
STATIC_ASSERT(N > 5, N_must_be_greater_than_5); /* compile error */

Before C99, you could not declare variables at arbitrary locations in a block, so you would have to be extremely
cautious about using this macro, ensuring that it only appears where a variable declaration would be valid.

Section 33.3: Assert Error Messages

A trick exists that can display an error message along with an assertion. Normally, you would write code like this

void f(void *p)

{
assert(p != NULL);
/* more code */
}

If the assertion failed, an error message would resemble

Assertion failed: p != NULL, file main.c, line 5

However, you can use logical AND (&&) to give an error message as well

void f(void *p)

{
assert(p != NULL && "function f: p cannot be NULL");
/* more code */
}

Now, if the assertion fails, an error message will read something like this

GoalKicker.com – C Notes for Professionals 208

 Assertion failed: p != NULL && "function f: p cannot be NULL", file main.c, line 5

The reason as to why this works is that a string literal always evaluates to non-zero (true). Adding && 1 to a Boolean
expression has no effect. Thus, adding && "error message" has no effect either, except that the compiler will
display the entire expression that failed.

Section 33.4: Assertion of Unreachable Code

During development, when certain code paths must be prevented from the reach of control flow, you may use
assert(0) to indicate that such a condition is erroneous:

switch (color) {
case COLOR_RED:
case COLOR_GREEN:
case COLOR_BLUE:
break;

default:
assert(0);
}

Whenever the argument of the assert() macro evaluates false, the macro will write diagnostic information to the
standard error stream and then abort the program. This information includes the file and line number of the
assert() statement and can be very helpful in debugging. Asserts can be disabled by defining the macro NDEBUG.

Another way to terminate a program when an error occurs are with the standard library functions exit, quick_exit
or abort. exit and quick_exit take an argument that can be passed back to your environment. abort() (and thus
assert) can be a really severe termination of your program, and certain cleanups that would otherwise be
performed at the end of the execution, may not be performed.

The primary advantage of assert() is that it automatically prints debugging information. Calling abort() has the
advantage that it cannot be disabled like an assert, but it may not cause any debugging information to be displayed.
In some situations, using both constructs together may be beneficial:

if (color == COLOR_RED || color == COLOR_GREEN) {

...
} else if (color == COLOR_BLUE) {
...
} else {
assert(0), abort();
}

When asserts are enabled, the assert() call will print debug information and terminate the program. Execution
never reaches the abort() call. When asserts are disabled, the assert() call does nothing and abort() is called.
This ensures that the program always terminates for this error condition; enabling and disabling asserts only effects
whether or not debug output is printed.

You should never leave such an assert in production code, because the debug information is not helpful for end
users and because abort is generally a much too severe termination that inhibit cleanup handlers that are installed
for exit or quick_exit to run.

Section 33.5: Precondition and Postcondition

One use case for assertion is precondition and postcondition. This can be very useful to maintain invariant and

GoalKicker.com – C Notes for Professionals 209

design by contract. For a example a length is always zero or positive so this function must return a zero or positive
value.

#include <stdio.h>
/* Uncomment to disable `assert()` */
/* #define NDEBUG */
#include <assert.h>

int length2 (int *a, int count)

{
int i, result = 0;

/* Precondition: */
/* NULL is an invalid vector */
assert (a != NULL);
/* Number of dimensions can not be negative.*/
assert (count >= 0);

/* Calculation */
for (i = 0; i < count; ++i)
{
result = result + (a[i] * a[i]);
}

/* Postcondition: */
/* Resulting length can not be negative. */
assert (result >= 0);
return result;
}

#define COUNT 3

int main (void)

{
int a[COUNT] = {1, 2, 3};
int *b = NULL;
int r;
r = length2 (a, COUNT);
printf ("r = %i\n", r);
r = length2 (b, COUNT);
printf ("r = %i\n", r);
return 0;
}

GoalKicker.com – C Notes for Professionals 210

Chapter 34: Generic selection
Parameter Details
generic-assoc-list generic-association OR generic-assoc-list , generic-association
generic-association type-name : assignment-expression OR default : assignment-expression

Section 34.1: Check whether a variable is of a certain qualified

type
#include <stdio.h>

#define is_const_int(x) _Generic((&x), \

const int *: "a const int", \
int *: "a non-const int", \
default: "of other type")

int main(void)
{
const int i = 1;
int j = 1;
double k = 1.0;
printf("i is %s\n", is_const_int(i));
printf("j is %s\n", is_const_int(j));
printf("k is %s\n", is_const_int(k));
}

Output:

i is a const int
j is a non-const int
k is of other type

However, if the type generic macro is implemented like this:

#define is_const_int(x) _Generic((x), \

const int: "a const int", \
int: "a non-const int", \
default: "of other type")

The output is:

i is a non-const int
j is a non-const int
k is of other type

This is because all type qualifiers are dropped for the evaluation of the controlling expression of a _Generic
primary expression.

Section 34.2: Generic selection based on multiple arguments

If a selection on multiple arguments for a type generic expression is wanted, and all types in question are
arithmetic types, an easy way to avoid nested _Generic expressions is to use addition of the parameters in the
controlling expression:

GoalKicker.com – C Notes for Professionals 211

int max_int(int, int);
unsigned max_unsigned(unsigned, unsigned);
double max_double(double, double);

#define MAX(X, Y) _Generic((X)+(Y), \

int: max_int, \
unsigned: max_unsigned, \
default: max_double) \
((X), (Y))

Here, the controlling expression (X)+(Y) is only inspected according to its type and not evaluated. The usual
conversions for arithmetic operands are performed to determine the selected type.

For more complex situation, a selection can be made based on more than one argument to the operator, by nesting
them together.

This example selects between four externally implemented functions, that take combinations of two int and/or
string arguments, and return their sum.

int AddIntInt(int a, int b);

int AddIntStr(int a, const char* b);
int AddStrInt(const char* a, int b );
int AddStrStr(const char* a, const char* b);

#define AddStr(y) \
_Generic((y), int: AddStrInt, \
char*: AddStrStr, \
const char*: AddStrStr )

#define AddInt(y) \
_Generic((y), int: AddIntInt, \
char*: AddIntStr, \
const char*: AddIntStr )

#define Add(x, y) \
_Generic((x) , int: AddInt(y) , \
char*: AddStr(y) , \
const char*: AddStr(y)) \
((x), (y))

int main( void )

{
int result = 0;
result = Add( 100 , 999 );
result = Add( 100 , "999" );
result = Add( "100" , 999 );
result = Add( "100" , "999" );

const int a = -123;

char b[] = "4321";
result = Add( a , b );

int c = 1;
const char d[] = "0";
result = Add( d , ++c );
}

Even though it appears as if argument y is evaluated more than once, it isn't 1. Both arguments are evaluated only
once, at the end of macro Add: ( x , y ), just like in an ordinary function call.

GoalKicker.com – C Notes for Professionals 212

1 (Quoted from: ISO:IEC 9899:201X 6.5.1.1 Generic selection 3)
The controlling expression of a generic selection is not evaluated.

Section 34.3: Type-generic printing macro

#include <stdio.h>

void print_int(int x) { printf("int: %d\n", x); }

void print_dbl(double x) { printf("double: %g\n", x); }
void print_default() { puts("unknown argument"); }

#define print(X) _Generic((X), \

int: print_int, \
double: print_dbl, \
default: print_default)(X)

int main(void) {
print(42);
print(3.14);
print("hello, world");
}

Output:

int: 42
double: 3.14
unknown argument

Note that if the type is neither int nor double, a warning would be generated. To eliminate the warning, you can
add that type to the print(X) macro.

GoalKicker.com – C Notes for Professionals 213

Chapter 35: X-macros
X-macros are a preprocessor-based technique for minimizing repetitious code and maintaining data / code
correspondences. Multiple distinct macro expansions based on a common set of data are supported by
representing the whole group of expansions via a single master macro, with that macro's replacement text
consisting of a sequence of expansions of an inner macro, one for each datum. The inner macro is traditionally
named X(), hence the name of the technique.

Section 35.1: Trivial use of X-macros for printfs

/* define a list of preprocessor tokens on which to call X */
#define X_123 X(1) X(2) X(3)

/* define X to use */
#define X(val) printf("X(%d) made this print\n", val);
X_123
#undef X
/* good practice to undef X to facilitate reuse later on */

This example will result in the preprocessor generating the following code:

printf("X(%d) made this print\n", 1);

printf("X(%d) made this print\n", 2);
printf("X(%d) made this print\n", 3);

Section 35.2: Extension: Give the X macro as an argument

The X-macro approach can be generalized a bit by making the name of the "X" macro an argument of the master
macro. This has the advantages of helping to avoid macro name collisions and of allowing use of a general-purpose
macro as the "X" macro.

As always with X macros, the master macro represents a list of items whose significance is specific to that macro. In
this variation, such a macro might be defined like so:

/* declare list of items */

#define ITEM_LIST(X) \
X(item1) \
X(item2) \
X(item3) \
/* end of list */

One might then generate code to print the item names like so:

/* define macro to apply */

#define PRINTSTRING(value) printf( #value "\n");

/* apply macro to the list of items */

ITEM_LIST(PRINTSTRING)

That expands to this code:

printf( "item1" "\n"); printf( "item2" "\n"); printf( "item3" "\n");

In contrast to standard X macros, where the "X" name is a built-in characteristic of the master macro, with this style

GoalKicker.com – C Notes for Professionals 214

it may be unnecessary or even undesirable to afterward undefine the macro used as the argument (PRINTSTRING in
this example).

Section 35.3: Enum Value and Identifier

/* declare items of the enum */
#define FOREACH \
X(item1) \
X(item2) \
X(item3) \
/* end of list */

/* define the enum values */

#define X(id) MyEnum_ ## id,
enum MyEnum { FOREACH };
#undef X

/* convert an enum value to its identifier */

const char * enum2string(int enumValue)
{
const char* stringValue = NULL;
#define X(id) if (enumValue == MyEnum_ ## id) stringValue = #id;
FOREACH
#undef X
return stringValue;
}

Next you can use the enumerated value in your code and easily print its identifier using :

printf("%s\n", enum2string(MyEnum_item2));

Section 35.4: Code generation

X-Macros can be used for code generation, by writing repetitive code: iterate over a list to do some tasks, or to
declare a set of constants, objects or functions.

Here we use X-macros to declare an enum containing 4 commands and a map of their names as strings

Then we can print the string values of the enum.

/* All our commands */

#define COMMANDS(OP) OP(Open) OP(Close) OP(Save) OP(Quit)

/* generate the enum Commands: {cmdOpen, cmdClose, cmdSave, cmdQuit, }; */

#define ENUM_NAME(name) cmd##name,
enum Commands {
COMMANDS(ENUM_NAME)
};
#undef ENUM_NAME

/* generate the string table */

#define COMMAND_OP(name) #name,
const char* const commandNames[] = {
COMMANDS(COMMAND_OP)
};
#undef COMMAND_OP

/* the following prints "Quit\n": */

GoalKicker.com – C Notes for Professionals 215

printf("%s\n", commandNames[cmdQuit]());

Similarly, we can generate a jump table to call functions by the enum value.

This requires all functions to have the same signature. If they take no arguments and return an int, we would put
this in a header with the enum definition:

/* declare all functions as extern */

#define EXTERN_FUNC(name) extern int doCmd##name(void);
COMMANDS(EXTERN_FUNC)
#undef EXTERN_FUNC

/* declare the function pointer type and the jump table */

typedef int (*CommandFunc)(void);
extern CommandFunc commandJumpTable[];

All of the following can be in different compilation units assuming the part above is included as a header:

/* generate the jump table */

#define FUNC_NAME(name) doCmd##name,
CommandFunc commandJumpTable[] = {
COMMANDS(FUNC_NAME)
};
#undef FUNC_NAME

/* call the save command like this: */

int result = commandJumpTable[cmdSave]();

/* somewhere else, we need the implementations of the commands */

int doCmdOpen(void) {/* code performing open command */}
int doCmdClose(void) {/* code performing close command */}
int doCmdSave(void) {/* code performing save command */}
int doCmdQuit(void) {/* code performing quit command */}

An example of this technique being used in real code is for GPU command dispatching in Chromium.

GoalKicker.com – C Notes for Professionals 216

Chapter 36: Aliasing and eective type
Section 36.1: Eective type
The effective type of a data object is the last type information that was associated with it, if any.

// a normal variable, effective type uint32_t, and this type never changes
uint32_t a = 0.0;

// effective type of *pa is uint32_t, too, simply

// because *pa is the object a
uint32_t* pa = &a;

// the object pointed to by q has no effective type, yet

void* q = malloc(sizeof uint32_t);
// the object pointed to by q still has no effective type,
// because nobody has written to it
uint32_t* qb = q;
// *qb now has effective type uint32_t because a uint32_t value was written
*qb = 37;

// the object pointed to by r has no effective type, yet, although

// it is initialized
void* r = calloc(1, sizeof uint32_t);
// the object pointed to by r still has no effective type,
// because nobody has written to or read from it
uint32_t* rc = r;
// *rc now has effective type uint32_t because a value is read
// from it with that type. The read operation is valid because we used calloc.
// Now the object pointed to by r (which is the same as *rc) has
// gained an effective type, although we didn't change its value.
uint32_t c = *rc;

// the object pointed to by s has no effective type, yet.

void* s = malloc(sizeof uint32_t);
// the object pointed to by s now has effective type uint32_t
// because an uint32_t value is copied into it.
memcpy(s, r, sizeof uint32_t);

Observe that for the latter, it was not necessary that we even have an uint32_t* pointer to that object. The fact
that we have copied another uint32_t object is sufficient.

Section 36.2: restrict qualification

If we have two pointer arguments of the same type, the compiler can't make any assumption and will always have
to assume that the change to *e may change *f:

void fun(float* e, float* f) {

float a = *f
*e = 22;
float b = *f;
print("is %g equal to %g?\n", a, b);
}

float fval = 4;
float eval = 77;
fun(&eval, &fval);

GoalKicker.com – C Notes for Professionals 217

all goes well and something like

is 4 equal to 4?

is printed. If we pass the same pointer, the program will still do the right thing and print

is 4 equal to 22?

This can turn out to be inefficient, if we know by some outside information that e and f will never point to the same
data object. We can reflect that knowledge by adding restrict qualifiers to the pointer parameters:

void fan(float*restrict e, float*restrict f) {

float a = *f
*e = 22;
float b = *f;
print("is %g equal to %g?\n", a, b);
}

Then the compiler may always suppose that e and f point to different objects.

Section 36.3: Changing bytes

Once an object has an effective type, you should not attempt to modify it through a pointer of another type, unless
that other type is a character type, char, signed char or unsigned char.

#include <inttypes.h>
#include <stdio.h>

int main(void) {
uint32_t a = 57;
// conversion from incompatible types needs a cast !
unsigned char* ap = (unsigned char*)&a;
for (size_t i = 0; i < sizeof a; ++i) {
/* set each byte of a to 42 */
ap[i] = 42;
}
printf("a now has value %" PRIu32 "\n", a);
}

This is a valid program that prints

a now has value 707406378

This works because:

The access is made to the individual bytes seen with type unsigned char so each modification is well
defined.
The two views to the object, through a and through *ap, alias, but since ap is a pointer to a character type, the
strict aliasing rule does not apply. Thus the compiler has to assume that the value of a may have been
changed in the for loop. The modified value of a must be constructed from the bytes that have been
changed.
The type of a, uint32_t has no padding bits. All its bits of the representation count for the value, here

GoalKicker.com – C Notes for Professionals 218

 707406378, and there can be no trap representation.

Section 36.4: Character types cannot be accessed through

non-character types
If an object is defined with static, thread, or automatic storage duration, and it has a character type, either: char,
unsigned char, or signed char, it may not be accessed by a non-character type. In the below example a char array
is reinterpreted as the type int, and the behavior is undefined on every dereference of the int pointer b.

int main( void )

{
char a[100];
int* b = ( int* )&a;
*b = 1;

static char c[100];

b = ( int* )&c;
*b = 2;

_Thread_local char d[100];

b = ( int* )&d;
*b = 3;
}

This is undefined because it violates the "effective type" rule, no data object that has an effective type may be
accessed through another type that is not a character type. Since the other type here is int, this is not allowed.

