Difference between revisions of "cpp/language/default initialization"

Latest revision as of 23:42, 22 April 2024

This is the initialization performed when an object is constructed with no initializer.

[edit] Syntax

[edit] Explanation

Default-initialization is performed in three situations:

The effects of default-initialization are:

• if T is a (possibly cv-qualified) non-POD(until C++11) class type, the constructors are considered and subjected to overload resolution against the empty argument list. The constructor selected (which is one of the default constructors) is called to provide the initial value for the new object;
• if T is an array type, every element of the array is default-initialized;
• otherwise, no initialization is performed (see notes).

[edit] Default-initialization of a const object

If a program calls for the default-initialization of an object of a const-qualified type T, T shall be a const-default-constructible class type or array thereof.

A class type T is const-default-constructible if default-initialization of T would invoke a user-provided constructor of T (not inherited from a base class)(since C++11) or if

each potentially constructed base class of T is const-default-constructible.

[edit] Indeterminate and erroneous values

If an evaluation produces an indeterminate value, the behavior is undefined.

[edit] Special cases

The following types are uninitialized-friendly:

• unsigned char
• char, if its underlying type is unsigned char

Given an indeterminate or erroneous(since C++26) value value, the uninitialized result value of value is:

• An indeterminate value, if value is also an indeterminate value.

If an evaluation eval produces an indeterminate or erroneous(since C++26) value value of an uninitialized-friendly type, the behavior is well-defined in the following cases:

• eval is the evaluation of one of the following expressions and operands:

• The second or third operand of a conditional expression.
• The right operand of a comma expression.
• The operand of an integral conversion, explicit cast or static_cast to an uninitialized-friendly type.
• A discarded-value expression.

In this case, the result of the operation is the uninitialized result value of value.

• eval is an evaluation of the right operand of a simple assignment operator whose left operand is an lvalue of an uninitialized-friendly type.

In this case, the value of the object referred to by the left operand is replaced by the uninitialized result value of value.

• eval is the evaluation of the initialization expression when initializing an object of an uninitialized-friendly type.

In this case, that object is initialized to the uninitialized result value of value.

Converting an indeterminate value of an uninitialized-friendly type produces an indeterminate value.

// Case 1: Uninitialized objects with dynamic storage duration
// All C++ versions: indeterminate value + undefined behavior
int f(bool b)
{
    unsigned char* c = new unsigned char;
    unsigned char d = *c; // OK, “d” has an indeterminate value
    int e = d;            // undefined behavior
    return b ? d : 0;     // undefined behavior if “b” is true
}
 
// Case 2: Uninitialized objects with automatic storage duration
// until C++26: indeterminate value + undefined behavior
// since C++26: erroneous value + erroneous behavior
int g(bool b)
{
    unsigned char c;     // “c” has an indeterminate/erroneous value
 
    unsigned char d = c; // no undefined/erroneous behavior,
                         // but “d” has an indeterminate/erroneous value
 
    assert(c == d);      // holds, but both integral promotions have
                         // undefined/erroneous behavior
 
    int e = d;           // undefined/erroneous behavior
    return b ? d : 0;    // undefined/erroneous behavior if “b” is true
}
 
// Same as case 2
void h()
{
    int d1, d2;  // “d1” and “d2” have indeterminate/erroneous values
    int e1 = d1; // undefined/erroneous behavior
    int e2 = d1; // undefined/erroneous behavior
 
    assert(e1 == e2); // holds
    assert(e1 == d1); // holds, undefined/erroneous behavior
    assert(e2 == d1); // holds, undefined/erroneous behavior
 
    // no undefined/erroneous behavior,
    // but “d2” has an indeterminate/erroneous value
    std::memcpy(&d2, &d1, sizeof(int));
 
    assert(e1 == d2); // holds, undefined/erroneous behavior
    assert(e2 == d2); // holds, undefined/erroneous behavior
}

[edit] Notes

References and const scalar objects cannot be default-initialized.

[edit] Example

#include <string>
 
struct T1 { int mem; };
 
struct T2
{
    int mem;
    T2() {} // “mem” is not in the initializer list
};
 
int n; // static non-class, a two-phase initialization is done:
       // 1) zero-initialization initializes n to zero
       // 2) default-initialization does nothing, leaving n being zero
 
int main()
{
    [[maybe_unused]]
    int n;            // non-class, the value is indeterminate
    std::string s;    // class, calls default constructor, the value is ""
    std::string a[2]; // array, default-initializes the elements, the value is {"", ""}
//  int& r;           // Error: a reference
//  const int n;      // Error: a const non-class
//  const T1 t1;      // Error: const class with implicit default constructor
    [[maybe_unused]]
    T1 t1;            // class, calls implicit default constructor
    const T2 t2;      // const class, calls the user-provided default constructor
                      // t2.mem is default-initialized
}

[edit] Defect reports

The following behavior-changing defect reports were applied retroactively to previously published C++ standards.

[edit] See also

• converting constructor
• default constructor
• explicit
• initialization
  • aggregate initialization
  • constant initialization
  • copy-initialization
  • direct-initialization
  • list-initialization
  • reference initialization
  • value-initialization
  • zero-initialization
• new