Default-initialization
This is the initialization performed when an object is constructed with no initializer.
Syntax
T object ;
|
(1) | ||||||||
new T
|
(2) | ||||||||
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).
Only (possibly cv-qualified) non-POD class types (or arrays thereof) with automatic storage duration were considered to be default-initialized when no initializer is used. Scalars and POD types with dynamic storage duration were considered to be not initialized (since C++11, this situation was reclassified as a form of default-initialization). |
(until C++11) |
|
(until C++11) |
|
(since C++11) |
each potentially constructed base class of T
is const-default-constructible.
Indeterminate and erroneous values
When storage for an object with automatic or dynamic storage duration is obtained, the object has an indeterminate value. If no initialization is performed for an object, that object retains an indeterminate value until that value is replaced. |
(until C++26) |
When storage for an object with automatic or dynamic storage duration is obtained, the bytes comprising the storage for the object have the following initial value:
If no initialization is performed for an object (including subobjects), such a byte retains its initial value until that value is replaced.
|
(since C++26) |
If an evaluation produces an indeterminate value, the behavior is undefined.
If an evaluation produces an erroneous value, the behavior is erroneous. |
(since C++26) |
Special cases
The following types are uninitialized-friendly:
(since C++17) |
- 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.
|
(since C++26) |
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.
(since C++17) |
- 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.
Converting an erroneous value of an uninitialized-friendly type produces an erroneous value, the result of the conversion is the value of the converted operand. |
(since C++26) |
// 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 }
Notes
References and const scalar objects cannot be default-initialized.
Feature-test macro | Value | Std | Feature |
---|---|---|---|
__cpp_constexpr |
201907L | (C++20) | Trivial default-initialization and asm-declaration in constexpr functions |
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 }
Defect reports
The following behavior-changing defect reports were applied retroactively to previously published C++ standards.
DR | Applied to | Behavior as published | Correct behavior |
---|---|---|---|
CWG 178 | C++98 | there was no value-initialization; empty initializer invoked default-initialization (though new T() also performs zero-initialization) |
empty initializer invokes value-initialization |
CWG 253 | C++98 | default-initialization of a const object could not call an implicitly declared default constructor |
allowed if all subobjects are initialized |
CWG 616 | C++98 | lvalue to rvalue conversion of any uninitialized object was always UB |
indeterminate unsigned char is allowed |
CWG 1787 | C++98 | read from an indeterminate unsigned char cached in a register was UB |
made well-defined |