Even if alignment and pointer sizes would be known to fit, this would not exempt from this rule, behavior would still
be undefined.

This means in particular that there is no way in standard C to reserve a buffer object of character type that can be
used through pointers with different types, as you would use a buffer that was received by malloc or similar
function.

A correct way to achieve the same goal as in the above example would be to use a union.

typedef union bufType bufType;

union bufType {
char c[sizeof(int[25])];
int i[25];
};

int main( void )

{
bufType a = { .c = { 0 } }; // reserve a buffer and initialize
int* b = a.i; // no cast necessary
*b = 1;

static bufType a = { .c = { 0 } };
int* b = a.i;
*b = 2;

_Thread_local bufType a = { .c = { 0 } };
int* b = a.i;
*b = 3;
}

Here, the union ensures that the compiler knows from the start that the buffer could be accessed through different

GoalKicker.com – C Notes for Professionals 219

views. This also has the advantage that now the buffer has a "view" a.i that already is of type int and no pointer
conversion is needed.

Section 36.5: Violating the strict aliasing rules

In the following code let us assume for simplicity that float and uint32_t have the same size.

void fun(uint32_t* u, float* f) {

float a = *f
*u = 22;
float b = *f;
print("%g should equal %g\n", a, b);
}

u and f have different base type, and thus the compiler can assume that they point to different objects. There is no
possibility that *f could have changed between the two initializations of a and b, and so the compiler may optimize
the code to something equivalent to

void fun(uint32_t* u, float* f) {

float a = *f
*u = 22;
print("%g should equal %g\n", a, a);
}

That is, the second load operation of *f can be optimized out completely.

If we call this function "normally"

float fval = 4;
uint32_t uval = 77;
fun(&uval, &fval);

all goes well and something like

4 should equal 4

is printed. But if we cheat and pass the same pointer, after converting it,

float fval = 4;
uint32_t* up = (uint32_t*)&fval;
fun(up, &fval);

we violate the strict aliasing rule. Then the behavior becomes undefined. The output could be as above, if the
compiler had optimized the second access, or something completely different, and so your program ends up in a
completely unreliable state.

GoalKicker.com – C Notes for Professionals 220

Chapter 37: Compilation
The C language is traditionally a compiled language (as opposed to interpreted). The C Standard defines translation
phases, and the product of applying them is a program image (or compiled program). In c11, the phases are listed
in §5.1.1.2.

Section 37.1: The Compiler

After the C pre-processor has included all the header files and expanded all macros, the compiler can compile the
program. It does this by turning the C source code into an object code file, which is a file ending in .o which
contains the binary version of the source code. Object code is not directly executable, though. In order to make an
executable, you also have to add code for all of the library functions that were #included into the file (this is not the
same as including the declarations, which is what #include does). This is the job of the linker.

In general, the exact sequence how to invoke a C compiler depends much on the system that you are using. Here
we are using the GCC compiler, though it should be noted that many more compilers exist:

% gcc -Wall -c foo.c

% is the OS' command prompt. This tells the compiler to run the pre-processor on the file foo.c and then compile it
into the object code file foo.o. The -c option means to compile the source code file into an object file but not to
invoke the linker. This option -c is available on POSIX systems, such as Linux or macOS; other systems may use
different syntax.

If your entire program is in one source code file, you can instead do this:

% gcc -Wall foo.c -o foo

This tells the compiler to run the pre-processor on foo.c, compile it and then link it to create an executable called
foo. The -o option states that the next word on the line is the name of the binary executable file (program). If you
don't specify the -o, (if you just type gcc foo.c), the executable will be named a.out for historical reasons.

In general the compiler takes four steps when converting a .c file into an executable:

1. pre-processing - textually expands #include directives and #define macros in your .c file
2. compilation - converts the program into assembly (you can stop the compiler at this step by adding the -S
option)
3. assembly - converts the assembly into machine code
4. linkage - links the object code to external libraries to create an executable

Note also that the name of the compiler we are using is GCC, which stands for both "GNU C compiler" and "GNU
compiler collection", depending on context. Other C compilers exist. For Unix-like operating systems, many of them
have the name cc, for "C compiler", which is often a symbolic link to some other compiler. On Linux systems, cc is
often an alias for GCC. On macOS or OS-X, it points to clang.

The POSIX standards currently mandates c99 as the name of a C compiler — it supports the C99 standard by
default. Earlier versions of POSIX mandated c89 as the compiler. POSIX also mandates that this compiler
understands the options -c and -o that we used above.

Note: The -Wall option present in both gcc examples tells the compiler to print warnings about questionable
constructions, which is strongly recommended. It is a also good idea to add other warning options, e.g. -Wextra.

GoalKicker.com – C Notes for Professionals 221

Section 37.2: File Types
Compiling C programs requires you to work with five kinds of files:

1. Source files: These files contain function definitions, and have names which end in .c by convention. Note:
.cc and .cpp are C++ files; not C files.
e.g., foo.c

2. Header files: These files contain function prototypes and various pre-processor statements (see below).
They are used to allow source code files to access externally-defined functions. Header files end in .h by
convention.
e.g., foo.h

3. Object files: These files are produced as the output of the compiler. They consist of function definitions in
binary form, but they are not executable by themselves. Object files end in .o by convention, although on
some operating systems (e.g. Windows, MS-DOS), they often end in .obj.
e.g., foo.o foo.obj

4. Binary executables: These are produced as the output of a program called a "linker". The linker links
together a number of object files to produce a binary file which can be directly executed. Binary executables
have no special suffix on Unix operating systems, although they generally end in .exe on Windows.
e.g., foo foo.exe

5. Libraries: A library is a compiled binary but is not in itself an an executable (i.e., there is no main() function
in a library). A library contains functions that may be used by more than one program. A library should ship
with header files which contain prototypes for all functions in the library; these header files should be
referenced (e.g; #include <library.h>) in any source file that uses the library. The linker then needs to be
referred to the library so the program can successfully compiled. There are two types of libraries: static and
dynamic.

Static library: A static library (.a files for POSIX systems and .lib files for Windows — not to be
confused with DLL import library files, which also use the .lib extension) is statically built into the
program . Static libraries have the advantage that the program knows exactly which version of a library
is used. On the other hand, the sizes of executables are bigger as all used library functions are
included.
e.g., libfoo.a foo.lib
Dynamic library: A dynamic library (.so files for most POSIX systems, .dylib for OSX and .dll files
for Windows) is dynamically linked at runtime by the program. These are also sometimes referred to
as shared libraries because one library image can be shared by many programs. Dynamic libraries
have the advantage of taking up less disk space if more than one application is using the library. Also,
they allow library updates (bug fixes) without having to rebuild executables.
e.g., foo.so foo.dylib foo.dll

Section 37.3: The Linker

The job of the linker is to link together a bunch of object files (.o files) into a binary executable. The process of
linking mainly involves resolving symbolic addresses to numerical addresses. The result of the link process is normally
an executable program.

During the link process, the linker will pick up all the object modules specified on the command line, add some
system-specific startup code in front and try to resolve all external references in the object module with external
definitions in other object files (object files can be specified directly on the command line or may implicitly be added
through libraries). It will then assign load addresses for the object files, that is, it specifies where the code and data

GoalKicker.com – C Notes for Professionals 222

will end up in the address space of the finished program. Once it's got the load addresses, it can replace all the
symbolic addresses in the object code with "real", numerical addresses in the target's address space. The program
is ready to be executed now.

This includes both the object files that the compiler created from your source code files as well as object files that
have been pre-compiled for you and collected into library files. These files have names which end in .a or .so, and
you normally don't need to know about them, as the linker knows where most of them are located and will link
them in automatically as needed.

Implicit invocation of the linker

Like the pre-processor, the linker is a separate program, often called ld (but Linux uses collect2, for example).
Also like the pre-processor, the linker is invoked automatically for you when you use the compiler. Thus, the normal
way of using the linker is as follows:

% gcc foo.o bar.o baz.o -o myprog

This line tells the compiler to link together three object files (foo.o, bar.o, and baz.o) into a binary executable file
named myprog. Now you have a file called myprog that you can run and which will hopefully do something cool
and/or useful.

Explicit invocation of the linker

It is possible to invoke the linker directly, but this is seldom advisable, and is typically very platform-specific. That is,
options that work on Linux won't necessarily work on Solaris, AIX, macOS, Windows, and similarly for any other
platform. If you work with GCC, you can use gcc -v to see what is executed on your behalf.

Options for the linker

The linker also takes some arguments to modify it's behavior. The following command would tell gcc to link foo.o
and bar.o, but also include the ncurses library.

% gcc foo.o bar.o -o foo -lncurses

This is actually (more or less) equivalent to

% gcc foo.o bar.o /usr/lib/libncurses.so -o foo

(although libncurses.so could be libncurses.a, which is just an archive created with ar). Note that you should list
the libraries (either by pathname or via -lname options) after the object files. With static libraries, the order that
they are specified matters; often, with shared libraries, the order doesn't matter.

Note that on many systems, if you are using mathematical functions (from <math.h>), you need to specify -lm to
load the mathematics library — but Mac OS X and macOS Sierra do not require this. There are other libraries that
are separate libraries on Linux and other Unix systems, but not on macOS — POSIX threads, and POSIX realtime,
and networking libraries are examples. Consequently, the linking process varies between platforms.

Other compilation options

This is all you need to know to begin compiling your own C programs. Generally, we also recommend that you use
the -Wall command-line option:

% gcc -Wall -c foo.cc

GoalKicker.com – C Notes for Professionals 223

The -Wall option causes the compiler to warn you about legal but dubious code constructs, and will help you catch
a lot of bugs very early.

If you want the compiler to throw more warnings at you (including variables that are declared but not used,
forgetting to return a value etc.), you can use this set of options, as -Wall, despite the name, doesn't turn all of the
possible warnings on:

% gcc -Wall -Wextra -Wfloat-equal -Wundef -Wcast-align -Wwrite-strings -Wlogical-op \

> -Wmissing-declarations -Wredundant-decls -Wshadow …

Note that clang has an option -Weverything which really does turn on all warnings in clang.

Section 37.4: The Preprocessor

Before the C compiler starts compiling a source code file, the file is processed in a preprocessing phase. This phase
can be done by a separate program or be completely integrated in one executable. In any case, it is invoked
automatically by the compiler before compilation proper begins. The preprocessing phase converts your source
code into another source code or translation unit by applying textual replacements. You can think of it as a
"modified" or "expanded" source code. That expanded source may exist as a real file in the file system, or it may
only be stored in memory for a short time before being processed further.

Preprocessor commands start with the pound sign ("#"). There are several preprocessor commands; two of the
most important are:

1. Defines:

#define is mainly used to define constants. For instance,

#define BIGNUM 1000000

int a = BIGNUM;

becomes

int a = 1000000;

#define is used in this way so as to avoid having to explicitly write out some constant value in many different
places in a source code file. This is important in case you need to change the constant value later on; it's
much less bug-prone to change it once, in the #define, than to have to change it in multiple places scattered
all over the code.

Because #define just does advanced search and replace, you can also declare macros. For instance:

#define ISTRUE(stm) do{stm = stm ? 1 : 0;}while(0)

// in the function:
a = x;
ISTRUE(a);

becomes:

// in the function:
a = x;
do {
a = a ? 1 : 0;
} while(0);

GoalKicker.com – C Notes for Professionals 224

 At first approximation, this effect is roughly the same as with inline functions, but the preprocessor doesn't
provide type checking for #define macros. This is well known to be error-prone and their use necessitates
great caution.

Also note here, that the preprocessor would also replace comments with a blanks as explained below.

2. Includes:

#include is used to access function definitions defined outside of a source code file. For instance:

#include <stdio.h>

causes the preprocessor to paste the contents of <stdio.h> into the source code file at the location of the
#include statement before it gets compiled. #include is almost always used to include header files, which
are files which mainly contain function declarations and #define statements. In this case, we use #include in
order to be able to use functions such as printf and scanf, whose declarations are located in the file
stdio.h. C compilers do not allow you to use a function unless it has previously been declared or defined in
that file; #include statements are thus the way to re-use previously-written code in your C programs.

3. Logic operations:

#if defined A || defined B

variable = another_variable + 1;
#else
variable = another_variable * 2;
#endif

will be changed to:

variable = another_variable + 1;

if A or B were defined somewhere in the project before. If this is not the case, of course the preprocessor will
do this:

variable = another_variable * 2;

This is often used for code, that runs on different systems or compiles on different compilers. Since there are
global defines, that are compiler/system specific you can test on those defines and always let the compiler
just use the code he will compile for sure.

4. Comments

The Preprocessor replaces all comments in the source file by single spaces. Comments are indicated by // up
to the end of the line, or a combination of opening /* and closing */ comment brackets.

Section 37.5: The Translation Phases

As of the C 2011 Standard, listed in §5.1.1.2 Translation Phases, the translation of source code to program image
(e.g., the executable) are listed to occur in 8 ordered steps.

GoalKicker.com – C Notes for Professionals 225

 1. The source file input is mapped to the source character set (if necessary). Trigraphs are replaced in this step.
2. Continuation lines (lines that end with \) are spliced with the next line.
3. The source code is parsed into whitespace and preprocessing tokens.
4. The preprocessor is applied, which executes directives, expands macros, and applies pragmas. Each source
file pulled in by #include undergoes translation phases 1 through 4 (recursively if necessary). All
preprocessor related directives are then deleted.
5. Source character set values in character constants and string literals are mapped to the execution character
set.
6. String literals adjacent to each other are concatenated.
7. The source code is parsed into tokens, which comprise the translation unit.
8. External references are resolved, and the program image is formed.

An implementation of a C compiler may combine several steps together, but the resulting image must still behave
as if the above steps had occurred separately in the order listed above.

GoalKicker.com – C Notes for Professionals 226

Chapter 38: Inline assembly
Section 38.1: gcc Inline assembly in macros
We can put assembly instructions inside a macro and use the macro like you would call a function.

#define mov(x,y) \
{ \
__asm__ ("l.cmov %0,%1,%2" : "=r" (x) : "r" (y), "r" (0x0000000F)); \
}

/// some definition and assignment

unsigned char sbox[size][size];
unsigned char sbox[size][size];

///Using
mov(state[0][1], sbox[si][sj]);

Using inline assembly instructions embedded in C code can improve the run time of a program. This is very helpful
in time critical situations like cryptographic algorithms such as AES. For example, for a simple shift operation that is
needed in the AES algorithm, we can substitute a direct Rotate Right assembly instruction with C shift operator >>.

In an implementation of 'AES256', in 'AddRoundKey()' function we have some statements like this:

unsigned int w; // 32-bit

unsigned char subkey[4]; // 8-bit, 4*8 = 32

subkey[0] = w >> 24; // hold 8 bit, MSB, leftmost group of 8-bits

subkey[1] = w >> 16; // hold 8 bit, second group of 8-bit from left
subkey[2] = w >> 8; // hold 8 bit, second group of 8-bit from right
subkey[3] = w; // hold 8 bit, LSB, rightmost group of 8-bits

/// subkey <- w

They simply assign the bit value of w to subkey array.

We can change three shift + assign and one assign C expression with only one assembly Rotate Right operation.

__asm__ ("l.ror %0,%1,%2" : "=r" (* (unsigned int *) subkey) : "r" (w), "r" (0x10));

The final result is exactly same.

Section 38.2: gcc Basic asm support

Basic assembly support with gcc has the following syntax:

asm [ volatile ] ( AssemblerInstructions )

where AssemblerInstructions is the direct assembly code for the given processor. The volatile keyword is optional
and has no effect as gcc does not optimize code within a basic asm statement. AssemblerInstructions can contain
multiple assembly instructions. A basic asm statement is used if you have an asm routine that must exist outside of
a C function. The following example is from the GCC manual:

/* Note that this code will not compile with -masm=intel */

#define DebugBreak() asm("int $3")

GoalKicker.com – C Notes for Professionals 227

In this example, you could then use DebugBreak() in other places in your code and it will execute the assembly
instruction int $3. Note that even though gcc will not modify any code in a basic asm statement, the optimizer may
still move consecutive asm statements around. If you have multiple assembly instructions that must occur in a
specific order, include them in one asm statement.

Section 38.3: gcc Extended asm support

Extended asm support in gcc has the following syntax:

asm [volatile] ( AssemblerTemplate

: OutputOperands
[ : InputOperands
[ : Clobbers ] ])

asm [volatile] goto ( AssemblerTemplate

:
: InputOperands
: Clobbers
: GotoLabels)

where AssemblerTemplate is the template for the assembler instruction, OutputOperands are any C variables that
can be modified by the assembly code, InputOperands are any C variables used as input parameters, Clobbers are
a list or registers that are modified by the assembly code, and GotoLabels are any goto statement labels that may
be used in the assembly code.

The extended format is used within C functions and is the more typical usage of inline assembly. Below is an
example from the Linux kernel for byte swapping 16-bit and 32-bit numbers for an ARM processor:

/* From arch/arm/include/asm/swab.h in Linux kernel version 4.6.4 */

#if __LINUX_ARM_ARCH__ >= 6

static inline __attribute_const__ __u32 __arch_swahb32(__u32 x)

{
__asm__ ("rev16 %0, %1" : "=r" (x) : "r" (x));
return x;
}
#define __arch_swahb32 __arch_swahb32
#define __arch_swab16(x) ((__u16)__arch_swahb32(x))

static inline __attribute_const__ __u32 __arch_swab32(__u32 x)

{
__asm__ ("rev %0, %1" : "=r" (x) : "r" (x));
return x;
}
#define __arch_swab32 __arch_swab32

#endif

Each asm section uses the variable x as its input and output parameter. The C function then returns the
manipulated result.

With the extended asm format, gcc may optimize the assembly instructions in an asm block following the same
rules it uses for optimizing C code. If you want your asm section to remain untouched, use the volatile keyword
for the asm section.

GoalKicker.com – C Notes for Professionals 228

Chapter 39: Identifier Scope
Section 39.1: Function Prototype Scope
#include <stdio.h>

/* The parameter name, apple, has function prototype scope. These names
are not significant outside the prototype itself. This is demonstrated
below. */

int test_function(int apple);

int main(void)
{
int orange = 5;

orange = test_function(orange);
printf("%d\r\n", orange); //orange = 6

return 0;
}

int test_function(int fruit)

{
fruit += 1;
return fruit;
}

Note that you get puzzling error messages if you introduce a type name in a prototype:

int function(struct whatever *arg);

struct whatever
{
int a;
// ...
};

int function(struct whatever *arg)

{
return arg->a;
}

With GCC 6.3.0, this code (source file dc11.c) produces:

$ gcc -O3 -g -std=c11 -Wall -Wextra -Werror -c dc11.c

dc11.c:1:25: error: ‘struct whatever’ declared inside parameter list will not be visible outside of
this definition or declaration [-Werror]
int function(struct whatever *arg);
^~~~~~~~
dc11.c:9:9: error: conflicting types for ‘function’
int function(struct whatever *arg)
^~~~~~~~
dc11.c:1:9: note: previous declaration of ‘function’ was here
int function(struct whatever *arg);
^~~~~~~~
cc1: all warnings being treated as errors
$

GoalKicker.com – C Notes for Professionals 229

Place the structure definition before the function declaration, or add struct whatever; as a line before the
function declaration, and there is no problem. You should not introduce new type names in a function prototype
because there's no way to use that type, and hence no way to define or use that function.

Section 39.2: Block Scope

An identifier has block scope if its corresponding declaration appears inside a block (parameter declaration in
function definition apply). The scope ends at the end of the corresponding block.

No different entities with the same identifier can have the same scope, but scopes may overlap. In case of
overlapping scopes the only visible one is the one declared in the innermost scope.

#include <stdio.h>

void test(int bar) // bar has scope test function block

{
int foo = 5; // foo has scope test function block
{
int bar = 10; // bar has scope inner block, this overlaps with previous
test:bar declaration, and it hides test:bar
printf("%d %d\n", foo, bar); // 5 10
} // end of scope for inner bar
printf("%d %d\n", foo, bar); // 5 5, here bar is test:bar
} // end of scope for test:foo and test:bar

int main(void)
{
int foo = 3; // foo has scope main function block

printf("%d\n", foo); // 3
test(5);
printf("%d\n", foo); // 3
return 0;
} // end of scope for main:foo

Section 39.3: File Scope

#include <stdio.h>

/* The identifier, foo, is declared outside all blocks.

It can be used anywhere after the declaration until the end of
the translation unit. */
static int foo;

void test_function(void)
{
foo += 2;
}

int main(void)
{
foo = 1;

test_function();
printf("%d\r\n", foo); //foo = 3;

return 0;
}

GoalKicker.com – C Notes for Professionals 230

Section 39.4: Function scope
Function scope is the special scope for labels. This is due to their unusual property. A label is visible through the
entire function it is defined and one can jump (using instruction gotolabel) to it from any point in the same
function. While not useful, the following example illustrate the point:

#include <stdio.h>

int main(int argc,char *argv[]) {

int a = 0;
goto INSIDE;
OUTSIDE:
if (a!=0) {
int i=0;
INSIDE:
printf("a=%d\n",a);
goto OUTSIDE;
}
}

INSIDE may seem defined inside the if block, as it is the case for i which scope is the block, but it is not. It is visible
in the whole function as the instruction goto INSIDE; illustrates. Thus there can't be two labels with the same
identifier in a single function.

A possible usage is the following pattern to realize correct complex cleanups of allocated ressources:

#include <stdlib.h>
#include <stdio.h>

void a_function(void) {
double* a = malloc(sizeof(double[34]));
if (!a) {
fprintf(stderr,"can't allocate\n");
return; /* No point in freeing a if it is null */
}
FILE* b = fopen("some_file","r");
if (!b) {
fprintf(stderr,"can't open\n");
goto CLEANUP1; /* Free a; no point in closing b */
}
/* do something reasonable */
if (error) {
fprintf(stderr,"something's wrong\n");
goto CLEANUP2; /* Free a and close b to prevent leaks */
}
/* do yet something else */
CLEANUP2:
close(b);
CLEANUP1:
free(a);
}

Labels such as CLEANUP1 and CLEANUP2 are special identifiers that behave differently from all other identifiers. They
are visible from everywhere inside the function, even in places that are executed before the labeled statement, or
even in places that could never be reached if none of the goto is executed. Labels are often written in lower-case
rather than upper-case.

GoalKicker.com – C Notes for Professionals 231

Chapter 40: Implicit and Explicit
Conversions
Section 40.1: Integer Conversions in Function Calls
Given that the function has a proper prototype, integers are widened for calls to functions according to the rules of
integer conversion, C11 6.3.1.3.

6.3.1.3 Signed and unsigned integers

When a value with integer type is converted to another integer type other than _Bool, if the value can be
represented by the new type, it is unchanged.

Otherwise, if the new type is unsigned, the value is converted by repeatedly adding or subtracting one
more than the maximum value that can be represented in the new type until the value is in the range of
the new type.

Otherwise, the new type is signed and the value cannot be represented in it; either the result is
implementation-defined or an implementation-defined signal is raised.

Usually you should not truncate a wide signed type to a narrower signed type, because obviously the values can't fit
and there is no clear meaning that this should have. The C standard cited above defines these cases to be
"implementation-defined", that is, they are not portable.

The following example supposes that int is 32 bit wide.

#include <stdio.h>
#include <stdint.h>

void param_u8(uint8_t val) {

printf("%s val is %d\n", __func__, val); /* val is promoted to int */
}

void param_u16(uint16_t val) {

printf("%s val is %d\n", __func__, val); /* val is promoted to int */
}

void param_u32(uint32_t val) {

printf("%s val is %u\n", __func__, val); /* here val fits into unsigned */
}

void param_u64(uint64_t val) {

printf("%s val is " PRI64u "\n", __func__, val); /* Fixed with format string */
}

void param_s8(int8_t val) {

printf("%s val is %d\n", __func__, val); /* val is promoted to int */
}

void param_s16(int16_t val) {

printf("%s val is %d\n", __func__, val); /* val is promoted to int */
}

void param_s32(int32_t val) {

printf("%s val is %d\n", __func__, val); /* val has same width as int */
}

GoalKicker.com – C Notes for Professionals 232

void param_s64(int64_t val) {
printf("%s val is " PRI64d "\n", __func__, val); /* Fixed with format string */
}

int main(void) {

/* Declare integers of various widths */

uint8_t u8 = 127;
uint8_t s64 = INT64_MAX;

/* Integer argument is widened when function parameter is wider */

param_u8(u8); /* param_u8 val is 127 */
param_u16(u8); /* param_u16 val is 127 */
param_u32(u8); /* param_u32 val is 127 */
param_u64(u8); /* param_u64 val is 127 */
param_s8(u8); /* param_s8 val is 127 */
param_s16(u8); /* param_s16 val is 127 */
param_s32(u8); /* param_s32 val is 127 */
param_s64(u8); /* param_s64 val is 127 */

/* Integer argument is truncated when function parameter is narrower */

param_u8(s64); /* param_u8 val is 255 */
param_u16(s64); /* param_u16 val is 65535 */
param_u32(s64); /* param_u32 val is 4294967295 */
param_u64(s64); /* param_u64 val is 9223372036854775807 */
param_s8(s64); /* param_s8 val is implementation defined */
param_s16(s64); /* param_s16 val is implementation defined */
param_s32(s64); /* param_s32 val is implementation defined */
param_s64(s64); /* param_s64 val is 9223372036854775807 */

return 0;
}

Section 40.2: Pointer Conversions in Function Calls

Pointer conversions to void* are implicit, but any other pointer conversion must be explicit. While the compiler
allows an explicit conversion from any pointer-to-data type to any other pointer-to-data type, accessing an object
through a wrongly typed pointer is erroneous and leads to undefined behavior. The only case that these are
allowed are if the types are compatible or if the pointer with which your are looking at the object is a character type.

#include <stdio.h>

void func_voidp(void* voidp) {

printf("%s Address of ptr is %p\n", __func__, voidp);
}

/* Structures have same shape, but not same type */

struct struct_a {
int a;
int b;
} data_a;

struct struct_b {
int a;
int b;
} data_b;

void func_struct_b(struct struct_b* bp) {

printf("%s Address of ptr is %p\n", __func__, (void*) bp);
}

GoalKicker.com – C Notes for Professionals 233

int main(void) {

/* Implicit ptr conversion allowed for void* */

func_voidp(&data_a);

/*
* Explicit ptr conversion for other types
*
* Note that here although the have identical definitions,
* the types are not compatible, and that the this call is
* erroneous and leads to undefined behavior on execution.
*/
func_struct_b((struct struct_b*)&data_a);

/* My output shows: */
/* func_charp Address of ptr is 0x601030 */
/* func_voidp Address of ptr is 0x601030 */
/* func_struct_b Address of ptr is 0x601030 */

return 0;
}

GoalKicker.com – C Notes for Professionals 234

Chapter 41: Type Qualifiers
Section 41.1: Volatile variables
The volatile keyword tells the compiler that the value of the variable may change at any time as a result of
external conditions, not only as a result of program control flow.

The compiler will not optimize anything that has to do with the volatile variable.

volatile int foo; /* Different ways to declare a volatile variable */

int volatile foo;

volatile uint8_t * pReg; /* Pointers to volatile variable */

uint8_t volatile * pReg;

There are two main reasons to uses volatile variables:

To interface with hardware that has memory-mapped I/O registers.

When using variables that are modified outside the program control flow (e.g., in an interrupt service routine)

Let's see this example:

int quit = false;

void main()
{
...
while (!quit) {
// Do something that does not modify the quit variable
}
...
}

void interrupt_handler(void)
{
quit = true;
}

The compiler is allowed to notice the while loop does not modify the quit variable and convert the loop to a
endless while (true) loop. Even if the quit variable is set on the signal handler for SIGINT and SIGTERM, the
compiler does not know that.

Declaring quit as volatile will tell the compiler to not optimize the loop and the problem will be solved.

The same problem happens when accessing hardware, as we see in this example:

uint8_t * pReg = (uint8_t *) 0x1717;

// Wait for register to become non-zero

while (*pReg == 0) { } // Do something else

The behavior of the optimizer is to read the variable's value once, there is no need to reread it, since the value will
always be the same. So we end up with an infinite loop. To force the compiler to do what we want, we modify the
declaration to:

uint8_t volatile * pReg = (uint8_t volatile *) 0x1717;

GoalKicker.com – C Notes for Professionals 235

Section 41.2: Unmodifiable (const) variables
const int a = 0; /* This variable is "unmodifiable", the compiler
should throw an error when this variable is changed */
int b = 0; /* This variable is modifiable */

b += 10; /* Changes the value of 'b' */

a += 10; /* Throws a compiler error */

The const qualification only means that we don't have the right to change the data. It doesn't mean that the value
cannot change behind our back.

_Bool doIt(double const* a) {

double rememberA = *a;
// do something long and complicated that calls other functions

return rememberA == *a;

}

During the execution of the other calls *a might have changed, and so this function may return either false or
true.

Warning

Variables with const qualification could still be changed using pointers:

const int a = 0;

int *a_ptr = (int*)&a; /* This conversion must be explicitly done with a cast */
*a_ptr += 10; /* This has undefined behavior */

printf("a = %d\n", a); /* May print: "a = 10" */

But doing so is an error that leads to undefined behavior. The difficulty here is that this may behave as expected in
simple examples as this, but then go wrong when the code grows.

GoalKicker.com – C Notes for Professionals 236

Chapter 42: Typedef
The typedef mechanism allows the creation of aliases for other types. It does not create new types. People often
use typedef to improve the portability of code, to give aliases to structure or union types, or to create aliases for
function (or function pointer) types.

In the C standard, typedef is classified as a 'storage class' for convenience; it occurs syntactically where storage
classes such as static or extern could appear.

Section 42.1: Typedef for Structures and Unions

You can give alias names to a struct:

typedef struct Person {

char name[32];
int age;
} Person;

Person person;

Compared to the traditional way of declaring structs, programmers wouldn't need to have struct every time they
declare an instance of that struct.

Note that the name Person (as opposed to struct Person) is not defined until the final semicolon. Thus for linked
lists and tree structures which need to contain a pointer to the same structure type, you must use either:

typedef struct Person {

char name[32];
int age;
struct Person *next;
} Person;

or:

typedef struct Person Person;

struct Person {
char name[32];
int age;
Person *next;
};

The use of a typedef for a union type is very similar.

typedef union Float Float;

union Float
{
float f;
char b[sizeof(float)];
};

A structure similar to this can be used to analyze the bytes that make up a float value.

GoalKicker.com – C Notes for Professionals 237

Section 42.2: Typedef for Function Pointers
We can use typedef to simplify the usage of function pointers. Imagine we have some functions, all having the
same signature, that use their argument to print out something in different ways:

#include<stdio.h>

void print_to_n(int n)
{
for (int i = 1; i <= n; ++i)
printf("%d\n", i);
}

void print_n(int n)
{
printf("%d\n, n);
}

Now we can use a typedef to create a named function pointer type called printer:

typedef void (*printer_t)(int);

This creates a type, named printer_t for a pointer to a function that takes a single int argument and returns
nothing, which matches the signature of the functions we have above. To use it we create a variable of the created
type and assign it a pointer to one of the functions in question:

printer_t p = &print_to_n;
void (*p)(int) = &print_to_n; // This would be required without the type

Then to call the function pointed to by the function pointer variable:

p(5); // Prints 1 2 3 4 5 on separate lines

(*p)(5); // So does this

Thus the typedef allows a simpler syntax when dealing with function pointers. This becomes more apparent when
function pointers are used in more complex situations, such as arguments to functions.

If you are using a function that takes a function pointer as a parameter without a function pointer type defined the
function definition would be,

void foo (void (*printer)(int), int y){

//code
printer(y);
//code
}

However, with the typedef it is:

void foo (printer_t printer, int y){

//code
printer(y);
//code
}

Likewise functions can return function pointers and again, the use of a typedef can make the syntax simpler when
doing so.

GoalKicker.com – C Notes for Professionals 238

A classic example is the signal function from <signal.h>. The declaration for it (from the C standard) is:

void (*signal(int sig, void (*func)(int)))(int);

That's a function that takes two arguments — an int and a pointer to a function which takes an int as an argument
and returns nothing — and which returns a pointer to function like its second argument.

If we defined a type SigCatcher as an alias for the pointer to function type:

typedef void (*SigCatcher)(int);

then we could declare signal() using:

SigCatcher signal(int sig, SigCatcher func);

On the whole, this is easier to understand (even though the C standard did not elect to define a type to do the job).
The signal function takes two arguments, an int and a SigCatcher, and it returns a SigCatcher — where a
SigCatcher is a pointer to a function that takes an int argument and returns nothing.

Although using typedef names for pointer to function types makes life easier, it can also lead to confusion for
others who will maintain your code later on, so use with caution and proper documentation. See also Function
Pointers.

Section 42.3: Simple Uses of Typedef

For giving short names to a data type

Instead of:

long long int foo;

struct mystructure object;

one can use

/* write once */
typedef long long ll;
typedef struct mystructure mystruct;

/* use whenever needed */

ll foo;
mystruct object;

This reduces the amount of typing needed if the type is used many times in the program.

Improving portability

The attributes of data types vary across different architectures. For example, an int may be a 2-byte type in one
implementation and an 4-byte type in another. Suppose a program needs to use a 4-byte type to run correctly.

In one implementation, let the size of int be 2 bytes and that of long be 4 bytes. In another, let the size of int be 4
bytes and that of long be 8 bytes. If the program is written using the second implementation,

/* program expecting a 4 byte integer */

int foo; /* need to hold 4 bytes to work */
/* some code involving many more ints */

GoalKicker.com – C Notes for Professionals 239

For the program to run in the first implementation, all the int declarations will have to be changed to long.

/* program now needs long */

long foo; /*need to hold 4 bytes to work */
/* some code involving many more longs - lot to be changed */

To avoid this, one can use typedef

/* program expecting a 4 byte integer */

typedef int myint; /* need to declare once - only one line to modify if needed */
myint foo; /* need to hold 4 bytes to work */
/* some code involving many more myints */

Then, only the typedef statement would need to be changed each time, instead of examining the whole program.

Version ≥ C99

The <stdint.h> header and the related <inttypes.h> header define standard type names (using typedef) for
integers of various sizes, and these names are often the best choice in modern code that needs fixed size integers.
For example, uint8_t is an unsigned 8-bit integer type; int64_t is a signed 64-bit integer type. The type uintptr_t
is an unsigned integer type big enough to hold any pointer to object. These types are theoretically optional — but it
is rare for them not to be available. There are variants like uint_least16_t (the smallest unsigned integer type with
at least 16 bits) and int_fast32_t (the fastest signed integer type with at least 32 bits). Also, intmax_t and
uintmax_t are the largest integer types supported by the implementation. These types are mandatory.

To specify a usage or to improve readability

If a set of data has a particular purpose, one can use typedef to give it a meaningful name. Moreover, if the
property of the data changes such that the base type must change, only the typedef statement would have to be
changed, instead of examining the whole program.

GoalKicker.com – C Notes for Professionals 240

Chapter 43: Storage Classes
A storage class is used to set the scope of a variable or function. By knowing the storage class of a variable, we can
determine the life-time of that variable during the run-time of the program.

Section 43.1: auto

This storage class denotes that an identifier has automatic storage duration. This means once the scope in which
the identifier was defined ends, the object denoted by the identifier is no longer valid.

Since all objects, not living in global scope or being declared static, have automatic storage duration by default
when defined, this keyword is mostly of historical interest and should not be used:

int foo(void)
{
/* An integer with automatic storage duration. */
auto int i = 3;

/* Same */
int j = 5;

return 0;
} /* The values of i and j are no longer able to be used. */

Section 43.2: register

Hints to the compiler that access to an object should be as fast as possible. Whether the compiler actually uses the
hint is implementation-defined; it may simply treat it as equivalent to auto.

The only property that is definitively different for all objects that are declared with register is that they cannot
have their address computed. Thereby register can be a good tool to ensure certain optimizations:

register size_t size = 467;

is an object that can never alias because no code can pass its address to another function where it might be
changed unexpectedly.

This property also implies that an array

register int array[5];

cannot decay into a pointer to its first element (i.e. array turning into &array[0]). This means that the elements of
such an array cannot be accessed and the array itself cannot be passed to a function.

In fact, the only legal usage of an array declared with a register storage class is the sizeof operator; any other
operator would require the address of the first element of the array. For that reason, arrays generally should not be
declared with the register keyword since it makes them useless for anything other than size computation of the
entire array, which can be done just as easily without the register keyword.

The register storage class is more appropriate for variables that are defined inside a block and are accessed with
high frequency. For example,

/* prints the sum of the first 5 integers*/

/* code assumed to be part of a function body*/

GoalKicker.com – C Notes for Professionals 241

{
register int k, sum;
for(k = 1, sum = 0; k < 6; sum += k, k++);
printf("\t%d\n",sum);
}
Version ≥ C11

The _Alignof operator is also allowed to be used with register arrays.

Section 43.3: static

The static storage class serves different purposes, depending on the location of the declaration in the file:

1. To confine the identifier to that translation unit only (scope=file).

/* No other translation unit can use this variable. */

static int i;

/* Same; static is attached to the function type of f, not the return type int. */
static int f(int n);

2. To save data for use with the next call of a function (scope=block):

void foo()
{
static int a = 0; /* has static storage duration and its lifetime is the
* entire execution of the program; initialized to 0 on
* first function call */
int b = 0; /* b has block scope and has automatic storage duration and
* only "exists" within function */

a += 10;
b += 10;

printf("static int a = %d, int b = %d\n", a, b);

}

int main(void)
{
int i;
for (i = 0; i < 5; i++)
{
foo();
}

return 0;
}

This code prints:

static int a = 10, int b = 10

static int a = 20, int b = 10
static int a = 30, int b = 10
static int a = 40, int b = 10
static int a = 50, int b = 10

Static variables retain their value even when called from multiple different threads.

GoalKicker.com – C Notes for Professionals 242

Version ≥ C99

3. Used in function parameters to denote an array is expected to have a constant minimum number of
elements and a non-null parameter:

/* a is expected to have at least 512 elements. */

void printInts(int a[static 512])
{
size_t i;
for (i = 0; i < 512; ++i)
printf("%d\n", a[i]);
}

The required number of items (or even a non-null pointer) is not necessarily checked by the compiler, and
compilers are not required to notify you in any way if you don't have enough elements. If a programmer
passes fewer than 512 elements or a null pointer, undefined behavior is the result. Since it is impossible to
enforce this, extra care must be used when passing a value for that parameter to such a function.

Section 43.4: typedef

Defines a new type based on an existing type. Its syntax mirrors that of a variable declaration.

/* Byte can be used wherever `unsigned char` is needed */

typedef unsigned char Byte;

/* Integer is the type used to declare an array consisting of a single int */

typedef int Integer[1];

/* NodeRef is a type used for pointers to a structure type with the tag "node" */
typedef struct node *NodeRef;

/* SigHandler is the function pointer type that gets passed to the signal function. */
typedef void (*SigHandler)(int);

While not technically a storage class, a compiler will treat it as one since none of the other storage classes are
allowed if the typedef keyword is used.

The typedefs are important and should not be substituted with #define macro.

typedef int newType;

newType *ptr; // ptr is pointer to variable of type 'newType' aka int

However,

#define int newType

newType *ptr; // Even though macros are exact replacements to words, this doesn't result to
a pointer to variable of type 'newType' aka int

Section 43.5: extern

Used to declare an object or function that is defined elsewhere (and that has external linkage). In general, it is
used to declare an object or function to be used in a module that is not the one in which the corresponding object
or function is defined:

GoalKicker.com – C Notes for Professionals 243

/* file1.c */
int foo = 2; /* Has external linkage since it is declared at file scope. */

/* file2.c */
#include <stdio.h>
int main(void)
{
/* `extern` keyword refers to external definition of `foo`. */
extern int foo;
printf("%d\n", foo);
return 0;
}
Version ≥ C99

Things get slightly more interesting with the introduction of the inline keyword in C99:

/* Should usually be place in a header file such that all users see the definition */
/* Hints to the compiler that the function `bar` might be inlined */
/* and suppresses the generation of an external symbol, unless stated otherwise. */
inline void bar(int drink)
{
printf("You ordered drink no.%d\n", drink);
}

/* To be found in just one .c file.

Creates an external function definition of `bar` for use by other files.
The compiler is allowed to choose between the inline version and the external
definition when `bar` is called. Without this line, `bar` would only be an inline
function, and other files would not be able to call it. */
extern void bar(int);

Section 43.6: _Thread_local

Version ≥ C11

This was a new storage specifier introduced in C11 along with multi-threading. This isn't available in earlier C
standards.

Denotes thread storage duration. A variable declared with _Thread_local storage specifier denotes that the object is
local to that thread and its lifetime is the entire execution of the thread in which it's created. It can also appear along
with static or extern.

#include <threads.h>
#include <stdio.h>
#define SIZE 5

int thread_func(void *id)

{
/* thread local variable i. */
static _Thread_local int i;

/* Prints the ID passed from main() and the address of the i.

* Running this program will print different addresses for i, showing
* that they are all distinct objects. */
printf("From thread:[%d], Address of i (thread local): %p\n", *(int*)id, (void*)&i);

return 0;
}

int main(void)

GoalKicker.com – C Notes for Professionals 244

{
thrd_t id[SIZE];
int arr[SIZE] = {1, 2, 3, 4, 5};

/* create 5 threads. */
for(int i = 0; i < SIZE; i++) {
thrd_create(&id[i], thread_func, &arr[i]);
}

/* wait for threads to complete. */

for(int i = 0; i < SIZE; i++) {
thrd_join(id[i], NULL);
}
}

GoalKicker.com – C Notes for Professionals 245

Chapter 44: Declarations
Section 44.1: Calling a function from another C file
foo.h

#ifndef FOO_DOT_H /* This is an "include guard" */

#define FOO_DOT_H /* prevents the file from being included twice. */
/* Including a header file twice causes all kinds */
/* of interesting problems.*/

/**
* This is a function declaration.
* It tells the compiler that the function exists somewhere.
*/
void foo(int id, char *name);

#endif /* FOO_DOT_H */

foo.c

#include "foo.h" /* Always include the header file that declares something
* in the C file that defines it. This makes sure that the
* declaration and definition are always in-sync. Put this
* header first in foo.c to ensure the header is self-contained.
*/
#include <stdio.h>

/**
* This is the function definition.
* It is the actual body of the function which was declared elsewhere.
*/
void foo(int id, char *name)
{
fprintf(stderr, "foo(%d, \"%s\");\n", id, name);
/* This will print how foo was called to stderr - standard error.
* e.g., foo(42, "Hi!") will print `foo(42, "Hi!")`
*/
}

main.c

#include "foo.h"

int main(void)
{
foo(42, "bar");
return 0;
}

Compile and Link

First, we compile both foo.c and main.c to object files. Here we use the gcc compiler, your compiler may have a
different name and need other options.

$ gcc -Wall -c foo.c

$ gcc -Wall -c main.c

GoalKicker.com – C Notes for Professionals 246

Now we link them together to produce our final executable:

$ gcc -o testprogram foo.o main.o

Section 44.2: Using a Global Variable

Use of global variables is generally discouraged. It makes your program more difficult to understand, and harder to
debug. But sometimes using a global variable is acceptable.

global.h

#ifndef GLOBAL_DOT_H /* This is an "include guard" */

#define GLOBAL_DOT_H

/**
* This tells the compiler that g_myglobal exists somewhere.
* Without "extern", this would create a new variable named
* g_myglobal in _every file_ that included it. Don't miss this!
*/
extern int g_myglobal; /* _Declare_ g_myglobal, that is promise it will be _defined_ by
* some module. */

#endif /* GLOBAL_DOT_H */

global.c

#include "global.h" /* Always include the header file that declares something
* in the C file that defines it. This makes sure that the
* declaration and definition are always in-sync.
*/

int g_myglobal; /* _Define_ my_global. As living in global scope it gets initialised to 0

* on program start-up. */

main.c

#include "global.h"

int main(void)
{
g_myglobal = 42;
return 0;
}

See also How do I use extern to share variables between source files?

Section 44.3: Introduction

Example of declarations are:

int a; /* declaring single identifier of type int */

The above declaration declares single identifier named a which refers to some object with int type.

int a1, b1; /* declaring 2 identifiers of type int */

GoalKicker.com – C Notes for Professionals 247

The second declaration declares 2 identifiers named a1 and b1 which refers to some other objects though with the
same int type.

Basically, the way this works is like this - first you put some type, then you write a single or multiple expressions
separated via comma (,) (which will not be evaluated at this point - and which should otherwise be referred
to as declarators in this context). In writing such expressions, you are allowed to apply only the indirection (*),
function call (( )) or subscript (or array indexing - [ ]) operators onto some identifier (you can also not use any
operators at all). The identifier used is not required to be visible in the current scope. Some examples:

/* 1 */ int /* 2 */ (*z) /* 3 */ , /* 4 */ *x , /* 5 */ **c /* 6 */ ;

# Description
1 The name of integer type.
2 Un-evaluated expression applying indirection to some identifier z.
3 We have a comma indicating that one more expression will follow in the same declaration.
4 Un-evaluated expression applying indirection to some other identifier x.
5 Un-evaluated expression applying indirection to the value of the expression (*c).
6 End of declaration.

Note that none of the above identifiers were visible prior to this declaration and so the expressions used would not be
valid before it.

After each such expression, the identifier used in it is introduced into the current scope. (If the identifier has
assigned linkage to it, it may also be re-declared with the same type of linkage so that both identifiers refer to the
same object or function)

Additionally, the equal operator sign (=) may be used for initialization. If an unevaluated expression (declarator) is
followed by = inside the declaration - we say that the identifier being introduced is also being initialized. After the =
sign we can put once again some expression, but this time it'll be evaluated and its value will be used as initial for
the object declared.

Examples:

int l = 90; /* the same as: */

int l; l = 90; /* if it the declaration of l was in block scope */

int c = 2, b[c]; /* ok, equivalent to: */

int c = 2; int b[c];

Later in your code, you are allowed to write the exact same expression from the declaration part of the newly
introduced identifier, giving you an object of the type specified at the beginning of the declaration, assuming that
you've assigned valid values to all accessed objects in the way. Examples:

void f()
{
int b2; /* you should be able to write later in your code b2
which will directly refer to the integer object
that b2 identifies */

b2 = 2; /* assign a value to b2 */

printf("%d", b2); /*ok - should print 2*/

int *b3; /* you should be able to write later in your code *b3 */

GoalKicker.com – C Notes for Professionals 248

 b3 = &b2; /* assign valid pointer value to b3 */

printf("%d", *b3); /* ok - should print 2 */

int **b4; /* you should be able to write later in your code **b4 */

b4 = &b3;

printf("%d", **b4); /* ok - should print 2 */

void (*p)(); /* you should be able to write later in your code (*p)() */

p = &f; /* assign a valid pointer value */

(*p)(); /* ok - calls function f by retrieving the

pointer value inside p - p
and dereferencing it - *p
resulting in a function
which is then called - (*p)() -

it is not *p() because else first the () operator is

applied to p and then the resulting void object is
dereferenced which is not what we want here */
}

The declaration of b3 specifies that you can potentially use b3 value as a mean to access some integer object.

Of course, in order to apply indirection (*) to b3, you should also have a proper value stored in it (see pointers for
more info). You should also first store some value into an object before trying to retrieve it (you can see more about
this problem here). We've done all of this in the above examples.

int a3(); /* you should be able to call a3 */

This one tells the compiler that you'll attempt to call a3. In this case a3 refers to function instead of an object. One
difference between object and function is that functions will always have some sort of linkage. Examples:

void f1()
{
{
int f2(); /* 1 refers to some function f2 */
}

{
int f2(); /* refers to the exact same function f2 as (1) */
}
}

In the above example, the 2 declarations refer to the same function f2, whilst if they were declaring objects then in
this context (having 2 different block scopes), they would have be 2 different distinct objects.

int (*a3)(); /* you should be able to apply indirection to `a3` and then call it */

Now it may seems to be getting complicated, but if you know operators precedence you'll have 0 problems reading
the above declaration. The parentheses are needed because the * operator has less precedence then the ( ) one.

In the case of using the subscript operator, the resulting expression wouldn't be actually valid after the declaration
because the index used in it (the value inside [ and ]) will always be 1 above the maximum allowed value for this
object/function.

GoalKicker.com – C Notes for Professionals 249

int a4[5]; /* here a4 shouldn't be accessed using the index 5 later on */

But it should be accessible by all other indexes lower then 5. Examples:

a4[0], a4[1]; a4[4];

a4[5] will result into UB. More information about arrays can be found here.

int (*a5)[5](); /* here a4 could be applied indirection

indexed up to (but not including) 5
and called */

Unfortunately for us, although syntactically possible, the declaration of a5 is forbidden by the current standard.

Section 44.4: Typedef

Typedefs are declarations which have the keyword typedef in front and before the type. E.g.:

typedef int (*(*t0)())[5];

(you can technically put the typedef after the type too - like this int typedef (*(*t0)())[5]; but this is discouraged)

The above declarations declares an identifier for a typedef name. You can use it like this afterwards:

t0 pf;

Which will have the same effect as writing:

int (*(*pf)())[5];

As you can see the typedef name "saves" the declaration as a type to use later for other declarations. This way you
can save some keystrokes. Also as declaration using typedef is still a declaration you are not limited only by the
above example:

t0 (*pf1);

Is the same as:

int (*(**pf1)())[5];

Section 44.5: Using Global Constants

Headers may be used to declare globally used read-only resources, like string tables for example.

Declare those in a separate header which gets included by any file ("Translation Unit") which wants to make use of
them. It's handy to use the same header to declare a related enumeration to identify all string-resources:

resources.h:

#ifndef RESOURCES_H
#define RESOURCES_H

typedef enum { /* Define a type describing the possible valid resource IDs. */
RESOURCE_UNDEFINED = -1, /* To be used to initialise any EnumResourceID typed variable to be

GoalKicker.com – C Notes for Professionals 250

 marked as "not in use", "not in list", "undefined", wtf.
Will say un-initialised on application level, not on language level.
Initialised uninitialised, so to say ;-)
Its like NULL for pointers ;-)*/
RESOURCE_UNKNOWN = 0, /* To be used if the application uses some resource ID,
for which we do not have a table entry defined, a fall back in
case we _need_ to display something, but do not find anything
appropriate. */

/* The following identify the resources we have defined: */

RESOURCE_OK,
RESOURCE_CANCEL,
RESOURCE_ABORT,
/* Insert more here. */

RESOURCE_MAX /* The maximum number of resources defined. */

} EnumResourceID;

extern const char * const resources[RESOURCE_MAX]; /* Declare, promise to anybody who includes
this, that at linkage-time this symbol will be around.
The 1st const guarantees the strings will not change,
the 2nd const guarantees the string-table entries
will never suddenly point somewhere else as set during
initialisation. */
#endif

To actually define the resources created a related .c-file, that is another translation unit holding the actual instances
of the what had been declared in the related header (.h) file:

resources.c:

#include "resources.h" /* To make sure clashes between declaration and definition are
recognised by the compiler include the declaring header into
the implementing, defining translation unit (.c file).

/* Define the resources. Keep the promise made in resources.h. */

const char * const resources[RESOURCE_MAX] = {
"<unknown>",
"OK",
"Cancel",
"Abort"
};

A program using this could look like this:

main.c:

#include <stdlib.h> /* for EXIT_SUCCESS */

#include <stdio.h>

#include "resources.h"

int main(void)
{
EnumResourceID resource_id = RESOURCE_UNDEFINED;

while ((++resource_id) < RESOURCE_MAX)

{

GoalKicker.com – C Notes for Professionals 251

 printf("resource ID: %d, resource: '%s'\n", resource_id, resources[resource_id]);
}

return EXIT_SUCCESS;
}

Compile the three file above using GCC, and link them to become the program file main for example using this:

gcc -Wall -Wextra -pedantic -Wconversion -g main.c resources.c -o main

(use these -Wall -Wextra -pedantic -Wconversion to make the compiler really picky, so you don't miss anything
before posting the code to SO, will say the world, or even worth deploying it into production)

Run the program created:

$ ./main

And get:

resource ID: 0, resource: ''

resource ID: 1, resource: 'OK'
resource ID: 2, resource: 'Cancel'
resource ID: 3, resource: 'Abort'

Section 44.6: Using the right-left or spiral rule to decipher C

declaration
The "right-left" rule is a completely regular rule for deciphering C declarations. It can also be useful in creating
them.

Read the symbols as you encounter them in the declaration...

* as "pointer to" - always on the left side

[] as "array of" - always on the right side
() as "function returning" - always on the right side

How to apply the rule

STEP 1

Find the identifier. This is your starting point. Then say to yourself, "identifier is." You've started your declaration.

STEP 2

Look at the symbols on the right of the identifier. If, say, you find () there, then you know that this is the
declaration for a function. So you would then have "identifier is function returning". Or if you found a [] there, you
would say "identifier is array of". Continue right until you run out of symbols OR hit a right parenthesis ). (If you hit a
left parenthesis (, that's the beginning of a () symbol, even if there is stuff in between the parentheses. More on
that below.)

STEP 3

Look at the symbols to the left of the identifier. If it is not one of our symbols above (say, something like "int"), just
say it. Otherwise, translate it into English using that table above. Keep going left until you run out of symbols OR hit

GoalKicker.com – C Notes for Professionals 252

a left parenthesis (.

Now repeat steps 2 and 3 until you've formed your declaration.

Here are some examples:

int *p[];

First, find identifier:

int *p[];
^

"p is"

Now, move right until out of symbols or right parenthesis hit.

int *p[];
^^

"p is array of"

Can't move right anymore (out of symbols), so move left and find:

int *p[];
^

"p is array of pointer to"

Keep going left and find:

int *p[];
^^^

"p is array of pointer to int".

(or "p is an array where each element is of type pointer to int")

Another example:

int *(*func())();

Find the identifier.

int *(*func())();
^^^^

"func is"

Move right.

int *(*func())();
^^

"func is function returning"

GoalKicker.com – C Notes for Professionals 253

Can't move right anymore because of the right parenthesis, so move left.

int *(*func())();
^

"func is function returning pointer to"

Can't move left anymore because of the left parenthesis, so keep going right.

int *(*func())();
^^

"func is function returning pointer to function returning"

Can't move right anymore because we're out of symbols, so go left.

int *(*func())();
^

"func is function returning pointer to function returning pointer to"

And finally, keep going left, because there's nothing left on the right.

int *(*func())();
^^^

"func is function returning pointer to function returning pointer to int".

As you can see, this rule can be quite useful. You can also use it to sanity check yourself while you are creating
declarations, and to give you a hint about where to put the next symbol and whether parentheses are required.

Some declarations look much more complicated than they are due to array sizes and argument lists in prototype
form. If you see [3], that's read as "array (size 3) of...". If you see (char *,int) that's read as *"function expecting
(char ,int) and returning...".

Here's a fun one:

int (*(*fun_one)(char *,double))[9][20];

I won't go through each of the steps to decipher this one.

*"fun_one is pointer to function expecting (char ,double) and returning pointer to array (size 9) of array (size 20) of int."

As you can see, it's not as complicated if you get rid of the array sizes and argument lists:

int (*(*fun_one)())[][];

You can decipher it that way, and then put in the array sizes and argument lists later.

Some final words:

It is quite possible to make illegal declarations using this rule, so some knowledge of what's legal in C is necessary.
For instance, if the above had been:

int *((*fun_one)())[][];

GoalKicker.com – C Notes for Professionals 254

it would have read "fun_one is pointer to function returning array of array of pointer to int". Since a function cannot
return an array, but only a pointer to an array, that declaration is illegal.

Illegal combinations include:

[]() - cannot have an array of functions

()() - cannot have a function that returns a function
()[] - cannot have a function that returns an array

In all the above cases, you would need a set of parentheses to bind a * symbol on the left between these () and []
right-side symbols in order for the declaration to be legal.

Here are some more examples:

Legal

int i; an int
int *p; an int pointer (ptr to an int)
int a[]; an array of ints
int f(); a function returning an int
int **pp; a pointer to an int pointer (ptr to a ptr to an int)
int (*pa)[]; a pointer to an array of ints
int (*pf)(); a pointer to a function returning an int
int *ap[]; an array of int pointers (array of ptrs to ints)
int aa[][]; an array of arrays of ints
int *fp(); a function returning an int pointer
int ***ppp; a pointer to a pointer to an int pointer
int (**ppa)[]; a pointer to a pointer to an array of ints
int (**ppf)(); a pointer to a pointer to a function returning an int
int *(*pap)[]; a pointer to an array of int pointers
int (*paa)[][]; a pointer to an array of arrays of ints
int *(*pfp)(); a pointer to a function returning an int pointer
int **app[]; an array of pointers to int pointers
int (*apa[])[]; an array of pointers to arrays of ints
int (*apf[])(); an array of pointers to functions returning an int
int *aap[][]; an array of arrays of int pointers
int aaa[][][]; an array of arrays of arrays of int
int **fpp(); a function returning a pointer to an int pointer
int (*fpa())[]; a function returning a pointer to an array of ints
int (*fpf())(); a function returning a pointer to a function returning an int

Illegal

int af[](); an array of functions returning an int

int fa()[]; a function returning an array of ints
int ff()(); a function returning a function returning an int
int (*pfa)()[]; a pointer to a function returning an array of ints
int aaf[][](); an array of arrays of functions returning an int
int (*paf)[](); a pointer to a an array of functions returning an int
int (*pff)()(); a pointer to a function returning a function returning an int
int *afp[](); an array of functions returning int pointers
int afa[]()[]; an array of functions returning an array of ints
int aff[]()(); an array of functions returning functions returning an int
int *fap()[]; a function returning an array of int pointers
int faa()[][]; a function returning an array of arrays of ints
int faf()[](); a function returning an array of functions returning an int
int *ffp()(); a function returning a function returning an int pointer

Source: http://ieng9.ucsd.edu/~cs30x/rt_lt.rule.html

GoalKicker.com – C Notes for Professionals 255

Chapter 45: Structure Padding and
Packing
By default, C compilers lay out structures so that each member can be accessed fast, without incurring penalties for
'unaligned access, a problem with RISC machines such as the DEC Alpha, and some ARM CPUs.

Depending on the CPU architecture and the compiler, a structure may occupy more space in memory than the sum
of the sizes of its component members. The compiler can add padding between members or at the end of the
structure, but not at the beginning.

Packing overrides the default padding.

Section 45.1: Packing structures

By default structures are padded in C. If you want to avoid this behaviour, you have to explicitly request it. Under
GCC it's __attribute__((__packed__)). Consider this example on a 64-bit machine:

struct foo {
char *p; /* 8 bytes */
char c; /* 1 byte */
long x; /* 8 bytes */
};

The structure will be automatically padded to have8-byte alignment and will look like this:

struct foo {
char *p; /* 8 bytes */
char c; /* 1 byte */

char pad[7]; /* 7 bytes added by compiler */

long x; /* 8 bytes */
};

So sizeof(struct foo) will give us 24 instead of 17. This happened because of a 64 bit compiler read/write from/to
Memory in 8 bytes of word in each step and obvious when try to write char c; a one byte in memory a complete 8
bytes (i.e. word) fetched and consumes only first byte of it and its seven successive of bytes remains empty and not
accessible for any read and write operation for structure padding.

Structure packing

But if you add the attribute packed, the compiler will not add padding:

struct __attribute__((__packed__)) foo {

char *p; /* 8 bytes */
char c; /* 1 byte */
long x; /* 8 bytes */
};

Now sizeof(struct foo) will return 17.

Generally packed structures are used:

To save space.

GoalKicker.com – C Notes for Professionals 256

 To format a data structure to transmit over network without depending on each architecture alignment of
each node of the network.

It must be taken in consideration that some processors such as the ARM Cortex-M0 do not allow unaligned memory
access; in such cases, structure packing can lead to undefined behaviour and can crash the CPU.

Section 45.2: Structure padding

Suppose this struct is defined and compiled with a 32 bit compiler:

struct test_32 {
int a; // 4 byte
short b; // 2 byte
int c; // 4 byte
} str_32;

We might expect this struct to occupy only 10 bytes of memory, but by printing sizeof(str_32) we see it uses 12
bytes.

This happened because the compiler aligns variables for fast access. A common pattern is that when the base type
occupies N bytes (where N is a power of 2 such as 1, 2, 4, 8, 16 — and seldom any bigger), the variable should be
aligned on an N-byte boundary (a multiple of N bytes).

For the structure shown with sizeof(int) == 4 and sizeof(short) == 2, a common layout is:

int a; stored at offset 0; size 4.

short b; stored at offset 4; size 2.
unnamed padding at offset 6; size 2.
int c; stored at offset 8; size 4.

Thus struct test_32 occupies 12 bytes of memory. In this example, there is no trailing padding.

The compiler will ensure that any struct test_32 variables are stored starting on a 4-byte boundary, so that the
members within the structure will be properly aligned for fast access. Memory allocation functions such as
malloc(), calloc() and realloc() are required to ensure that the pointer returned is sufficiently well aligned for
use with any data type, so dynamically allocated structures will be properly aligned too.

You can end up with odd situations such as on a 64-bit Intel x86_64 processor (e.g. Intel Core i7 — a Mac running
macOS Sierra or Mac OS X), where when compiling in 32-bit mode, the compilers place double aligned on a 4-byte
boundary; but, on the same hardware, when compiling in 64-bit mode, the compilers place double aligned on an 8-
byte boundary.

GoalKicker.com – C Notes for Professionals 257

Chapter 46: Memory management
name description
size (malloc, realloc and total size of the memory in bytes. For aligned_alloc the size must be a
aligned_alloc) integral multiple of alignment.
size (calloc) size of each element
nelements number of elements
pointer to allocated memory previously returned by malloc, calloc, realloc
ptr
or aligned_alloc
alignment alignment of allocated memory

For managing dynamically allocated memory, the standard C library provides the functions malloc(), calloc(),
realloc() and free(). In C99 and later, there is also aligned_alloc(). Some systems also provide alloca().

Section 46.1: Allocating Memory

Standard Allocation

The C dynamic memory allocation functions are defined in the <stdlib.h> header. If one wishes to allocate
memory space for an object dynamically, the following code can be used:

int *p = malloc(10 * sizeof *p);

if (p == NULL)
{
perror("malloc() failed");
return -1;
}

This computes the number of bytes that ten ints occupy in memory, then requests that many bytes from malloc
and assigns the result (i.e., the starting address of the memory chunk that was just created using malloc) to a
pointer named p.

It is good practice to use sizeof to compute the amount of memory to request since the result of sizeof is
implementation defined (except for character types, which are char, signed char and unsigned char, for which
sizeof is defined to always give 1).

Because malloc might not be able to service the request, it might return a null pointer. It is important to
check for this to prevent later attempts to dereference the null pointer.

Memory dynamically allocated using malloc() may be resized using realloc() or, when no longer needed,
released using free().

Alternatively, declaring int array[10]; would allocate the same amount of memory. However, if it is declared
inside a function without the keyword static, it will only be usable within the function it is declared in and the
functions it calls (because the array will be allocated on the stack and the space will be released for reuse when the
function returns). Alternatively, if it is defined with static inside a function, or if it is defined outside any function,
then its lifetime is the lifetime of the program. Pointers can also be returned from a function, however a function in
C can not return an array.

Zeroed Memory

The memory returned by malloc may not be initialized to a reasonable value, and care should be taken to zero the
memory with memset or to immediately copy a suitable value into it. Alternatively, calloc returns a block of the

GoalKicker.com – C Notes for Professionals 258

desired size where all bits are initialized to 0. This need not be the same as the representation of floating-point zero
or a null pointer constant.

int *p = calloc(10, sizeof *p);

if (p == NULL)
{
perror("calloc() failed");
return -1;
}

A note on calloc: Most (commonly used) implementations will optimise calloc() for performance, so it will be
faster than calling malloc(), then memset(), even though the net effect is identical.

Aligned Memory
Version ≥ C11

C11 introduced a new function aligned_alloc() which allocates space with the given alignment. It can be used if
the memory to be allocated is needed to be aligned at certain boundaries which can't be satisfied by malloc() or
calloc(). malloc() and calloc() functions allocate memory that's suitably aligned for any object type (i.e. the
alignment is alignof(max_align_t)). But with aligned_alloc() greater alignments can be requested.

/* Allocates 1024 bytes with 256 bytes alignment. */

char *ptr = aligned_alloc(256, 1024);
if (ptr) {
perror("aligned_alloc()");
return -1;
}
free(ptr);

The C11 standard imposes two restrictions: 1) the size (second argument) requested must be an integral multiple of
the alignment (first argument) and 2) the value of alignment should be a valid alignment supported by the
implementation. Failure to meet either of them results in undefined behavior.

Section 46.2: Freeing Memory

It is possible to release dynamically allocated memory by calling free().

int *p = malloc(10 * sizeof *p); /* allocation of memory */

if (p == NULL)
{
perror("malloc failed");
return -1;
}

free(p); /* release of memory */

/* note that after free(p), even using the *value* of the pointer p
has undefined behavior, until a new value is stored into it. */

/* reusing/re-purposing the pointer itself */

int i = 42;
p = &i; /* This is valid, has defined behaviour */

The memory pointed to by p is reclaimed (either by the libc implementation or by the underlying OS) after the call
to free(), so accessing that freed memory block via p will lead to undefined behavior. Pointers that reference
memory elements that have been freed are commonly called dangling pointers, and present a security risk.
Furthermore, the C standard states that even accessing the value of a dangling pointer has undefined behavior.

GoalKicker.com – C Notes for Professionals 259

Note that the pointer p itself can be re-purposed as shown above.

Please note that you can only call free() on pointers that have directly been returned from the malloc(),
calloc(), realloc() and aligned_alloc() functions, or where documentation tells you the memory has been
allocated that way (functions like strdup () are notable examples). Freeing a pointer that is,

obtained by using the & operator on a variable, or

in the middle of an allocated block,

is forbidden. Such an error will usually not be diagnosed by your compiler but will lead the program execution in an
undefined state.

There are two common strategies to prevent such instances of undefined behavior.

The first and preferable is simple - have p itself cease to exist when it is no longer needed, for example:

if (something_is_needed())
{

int *p = malloc(10 * sizeof *p);

if (p == NULL)
{
perror("malloc failed");
return -1;
}

/* do whatever is needed with p */

free(p);
}

By calling free() directly before the end of the containing block (i.e. the }), p itself ceases to exist. The compiler will
give a compilation error on any attempt to use p after that.

A second approach is to also invalidate the pointer itself after releasing the memory to which it points:

free(p);
p = NULL; // you may also use 0 instead of NULL

Arguments for this approach:

On many platforms, an attempt to dereference a null pointer will cause instant crash: Segmentation fault.
Here, we get at least a stack trace pointing to the variable that was used after being freed.

Without setting pointer to NULL we have dangling pointer. The program will very likely still crash, but later,
because the memory to which the pointer points will silently be corrupted. Such bugs are difficult to trace
because they can result in a call stack that completely unrelated to the initial problem.

This approach hence follows the fail-fast concept.

It is safe to free a null pointer. The C Standard specifies that free(NULL) has no effect:

The free function causes the space pointed to by ptr to be deallocated, that is, made available for
further allocation. If ptr is a null pointer, no action occurs. Otherwise, if the argument does not

GoalKicker.com – C Notes for Professionals 260

 match a pointer earlier returned by the calloc, malloc, or realloc function, or if the space has
been deallocated by a call to free or realloc, the behavior is undefined.

Sometimes the first approach cannot be used (e.g. memory is allocated in one function, and deallocated
much later in a completely different function)

Section 46.3: Reallocating Memory

You may need to expand or shrink your pointer storage space after you have allocated memory to it. The void
*realloc(void *ptr, size_t size) function deallocates the old object pointed to by ptr and returns a pointer to
an object that has the size specified by size. ptr is the pointer to a memory block previously allocated with malloc,
calloc or realloc (or a null pointer) to be reallocated. The maximal possible contents of the original memory is
preserved. If the new size is larger, any additional memory beyond the old size are uninitialized. If the new size is
shorter, the contents of the shrunken part is lost. If ptr is NULL, a new block is allocated and a pointer to it is
returned by the function.

#include <stdio.h>
#include <stdlib.h>

int main(void)
{
int *p = malloc(10 * sizeof *p);
if (NULL == p)
{
perror("malloc() failed");
return EXIT_FAILURE;
}

p[0] = 42;
p[9] = 15;

/* Reallocate array to a larger size, storing the result into a

* temporary pointer in case realloc() fails. */
{
int *temporary = realloc(p, 1000000 * sizeof *temporary);

/* realloc() failed, the original allocation was not free'd yet. */

if (NULL == temporary)
{
perror("realloc() failed");
free(p); /* Clean up. */
return EXIT_FAILURE;
}

p = temporary;
}

/* From here on, array can be used with the new size it was
* realloc'ed to, until it is free'd. */

/* The values of p[0] to p[9] are preserved, so this will print:

42 15
*/
printf("%d %d\n", p[0], p[9]);

free(p);

GoalKicker.com – C Notes for Professionals 261

 return EXIT_SUCCESS;
}

The reallocated object may or may not have the same address as *p. Therefore it is important to capture the return
value from realloc which contains the new address if the call is successful.

Make sure you assign the return value of realloc to a temporary instead of the original p. realloc will return null in
case of any failure, which would overwrite the pointer. This would lose your data and create a memory leak.

Section 46.4: realloc(ptr, 0) is not equivalent to free(ptr)

realloc is conceptually equivalent to malloc + memcpy + free on the other pointer.

If the size of the space requested is zero, the behavior of realloc is implementation-defined. This is similar for all
memory allocation functions that receive a size parameter of value 0. Such functions may in fact return a non-null
pointer, but that must never be dereferenced.

Thus, realloc(ptr,0) is not equivalent to free(ptr). It may

be a "lazy" implementation and just return ptr

free(ptr), allocate a dummy element and return that
free(ptr) and return 0
just return 0 for failure and do nothing else.

So in particular the latter two cases are indistinguishable by application code.

This means realloc(ptr,0) may not really free/deallocate the memory, and thus it should never be used as a
replacement for free.

Section 46.5: Multidimensional arrays of variable size

Version ≥ C99

Since C99, C has variable length arrays, VLA, that model arrays with bounds that are only known at initialization
time. While you have to be careful not to allocate too large VLA (they might smash your stack), using pointers to VLA
and using them in sizeof expressions is fine.

double sumAll(size_t n, size_t m, double A[n][m]) {

double ret = 0.0;
for (size_t i = 0; i < n; ++i)
for (size_t j = 0; j < m; ++j)
ret += A[i][j]
return ret;
}

int main(int argc, char *argv[argc+1]) {

size_t n = argc*10;
size_t m = argc*8;
double (*matrix)[m] = malloc(sizeof(double[n][m]));
// initialize matrix somehow
double res = sumAll(n, m, matrix);
printf("result is %g\n", res);
free(matrix);
}

Here matrix is a pointer to elements of type double[m], and the sizeof expression with double[n][m] ensures that
it contains space for n such elements.

GoalKicker.com – C Notes for Professionals 262

All this space is allocated contiguously and can thus be deallocated by a single call to free.

The presence of VLA in the language also affects the possible declarations of arrays and pointers in function
headers. Now, a general integer expression is permitted inside the [] of array parameters. For both functions the
expressions in [] use parameters that have declared before in the parameter list. For sumAll these are the lengths
that the user code expects for the matrix. As for all array function parameters in C the innermost dimension is
rewritten to a pointer type, so this is equivalent to the declaration

double sumAll(size_t n, size_t m, double (*A)[m]);

That is, n is not really part of the function interface, but the information can be useful for documentation and it
could also be used by bounds checking compilers to warn about out-of-bounds access.

Likwise, for main, the expression argc+1 is the minimal length that the C standard prescribes for the argv argument.

Note that officially VLA support is optional in C11, but we know of no compiler that implements C11 and that
doesn't have them. You could test with the macro __STDC_NO_VLA__ if you must.

Section 46.6: alloca: allocate memory on stack

Caveat: alloca is only mentioned here for the sake of completeness. It is entirely non-portable (not covered by any
of the common standards) and has a number of potentially dangerous features that make it un-safe for the
unaware. Modern C code should replace it with Variable Length Arrays (VLA).

Manual page

#include <alloca.h>
// glibc version of stdlib.h include alloca.h by default

void foo(int size) {

char *data = alloca(size);
/*
function body;
*/
// data is automatically freed
}

Allocate memory on the stack frame of the caller, the space referenced by the returned pointer is automatically
free'd when the caller function finishes.

While this function is convenient for automatic memory management, be aware that requesting large allocation
could cause a stack overflow, and that you cannot use free with memory allocated with alloca (which could cause
more issue with stack overflow).

For these reason it is not recommended to use alloca inside a loop nor a recursive function.

And because the memory is free'd upon function return you cannot return the pointer as a function result (the
behavior would be undefined).

Summary

call identical to malloc

automatically free'd upon function return
incompatible with free,realloc functions (undefined behavior)
pointer cannot be returned as a function result (undefined behavior)

GoalKicker.com – C Notes for Professionals 263

 allocation size limited by stack space, which (on most machines) is a lot smaller than the heap space available
for use by malloc()
avoid using alloca() and VLAs (variable length arrays) in a single function
alloca() is not as portable as malloc() et al

Recommendation

Do not use alloca() in new code

Version ≥ C99

Modern alternative.

void foo(int size) {

char data[size];
/*
function body;
*/
// data is automatically freed
}

This works where alloca() does, and works in places where alloca() doesn't (inside loops, for example). It does
assume either a C99 implementation or a C11 implementation that does not define __STDC_NO_VLA__.

Section 46.7: User-defined memory management

malloc() often calls underlying operating system functions to obtain pages of memory. But there is nothing special
about the function and it can be implemented in straight C by declaring a large static array and allocating from it
(there is a slight difficulty in ensuring correct alignment, in practice aligning to 8 bytes is almost always adequate).

To implement a simple scheme, a control block is stored in the region of memory immediately before the pointer to
be returned from the call. This means that free() may be implemented by subtracting from the returned pointer
and reading off the control information, which is typically the block size plus some information that allows it to be
put back in the free list - a linked list of unallocated blocks.

When the user requests an allocation, the free list is searched until a block of identical or larger size to the amount
requested is found, then if necessary it is split. This can lead to memory fragmentation if the user is continually
making many allocations and frees of unpredictable size and and at unpredictable intervals (not all real programs
behave like that, the simple scheme is often adequate for small programs).

/* typical control block */

struct block
{
size_t size; /* size of block */
struct block *next; /* next block in free list */
struct block *prev; /* back pointer to previous block in memory */
void *padding; /* need 16 bytes to make multiple of 8 */
}

static struct block arena[10000]; /* allocate from here */

static struct block *firstfree;

Many programs require large numbers of allocations of small objects of the same size. This is very easy to
implement. Simply use a block with a next pointer. So if a block of 32 bytes is required:

union block

GoalKicker.com – C Notes for Professionals 264

{
union block * next;
unsigned char payload[32];
}

static union block arena[100];

static union block * head;
void init(void)
{
int i;
for (i = 0; i < 100 - 1; i++)
arena[i].next = &arena[i + 1];
arena[i].next = 0; /* last one, null */
head = &block[0];
}

void *block_alloc()
{
void *answer = head;
if (answer)
head = head->next;
return answer;
}

void block_free(void *ptr)

{
union block *block = ptr;
block->next = head;
head - block;
}

This scheme is extremely fast and efficient, and can be made generic with a certain loss of clarity.

GoalKicker.com – C Notes for Professionals 265

Chapter 47: Implementation-defined
behaviour
Section 47.1: Right shift of a negative integer
int signed_integer = -1;

// The right shift operation exhibits implementation-defined behavior:

int result = signed_integer >> 1;

Section 47.2: Assigning an out-of-range value to an integer

// Supposing SCHAR_MAX, the maximum value that can be represented by a signed char, is
// 127, the behavior of this assignment is implementation-defined:
signed char integer;
integer = 128;

Section 47.3: Allocating zero bytes

// The allocation functions have implementation-defined behavior when the requested size
// of the allocation is zero.
void *p = malloc(0);

Section 47.4: Representation of signed integers

Each signed integer type may be represented in any one of three formats; it is implementation-defined which one is
used. The implementation in use for any given signed integer type at least as wide as int can be determined at
runtime from the two lowest-order bits of the representation of value -1 in that type, like so:

enum { sign_magnitude = 1, ones_compl = 2, twos_compl = 3, };

#define SIGN_REP(T) ((T)-1 & (T)3)

switch (SIGN_REP(long)) {
case sign_magnitude: { /* do something */ break; }
case ones_compl: { /* do otherwise */ break; }
case twos_compl: { /* do yet else */ break; }
case 0: { _Static_assert(SIGN_REP(long), "bogus sign representation"); }
}

The same pattern applies to the representation of narrower types, but they cannot be tested by this technique
because the operands of & are subject to "the usual arithmetic conversions" before the result is computed.

GoalKicker.com – C Notes for Professionals 266

Chapter 48: Atomics
Section 48.1: atomics and operators
Atomic variables can be accessed concurrently between different threads without creating race conditions.

/* a global static variable that is visible by all threads */

static unsigned _Atomic active = ATOMIC_VAR_INIT(0);

int myThread(void* a) {
++active; // increment active race free
// do something
--active; // decrement active race free
return 0;
}

All lvalue operations (operations that modify the object) that are allowed for the base type are allowed and will not
lead to race conditions between different threads that access them.

Operations on atomic objects are generally orders of magnitude slower than normal arithmetic operations.
This also includes simple load or store operations. So you should only use them for critical tasks.
Usual arithmetic operations and assignment such as a = a+1; are in fact three operations on a: first a load,
then addition and finally a store. This is not race free. Only the operation a += 1; and a++; are.

GoalKicker.com – C Notes for Professionals 267

Chapter 49: Jump Statements
Section 49.1: Using return
Returning a value

One commonly used case: returning from main()

#include <stdlib.h> /* for EXIT_xxx macros */

int main(int argc, char ** argv)

{
if (2 < argc)
{
return EXIT_FAILURE; /* The code expects one argument:
leave immediately skipping the rest of the function's code */
}

/* Do stuff. */

return EXIT_SUCCESS;
}

Additional notes:

1. For a function having a return type as void (not including void * or related types), the return statement
should not have any associated expression; i.e, the only allowed return statement would be return;.

2. For a function having a non-void return type, the return statement shall not appear without an expression.

3. For main() (and only for main()), an explicit return statement is not required (in C99 or later). If the execution
reaches the terminating }, an implicit value of 0 is returned. Some people think omitting this return is bad
practice; others actively suggest leaving it out.

Returning nothing

Returning from a void function

void log(const char * message_to_log)

{
if (NULL == message_to_log)
{
return; /* Nothing to log, go home NOW, skip the logging. */
}

fprintf(stderr, "%s:%d %s\n", __FILE__, _LINE__, message_to_log);

return; /* Optional, as this function does not return a value. */

}

Section 49.2: Using goto to jump out of nested loops

Jumping out of nested loops would usually require use of a boolean variable with a check for this variable in the
loops. Supposing we are iterating over i and j, it could look like this

size_t i,j;
for (i = 0; i < myValue && !breakout_condition; ++i) {

GoalKicker.com – C Notes for Professionals 268

 for (j = 0; j < mySecondValue && !breakout_condition; ++j) {
... /* Do something, maybe modifying breakout_condition */
/* When breakout_condition == true the loops end */
}
}

But the C language offers the goto clause, which can be useful in this case. By using it with a label declared after the
loops, we can easily break out of the loops.

size_t i,j;
for (i = 0; i < myValue; ++i) {
for (j = 0; j < mySecondValue; ++j) {
...
if(breakout_condition)
goto final;
}
}
final:

However, often when this need comes up a return could be better used instead. This construct is also considered
"unstructured" in structural programming theory.

Another situation where goto might be useful is for jumping to an error-handler:

ptr = malloc(N * x);

if(!ptr)
goto out_of_memory;

/* normal processing */
free(ptr);
return SUCCESS;

out_of_memory:
free(ptr); /* harmless, and necessary if we have further errors */
return FAILURE;

Use of goto keeps error flow separate from normal program control flow. It is however also considered
"unstructured" in the technical sense.

Section 49.3: Using break and continue

Immediately continue reading on invalid input or break on user request or end-of-file:

#include <stdlib.h> /* for EXIT_xxx macros */

#include <stdio.h> /* for printf() and getchar() */
#include <ctype.h> /* for isdigit() */

void flush_input_stream(FILE * fp);

int main(void)
{
int sum = 0;
printf("Enter digits to be summed up or 0 to exit:\n");

do
{
int c = getchar();
if (EOF == c)

GoalKicker.com – C Notes for Professionals 269

 {
printf("Read 'end-of-file', exiting!\n");

break;
}

if ('\n' != c)
{
flush_input_stream(stdin);
}

if (!isdigit(c))
{
printf("%c is not a digit! Start over!\n", c);

continue;
}

if ('0' == c)
{
printf("Exit requested.\n");

break;
}

sum += c - '0';

printf("The current sum is %d.\n", sum);

} while (1);

return EXIT_SUCCESS;
}

void flush_input_stream(FILE * fp)

{
size_t i = 0;
int c;
while ((c = fgetc(fp)) != '\n' && c != EOF) /* Pull all until and including the next new-line. */
{
++i;
}

if (0 != i)
{
fprintf(stderr, "Flushed %zu characters from input.\n", i);
}
}

GoalKicker.com – C Notes for Professionals 270

Chapter 50: Create and include header
files
In modern C, header files are crucial tools that must be designed and used correctly. They allow the compiler to
cross-check independently compiled parts of a program.

Headers declare types, functions, macros etc that are needed by the consumers of a set of facilities. All the code
that uses any of those facilities includes the header. All the code that defines those facilities includes the header.
This allows the compiler to check that the uses and definitions match.

Section 50.1: Introduction

There are a number of guidelines to follow when creating and using header files in a C project:

Idemopotence

If a header file is included multiple times in a translation unit (TU), it should not break builds.

Self-containment

If you need the facilities declared in a header file, you should not have to include any other headers explicitly.

Minimality

You should not be able to remove any information from a header without causing builds to fail.

Include What You Use (IWYU)

Of more concern to C++ than C, but nevertheless important in C too. If the code in a TU (call it code.c)
directly uses the features declared by a header (call it "headerA.h"), then code.c should #include
"headerA.h" directly, even if the TU includes another header (call it "headerB.h") that happens, at the
moment, to include "headerA.h".

Occasionally, there might be good enough reasons to break one or more of these guidelines, but you should both
be aware that you are breaking the rule and be aware of the consequences of doing so before you break it.

Section 50.2: Self-containment

Modern headers should be self-contained, which means that a program that needs to use the facilities defined by
header.h can include that header (#include "header.h") and not worry about whether other headers need to be
included first.

Recommendation: Header files should be self-contained.

Historical rules

Historically, this has been a mildly contentious subject.

GoalKicker.com – C Notes for Professionals 271

Once upon another millennium, the AT&T Indian Hill C Style and Coding Standards stated:

Header files should not be nested. The prologue for a header file should, therefore, describe what other
headers need to be #included for the header to be functional. In extreme cases, where a large number
of header files are to be included in several different source files, it is acceptable to put all common
#includes in one include file.

This is the antithesis of self-containment.

Modern rules

However, since then, opinion has tended in the opposite direction. If a source file needs to use the facilities
declared by a header header.h, the programmer should be able to write:

#include "header.h"

and (subject only to having the correct search paths set on the command line), any necessary pre-requisite headers
will be included by header.h without needing any further headers added to the source file.

This provides better modularity for the source code. It also protects the source from the "guess why this header
was added" conundrum that arises after the code has been modified and hacked for a decade or two.

The NASA Goddard Space Flight Center (GSFC) coding standards for C is one of the more modern standards — but
is now a little hard to track down. It states that headers should be self-contained. It also provides a simple way to
ensure that headers are self-contained: the implementation file for the header should include the header as the
first header. If it is not self-contained, that code will not compile.

The rationale given by GSFC includes:

§2.1.1 Header include rationale

This standard requires a unit’s header to contain #include statements for all other headers required by
the unit header. Placing #include for the unit header first in the unit body allows the compiler to verify
that the header contains all required #include statements.

An alternate design, not permitted by this standard, allows no #include statements in headers; all
#includes are done in the body files. Unit header files then must contain #ifdef statements that check
that the required headers are included in the proper order.

One advantage of the alternate design is that the #include list in the body file is exactly the dependency
list needed in a makefile, and this list is checked by the compiler. With the standard design, a tool must be
used to generate the dependency list. However, all of the branch recommended development
environments provide such a tool.

A major disadvantage of the alternate design is that if a unit’s required header list changes, each file that
uses that unit must be edited to update the #include statement list. Also, the required header list for a

GoalKicker.com – C Notes for Professionals 272

 compiler library unit may be different on different targets.

Another disadvantage of the alternate design is that compiler library header files, and other third party
files, must be modified to add the required #ifdef statements.

Thus, self-containment means that:

If a header header.h needs a new nested header extra.h, you do not have to check every source file that
uses header.h to see whether you need to add extra.h.
If a header header.h no longer needs to include a specific header notneeded.h, you do not have to check
every source file that uses header.h to see whether you can safely remove notneeded.h (but see Include
what you use.
You do not have to establish the correct sequence for including the pre-requisite headers (which requires a
topological sort to do the job properly).

Checking self-containment

See Linking against a static library for a script chkhdr that can be used to test idempotence and self-containment of
a header file.

Section 50.3: Minimality

Headers are a crucial consistency checking mechanism, but they should be as small as possible. In particular, that
means that a header should not include other headers just because the implementation file will need the other
headers. A header should contain only those headers necessary for a consumer of the services described.

For example, a project header should not include <stdio.h> unless one of the function interfaces uses the type
FILE * (or one of the other types defined solely in <stdio.h>). If an interface uses size_t, the smallest header that
suffices is <stddef.h>. Obviously, if another header that defines size_t is included, there is no need to include
<stddef.h> too.

If the headers are minimal, then it keeps the compilation time to a minimum too.

It is possible to devise headers whose sole purpose is to include a lot of other headers. These seldom turn out to be
a good idea in the long run because few source files will actually need all the facilities described by all the headers.
For example, a <standard-c.h> could be devised that includes all the standard C headers — with care since some
headers are not always present. However, very few programs actually use the facilities of <locale.h> or
<tgmath.h>.

See also How to link multiple implementation files in C?

Section 50.4: Notation and Miscellany

The C standard says that there is very little difference between the #include <header.h> and #include
"header.h" notations.

[#include <header.h>] searches a sequence of implementation-defined places for a header identified

uniquely by the specified sequence between the < and > delimiters, and causes the replacement of that
directive by the entire contents of the header. How the places are specified or the header identified is
implementation-defined.

GoalKicker.com – C Notes for Professionals 273

 [#include "header.h"] causes the replacement of that directive by the entire contents of the source file
identified by the specified sequence between the "…" delimiters. The named source file is searched for in
an implementation-defined manner. If this search is not supported, or if the search fails, the directive is
reprocessed as if it read [#include <header.h>] …

So, the double quoted form may look in more places than the angle-bracketed form. The standard specifies by
example that the standard headers should be included in angle-brackets, even though the compilation works if you
use double quotes instead. Similarly, standards such as POSIX use the angle-bracketed format — and you should
too. Reserve double-quoted headers for headers defined by the project. For externally-defined headers (including
headers from other projects your project relies on), the angle-bracket notation is most appropriate.

Note that there should be a space between #include and the header, even though the compilers will accept no
space there. Spaces are cheap.

A number of projects use a notation such as:

#include <openssl/ssl.h>
#include <sys/stat.h>
#include <linux/kernel.h>

You should consider whether to use that namespace control in your project (it is quite probably a good idea). You
should steer clear of the names used by existing projects (in particular, both sys and linux would be bad choices).

If you use this, your code should be careful and consistent in the use of the notation.

Do not use #include "../include/header.h" notation.

Header files should seldom if ever define variables. Although you will keep global variables to a minimum, if you
need a global variable, you will declare it in a header, and define it in one suitable source file, and that source file
will include the header to cross-check the declaration and definition, and all source files that use the variable will
use the header to declare it.

Corollary: you will not declare global variables in a source file — a source file will only contain definitions.

Header files should seldom declare static functions, with the notable exception of static inline functions which
will be defined in headers if the function is needed in more than one source file.

Source files define global variables, and global functions.

Source files do not declare the existence of global variables or functions; they include the header that
declares the variable or function.
Header files declare global variable and functions (and types and other supporting material).
Header files do not define variables or any functions except (static) inline functions.

Cross-references

Where to document functions in C?

List of standard header files in C and C++
Is inline without static or extern ever useful in C99?
How do I use extern to share variables between source files?
What are the benefits of a relative path such as "../include/header.h" for a header?
Header inclusion optimization
Should I include every header?

GoalKicker.com – C Notes for Professionals 274

Section 50.5: Idempotence
If a particular header file is included more than once in a translation unit (TU), there should not be any compilation
problems. This is termed 'idempotence'; your headers should be idempotent. Think how difficult life would be if you
had to ensure that #include <stdio.h> was only included once.

There are two ways to achieve idempotence: header guards and the #pragma once directive.

Header guards

Header guards are simple and reliable and conform to the C standard. The first non-comment lines in a header file
should be of the form:

#ifndef UNIQUE_ID_FOR_HEADER
#define UNIQUE_ID_FOR_HEADER

The last non-comment line should be #endif, optionally with a comment after it:

#endif /* UNIQUE_ID_FOR_HEADER */

All the operational code, including other #include directives, should be between these lines.

Each name must be unique. Often, a name scheme such as HEADER_H_INCLUDED is used. Some older code uses a
symbol defined as the header guard (e.g. #ifndef BUFSIZ in <stdio.h>), but it is not as reliable as a unique name.

One option would be to use a generated MD5 (or other) hash for the header guard name. You should avoid
emulating the schemes used by system headers which frequently use names reserved to the implementation —
names starting with an underscore followed by either another underscore or an upper-case letter.

The #pragma once Directive

Alternatively, some compilers support the #pragma once directive which has the same effect as the three lines
shown for header guards.

#pragma once

The compilers which support #pragma once include MS Visual Studio and GCC and Clang. However, if portability is a
concern, it is better to use header guards, or use both. Modern compilers (those supporting C89 or later) are
required to ignore, without comment, pragmas that they do not recognize ('Any such pragma that is not recognized
by the implementation is ignored') but old versions of GCC were not so indulgent.

Section 50.6: Include What You Use (IWYU)

Google's Include What You Use project, or IWYU, ensures source files include all headers used in the code.

Suppose a source file source.c includes a header arbitrary.h which in turn coincidentally includes freeloader.h,
but the source file also explicitly and independently uses the facilities from freeloader.h. All is well to start with.
Then one day arbitrary.h is changed so its clients no longer need the facilities of freeloader.h. Suddenly,
source.c stops compiling — because it didn't meet the IWYU criteria. Because the code in source.c explicitly used
the facilities of freeloader.h, it should have included what it uses — there should have been an explicit #include
"freeloader.h" in the source too. (Idempotency would have ensured there wasn't a problem.)

The IWYU philosophy maximizes the probability that code continues to compile even with reasonable changes
made to interfaces. Clearly, if your code calls a function that is subsequently removed from the published interface,

GoalKicker.com – C Notes for Professionals 275

no amount of preparation can prevent changes becoming necessary. This is why changes to APIs are avoided when
possible, and why there are deprecation cycles over multiple releases, etc.

This is a particular problem in C++ because standard headers are allowed to include each other. Source file
file.cpp could include one header header1.h that on one platform includes another header header2.h. file.cpp
might turn out to use the facilities of header2.h as well. This wouldn't be a problem initially - the code would
compile because header1.h includes header2.h. On another platform, or an upgrade of the current platform,
header1.h could be revised so it no longer includes header2.h, and thenfile.cpp would stop compiling as a result.

IWYU would spot the problem and recommend that header2.h be included directly in file.cpp. This would ensure
it continues to compile. Analogous considerations apply to C code too.

GoalKicker.com – C Notes for Professionals 276

Chapter 51: <ctype.h> — character
classification & conversion
Section 51.1: Introduction
The header ctype.h is a part of the standard C library. It provides functions for classifying and converting
characters.

All of these functions take one parameter, an int that must be either EOF or representable as an unsigned char.

The names of the classifying functions are prefixed with 'is'. Each returns an integer non-zero value (TRUE) if the
character passed to it satisfies the related condition. If the condition is not satisfied then the function returns a zero
value (FALSE).

These classifying functions operate as shown, assuming the default C locale:

int a;
int c = 'A';
a = isalpha(c); /* Checks if c is alphabetic (A-Z, a-z), returns non-zero here. */
a = isalnum(c); /* Checks if c is alphanumeric (A-Z, a-z, 0-9), returns non-zero here. */
a = iscntrl(c); /* Checks is c is a control character (0x00-0x1F, 0x7F), returns zero here. */
a = isdigit(c); /* Checks if c is a digit (0-9), returns zero here. */
a = isgraph(c); /* Checks if c has a graphical representation (any printing character except space),
returns non-zero here. */
a = islower(c); /* Checks if c is a lower-case letter (a-z), returns zero here. */
a = isprint(c); /* Checks if c is any printable character (including space), returns non-zero here.
*/
a = isupper(c); /* Checks if c is a upper-case letter (a-z), returns zero here. */
a = ispunct(c); /* Checks if c is a punctuation character, returns zero here. */
a = isspace(c); /* Checks if c is a white-space character, returns zero here. */
a = isupper(c); /* Checks if c is an upper-case letter (A-Z), returns non-zero here. */
a = isxdigit(c); /* Checks if c is a hexadecimal digit (A-F, a-f, 0-9), returns non-zero here. */
Version ≥ C99
a = isblank(c); /* Checks if c is a blank character (space or tab), returns non-zero here. */

There are two conversion functions. These are named using the prefix 'to'. These functions take the same argument
as those above. However the return value is not a simple zero or non-zero but the passed argument changed in
some manner.

These conversion functions operate as shown, assuming the default C locale:

int a;
int c = 'A';

/* Converts c to a lower-case letter (a-z).

* If conversion is not possible the unchanged value is returned.
* Returns 'a' here.
*/
a = tolower(c);

/* Converts c to an upper-case letter (A-Z).

* If conversion is not possible the unchanged value is returned.
* Returns 'A' here.
*/
a = toupper(c);

GoalKicker.com – C Notes for Professionals 277

The below information is quoted from cplusplus.com mapping how the original 127-character ASCII set is
considered by each of the classifying type functions (a • indicates that the function returns non-zero for that
character)

ASCII
characters iscntrl isblank isspace isupper islower isalpha isdigit isxdigit isalnum ispunct isgraph isprint
values
0x00 ..
NUL, (other control codes) •
0x08
0x09 tab ('\t') • • •
0x0A .. (white-space control
• •
0x0D codes: '\f','\v','\n','\r')
0x0E ..
(other control codes) •
0x1F
0x20 space (' ') • • •
0x21 ..
!"#$%&'()*+,-./ • • •
0x2F
0x30 ..
0123456789 • • • • •
0x39
0x3a ..
:;<=>?@ • • •
0x40
0x41 ..
ABCDEF • • • • • •
0x46
0x47 ..
GHIJKLMNOPQRSTUVWXYZ • • • • •
0x5A
0x5B ..
[]^_` • • •
0x60
0x61 ..
abcdef • • • • • •
0x66
0x67 ..
ghijklmnopqrstuvwxyz • • • • •
0x7A
0x7B ..
{}~bar • • •
0x7E
0x7F (DEL) •

Section 51.2: Classifying characters read from a stream

#include <ctype.h>
#include <stdio.h>

typedef struct {
size_t space;
size_t alnum;
size_t punct;
} chartypes;

chartypes classify(FILE *f) {

chartypes types = { 0, 0, 0 };
int ch;

while ((ch = fgetc(f)) != EOF) {

types.space += !!isspace(ch);
types.alnum += !!isalnum(ch);
types.punct += !!ispunct(ch);
}

return types;
}

The classify function reads characters from a stream and counts the number of spaces, alphanumeric and

GoalKicker.com – C Notes for Professionals 278

punctuation characters. It avoids several pitfalls.

When reading a character from a stream, the result is saved as an int, since otherwise there would be an
ambiguity between reading EOF (the end-of-file marker) and a character that has the same bit pattern.
The character classification functions (e.g. isspace) expect their argument to be either representable as an
unsigned char, or the value of the EOF macro. Since this is exactly what the fgetc returns, there is no need for
conversion here.
The return value of the character classification functions only distinguishes between zero (meaning false)
and nonzero (meaning true). For counting the number of occurrences, this value needs to be converted to a
1 or 0, which is done by the double negation, !!.

Section 51.3: Classifying characters from a string

#include <ctype.h>
#include <stddef.h>

typedef struct {
size_t space;
size_t alnum;
size_t punct;
} chartypes;

chartypes classify(const char *s) {

chartypes types = { 0, 0, 0 };
const char *p;
for (p= s; p != '\0'; p++) {
types.space += !!isspace((unsigned char)*p);
types.alnum += !!isalnum((unsigned char)*p);
types.punct += !!ispunct((unsigned char)*p);
}

return types;
}

The classify function examines all characters from a string and counts the number of spaces, alphanumeric and
punctuation characters. It avoids several pitfalls.

The character classification functions (e.g. isspace) expect their argument to be either representable as an
unsigned char, or the value of the EOF macro.
The expression *p is of type char and must therefore be converted to match the above wording.
The char type is defined to be equivalent to either signed char or unsigned char.
When char is equivalent to unsigned char, there is no problem, since every possible value of the char type is
representable as unsigned char.
When char is equivalent to signed char, it must be converted to unsigned char before being passed to the
character classification functions. And although the value of the character may change because of this
conversion, this is exactly what these functions expect.
The return value of the character classification functions only distinguishes between zero (meaning false)
and nonzero (meaning true). For counting the number of occurrences, this value needs to be converted to a
1 or 0, which is done by the double negation, !!.

GoalKicker.com – C Notes for Professionals 279

Chapter 52: Side Eects
Section 52.1: Pre/Post Increment/Decrement operators
In C, there are two unary operators - '++' and '--' that are very common source of confusion. The operator ++ is
called the increment operator and the operator -- is called the decrement operator. Both of them can be used used
in either prefix form or postfix form. The syntax for prefix form for ++ operator is ++operand and the syntax for
postfix form is operand++. When used in the prefix form, the operand is incremented first by 1 and the resultant
value of the operand is used in the evaluation of the expression. Consider the following example:

int n, x = 5;
n = ++x; /* x is incremented by 1(x=6), and result is assigned to n(6) */
/* this is a short form for two statements: */
/* x = x + 1; */
/* n = x ; */

When used in the postfix form, the operand's current value is used in the expression and then the value of the
operand is incremented by 1. Consider the following example:

int n, x = 5;
n = x++; /* value of x(5) is assigned first to n(5), and then x is incremented by 1; x(6) */
/* this is a short form for two statements: */
/* n = x; */
/* x = x + 1; */

The working of the decrement operator -- can be understood similarly.

The following code demonstrates what each one does

int main()
{
int a, b, x = 42;
a = ++x; /* a and x are 43 */
b = x++; /* b is 43, x is 44 */
a = x--; /* a is is 44, x is 43 */
b = --x; /* b and x are 42 */

return 0;
}

From the above it is clear that post operators return the current value of a variable and then modify it, but pre
operators modify the variable and then return the modified value.

In all versions of C, the order of evaluation of pre and post operators are not defined, hence the following code can
return unexpected outputs:

int main()
{
int a, x = 42;
a = x++ + x; /* wrong */
a = x + x; /* right */
++x;

int ar[10];
x = 0;
ar[x] = x++; /* wrong */

GoalKicker.com – C Notes for Professionals 280

 ar[x++] = x; /* wrong */
ar[x] = x; /* right */
++x;
return 0;
}

Note that it is also good practice to use pre over post operators when used alone in a statement. Look at the above
code for this.

Note also, that when a function is called, all side effects on arguments must take place before the function runs.

int foo(int x)
{
return x;
}

int main()
{
int a = 42;
int b = foo(a++); /* This returns 43, even if it seems like it should return 42 */
return 0;
}

GoalKicker.com – C Notes for Professionals 281

Chapter 53: Multi-Character Character
Sequence
Section 53.1: Trigraphs
The symbols [ ] { } ^ \ | ~ # are frequently used in C programs, but in the late 1980s, there were code sets in
use (ISO 646 variants, for example, in Scandinavian countries) where the ASCII character positions for these were
used for national language variant characters (e.g. £ for # in the UK; Æ Å æ å ø Ø for { } { } | \ in Denmark;
there was no ~ in EBCDIC). This meant that it was hard to write C code on machines that used these sets.

To solve this problem, the C standard suggested the use of combinations of three characters to produce a single
character called a trigraph. A trigraph is a sequence of three characters, the first two of which are question marks.

The following is a simple example that uses trigraph sequences instead of #, { and }:

??=include <stdio.h>

int main()
??<
printf("Hello World!\n");
??>

This will be changed by the C preprocessor by replacing the trigraphs with their single-character equivalents as if
the code had been written:

#include <stdio.h>

int main()
{
printf("Hello World!\n");
}
Trigraph Equivalent
??= #
??/ \
??' ^
??( [
??) ]
??! |
??< {
??> }
??- ~

Note that trigraphs are problematic because, for example, ??/ is a backslash and can affect the meaning of
continuation lines in comments, and have to be recognized inside strings and character literals (e.g. '??/??/' is a
single character, a backslash).

Section 53.2: Digraphs

Version ≥ C99

In 1994 more readable alternatives to five of the trigraphs were supplied. These use only two characters and are
known as digraphs. Unlike trigraphs, digraphs are tokens. If a digraph occurs in another token (e.g. string literals or

GoalKicker.com – C Notes for Professionals 282

character constants) then it will not be treated as a digraph, but remain as it is.

The following shows the difference before and after processing the digraphs sequence.

#include <stdio.h>

int main()
<%
printf("Hello %> World!\n"); /* Note that the string contains a digraph */
%>

Which will be treated the same as:

#include <stdio.h>

int main()
{
printf("Hello %> World!\n"); /* Note the unchanged digraph within the string. */
}
Digraph Equivalent
<: [
:> ]
<% {
%> }
%: #

GoalKicker.com – C Notes for Professionals 283

Chapter 54: Constraints
Section 54.1: Duplicate variable names in the same scope
An example of a constraint as expressed in the C standard is having two variables of the same name declared in a
scope1), for example:

void foo(int bar)

{
int var;
double var;
}

This code breaches the constraint and must produce a diagnostic message at compile time. This is very useful as
compared to undefined behavior as the developer will be informed of the issue before the program is run,
potentially doing anything.

Constraints thus tend to be errors which are easily detectable at compile time such as this, issues which result in
undefined behavior but would be difficult or impossible to detect at compile time are thus not constraints.

1) exact wording:

Version = C99

If an identifier has no linkage, there shall be no more than one declaration of the identifier (in a declarator or type
specifier) with the same scope and in the same name space, except for tags as specified in 6.7.2.3.

Section 54.2: Unary arithmetic operators

The unary + and - operators are only usable on arithmetic types, therefore if for example one tries to use them on
a struct the program will produce a diagnostic eg:

struct foo
{
bool bar;
};

void baz(void)
{
struct foo testStruct;
-testStruct; /* This breaks the constraint so must produce a diagnostic */
}

GoalKicker.com – C Notes for Professionals 284

Chapter 55: Inlining
Section 55.1: Inlining functions used in more than one source
file
For small functions that get called often, the overhead associated with the function call can be a significant fraction
of the total execution time of that function. One way of improving performance, then, is to eliminate the overhead.

In this example we use four functions (plus main()) in three source files. Two of those (plusfive() and timestwo())
each get called by the other two located in "source1.c" and "source2.c". The main() is included so we have a
working example.

main.c:
#include <stdio.h>
#include <stdlib.h>
#include "headerfile.h"

int main(void) {
int start = 3;
int intermediate = complicated1(start);
printf("First result is %d\n", intermediate);
intermediate = complicated2(start);
printf("Second result is %d\n", intermediate);
return 0;
}

source1.c:
#include <stdio.h>
#include <stdlib.h>
#include "headerfile.h"

int complicated1(int input) {

int tmp = timestwo(input);
tmp = plusfive(tmp);
return tmp;
}

source2.c:
#include <stdio.h>
#include <stdlib.h>
#include "headerfile.h"

int complicated2(int input) {

int tmp = plusfive(input);
tmp = timestwo(tmp);
return tmp;
}

headerfile.h:
#ifndef HEADERFILE_H
#define HEADERFILE_H

int complicated1(int input);

int complicated2(int input);

inline int timestwo(int input) {

return input * 2;
}
inline int plusfive(int input) {

GoalKicker.com – C Notes for Professionals 285

 return input + 5;
}

#endif

Functions timestwo and plusfive get called by both complicated1 and complicated2, which are in different
"translation units", or source files. In order to use them in this way, we have to define them in the header.

Compile like this, assuming gcc:

cc -O2 -std=c99 -c -o main.o main.c

cc -O2 -std=c99 -c -o source1.o source1.c
cc -O2 -std=c99 -c -o source2.o source2.c
cc main.o source1.o source2.o -o main

We use the -O2 optimization option because some compilers don't inline without optimization turned on.

The effect of the inline keyword is that the function symbol in question is not emitted into the object file.
Otherwise an error would occur in the last line, where we are linking the object files to form the final executable. If
we would not have inline, the same symbol would be defined in both .o files, and a "multiply defined symbol"
error would occur.

In situations where the symbol is actually needed, this has the disadvantage that the symbol is not produced at all.
There are two possibilities to deal with that. The first is to add an extra extern declaration of the inlined functions
in exactly one of the .c files. So add the following to source1.c:

extern int timestwo(int input);

extern int plusfive(int input);

The other possibility is to define the function with static inline instead of inline. This method has the drawback
that eventually a copy of the function in question may be produced in every object file that is produced with this
header.

GoalKicker.com – C Notes for Professionals 286

Chapter 56: Unions
Section 56.1: Using unions to reinterpret values
Some C implementations permit code to write to one member of a union type then read from another in order to
perform a sort of reinterpreting cast (parsing the new type as the bit representation of the old one).

It is important to note however, this is not permitted by the C standard current or past and will result in undefined
behavior, none the less is is a very common extension offered by compilers (so check your compiler docs if you plan
to do this).

One real life example of this technique is the "Fast Inverse Square Root" algorithm which relies on implementation
details of IEEE 754 floating point numbers to perform an inverse square root more quickly than using floating point
operations, this algorithm can be performed either through pointer casting (which is very dangerous and breaks
the strict aliasing rule) or through a union (which is still undefined behavior but works in many compilers):

union floatToInt
{
int32_t intMember;
float floatMember; /* Float must be 32 bits IEEE 754 for this to work */
};

float inverseSquareRoot(float input)

{
union floatToInt x;
int32_t i;
float f;
x.floatMember = input; /* Assign to the float member */
i = x.intMember; /* Read back from the integer member */
i = 0x5f3759df - (i >> 1);
x.intMember = i; /* Assign to the integer member */
f = x.floatMember; /* Read back from the float member */
f = f * (1.5f - input * 0.5f * f * f);
return f * (1.5f - input * 0.5f * f * f);
}

This technique was widely used in computer graphics and games in the past due to its greater speed compared to
using floating point operations, and is very much a compromise, losing some accuracy and being very non portable
in exchange for speed.

Section 56.2: Writing to one union member and reading from

another
The members of a union share the same space in memory. This means that writing to one member overwrites the
data in all other members and that reading from one member results in the same data as reading from all other
members. However, because union members can have differing types and sizes, the data that is read can be
interpreted differently, see
http://stackoverflow.com/documentation/c/1119/structs-and-unions/9399/using-unions-to-reinterpret-values

The simple example below demonstrates a union with two members, both of the same type. It shows that writing
to member m_1 results in the written value being read from member m_2 and writing to member m_2 results in the
written value being read from member m_1.

#include <stdio.h>

GoalKicker.com – C Notes for Professionals 287

union my_union /* Define union */
{
int m_1;
int m_2;
};

int main (void)

{
union my_union u; /* Declare union */
u.m_1 = 1; /* Write to m_1 */
printf("u.m_2: %i\n", u.m_2); /* Read from m_2 */
u.m_2 = 2; /* Write to m_2 */
printf("u.m_1: %i\n", u.m_1); /* Read from m_1 */
return 0;
}

Result

u.m_2: 1
u.m_1: 2

Section 56.3: Dierence between struct and union

This illustrates that union members shares memory and that struct members does not share memory.

#include <stdio.h>
#include <string.h>

union My_Union
{
int variable_1;
int variable_2;
};

struct My_Struct
{
int variable_1;
int variable_2;
};

int main (void)

{
union My_Union u;
struct My_Struct s;
u.variable_1 = 1;
u.variable_2 = 2;
s.variable_1 = 1;
s.variable_2 = 2;
printf ("u.variable_1: %i\n", u.variable_1);
printf ("u.variable_2: %i\n", u.variable_2);
printf ("s.variable_1: %i\n", s.variable_1);
printf ("s.variable_2: %i\n", s.variable_2);
printf ("sizeof (union My_Union): %i\n", sizeof (union My_Union));
printf ("sizeof (struct My_Struct): %i\n", sizeof (struct My_Struct));
return 0;
}

GoalKicker.com – C Notes for Professionals 288

Chapter 57: Threads (native)
Section 57.1: Inititialization by one thread
In most cases all data that is accessed by several threads should be initialized before the threads are created. This
ensures that all threads start with a clear state and no race condition occurs.

If this is not possible once_flag and call_once can be used

#include <threads.h>
#include <stdlib.h>

// the user data for this example

double const* Big = 0;

// the flag to protect big, must be global and/or static

static once_flag onceBig = ONCE_INIT;

void destroyBig(void) {
free((void*)Big);
}

void initBig(void) {
// assign to temporary with no const qualification
double* b = malloc(largeNum);
if (!b) {
perror("allocation failed for Big");
exit(EXIT_FAILURE);
}
// now initialize and store Big
initializeBigWithSophisticatedValues(largeNum, b);
Big = b;
// ensure that the space is freed on exit or quick_exit
atexit(destroyBig);
at_quick_exit(destroyBig);
}

// the user thread function that relies on Big

int myThreadFunc(void* a) {
call_once(&onceBig, initBig);
// only use Big from here on
...
return 0;
}

The once_flag is used to coordinate different threads that might want to initialize the same data Big. The call to
call_once guarantees that

initBig is called exactly once

call_once blocks until such a call to initBig has been made, either by the same or another thread.

Besides allocation, a typical thing to do in such a once-called function is a dynamic initialization of a thread control
data structures such as mtx_t or cnd_t that can't be initialized statically, using mtx_init or cnd_init, respectively.

Section 57.2: Start several threads

#include <stdio.h>

GoalKicker.com – C Notes for Professionals 289

#include <threads.h>
#include <stdlib.h>

struct my_thread_data {
double factor;
};

int my_thread_func(void* a) {
struct my_thread_data* d = a;
// do something with d
printf("we found %g\n", d->factor);
// return an success or error code
return d->factor > 1.0;
}

int main(int argc, char* argv[argc+1]) {

unsigned n = 4;
if (argc > 1) n = strtoull(argv[1], 0, 0);
// reserve space for the arguments for the threads
struct my_thread_data D[n]; // can't be initialized
for (unsigned i = 0; i < n; ++i) {
D[i] = (struct my_thread_data){ .factor = 0.5*i, };
}
// reserve space for the ID's of the threads
thrd_t id[4];
// launch the threads
for (unsigned i = 0; i < n; ++i) {
thrd_create(&id[i], my_thread_func, &D[i]);
}
// Wait that all threads have finished, but throw away their
// return values
for (unsigned i = 0; i < n; ++i) {
thrd_join(id[i], 0);
}
return EXIT_SUCCESS;
}

GoalKicker.com – C Notes for Professionals 290

Chapter 58: Multithreading
In C11 there is a standard thread library, <threads.h>, but no known compiler that yet implements it. Thus, to use
multithreading in C you must use platform specific implementations such as the POSIX threads library (often
referred to as pthreads) using the pthread.h header.

Section 58.1: C11 Threads simple example

#include <threads.h>
#include <stdio.h>

int run(void *arg)

{
printf("Hello world of C11 threads.");

return 0;
}

int main(int argc, const char *argv[])

{
thrd_t thread;
int result;

thrd_create(&thread, run, NULL);

thrd_join(&thread, &result);

printf("Thread return %d at the end\n", result);

}

GoalKicker.com – C Notes for Professionals 291