This International Standard specifies requirements for implementations of the C++ programming language. The first such requirement is that they implement the language, and so this International Standard also defines C++. Other requirements and relaxations of the first requirement appear at various places within this International Standard.

C++ is a general purpose programming language based on the C programming language as described in ISO/IEC 9899:1999 Programming languages — C (hereinafter referred to as the C standard). In addition to the facilities provided by C, C++ provides additional data types, classes, templates, exceptions, namespaces, operator overloading, function name overloading, references, free store management operators, and additional library facilities.

The following referenced documents are indispensable for the application of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.

• (1.1)

  Ecma International, ECMAScript Language Specification, Standard Ecma-262, third edition, 1999.

• (1.2)

  ISO/IEC 2382 (all parts), Information technology — Vocabulary

• (1.3)

  ISO/IEC 9899:1999, Programming languages — C

• (1.4)

  ISO/IEC 9899:1999/Cor.1:2001(E), Programming languages — C, Technical Corrigendum 1

• (1.5)

  ISO/IEC 9899:1999/Cor.2:2004(E), Programming languages — C, Technical Corrigendum 2

• (1.6)

  ISO/IEC 9899:1999/Cor.3:2007(E), Programming languages — C, Technical Corrigendum 3

• (1.7)

  ISO/IEC 9945:2003, Information Technology — Portable Operating System Interface (POSIX)

• (1.8)

  ISO/IEC 10646-1:1993, Information technology — Universal Multiple-Octet Coded Character Set (UCS) — Part 1: Architecture and Basic Multilingual Plane

• (1.9)

  ISO/IEC TR 19769:2004, Information technology — Programming languages, their environments and system software interfaces — Extensions for the programming language C to support new character data types

The library described in Clause 7 of ISO/IEC 9899:1999 and Clause 7 of ISO/IEC 9899:1999/Cor.1:2001 and Clause 7 of ISO/IEC 9899:1999/Cor.2:2003 is hereinafter called the C standard library.1

The library described in ISO/IEC TR 19769:2004 is hereinafter called the C Unicode TR.

The operating system interface described in ISO/IEC 9945:2003 is hereinafter called POSIX.

The ECMAScript Language Specification described in Standard Ecma-262 is hereinafter called ECMA-262.

For the purposes of this document, the following definitions apply.

[definitions] defines additional terms that are used only in Clauses [library] through [thread] and Annex [depr].

Terms that are used only in a small portion of this International Standard are defined where they are used and italicized where they are defined.

The set of diagnosable rules consists of all syntactic and semantic rules in this International Standard except for those rules containing an explicit notation that “no diagnostic is required” or which are described as resulting in “undefined behavior.”

Although this International Standard states only requirements on C++ implementations, those requirements are often easier to understand if they are phrased as requirements on programs, parts of programs, or execution of programs. Such requirements have the following meaning:

• (2.1)

  If a program contains no violations of the rules in this International Standard, a conforming implementation shall, within its resource limits, accept and correctly execute2 that program.

• (2.2)

  If a program contains a violation of any diagnosable rule or an occurrence of a construct described in this Standard as “conditionally-supported” when the implementation does not support that construct, a conforming implementation shall issue at least one diagnostic message.

• (2.3)

  If a program contains a violation of a rule for which no diagnostic is required, this International Standard places no requirement on implementations with respect to that program.

For classes and class templates, the library Clauses specify partial definitions. Private members (Clause [class.access]) are not specified, but each implementation shall supply them to complete the definitions according to the description in the library Clauses.

For functions, function templates, objects, and values, the library Clauses specify declarations. Implementations shall supply definitions consistent with the descriptions in the library Clauses.

The names defined in the library have namespace scope ([basic.namespace]). A C++ translation unit ([lex.phases]) obtains access to these names by including the appropriate standard library header ([cpp.include]).

The templates, classes, functions, and objects in the library have external linkage ([basic.link]). The implementation provides definitions for standard library entities, as necessary, while combining translation units to form a complete C++ program ([lex.phases]).

Two kinds of implementations are defined: a hosted implementation and a freestanding implementation. For a hosted implementation, this International Standard defines the set of available libraries. A freestanding implementation is one in which execution may take place without the benefit of an operating system, and has an implementation-defined set of libraries that includes certain language-support libraries ([compliance]).

A conforming implementation may have extensions (including additional library functions), provided they do not alter the behavior of any well-formed program. Implementations are required to diagnose programs that use such extensions that are ill-formed according to this International Standard. Having done so, however, they can compile and execute such programs.

Each implementation shall include documentation that identifies all conditionally-supported constructs that it does not support and defines all locale-specific characteristics.3

Clauses [lex] through [cpp] describe the C++ programming language. That description includes detailed syntactic specifications in a form described in [syntax]. For convenience, Annex [gram] repeats all such syntactic specifications.

Clauses [language.support] through [thread] and Annex [depr] (the library clauses) describe the Standard C++ library. That description includes detailed descriptions of the templates, classes, functions, constants, and macros that constitute the library, in a form described in Clause [library].

Annex [implimits] recommends lower bounds on the capacity of conforming implementations.

Annex [diff] summarizes the evolution of C++ since its first published description, and explains in detail the differences between C++ and C. Certain features of C++ exist solely for compatibility purposes; Annex [depr] describes those features.

Throughout this International Standard, each example is introduced by “[ Example:” and terminated by “ — end example ]”. Each note is introduced by “[ Note:” and terminated by “ — end note ]”. Examples and notes may be nested.

In the syntax notation used in this International Standard, syntactic categories are indicated by italic type, and literal words and characters in constant width type. Alternatives are listed on separate lines except in a few cases where a long set of alternatives is marked by the phrase “one of.” If the text of an alternative is too long to fit on a line, the text is continued on subsequent lines indented from the first one. An optional terminal or non-terminal symbol is indicated by the subscript “_opt”, so

{ expressionopt }

indicates an optional expression enclosed in braces.

Names for syntactic categories have generally been chosen according to the following rules:

• (2.1)

  X-name is a use of an identifier in a context that determines its meaning (e.g., class-name, typedef-name).

• (2.2)

  X-id is an identifier with no context-dependent meaning (e.g., qualified-id).

• (2.3)

  X-seq is one or more X's without intervening delimiters (e.g., declaration-seq is a sequence of declarations).

• (2.4)

  X-list is one or more X's separated by intervening commas (e.g., expression-list is a sequence of expressions separated by commas).

The fundamental storage unit in the C++ memory model is the byte. A byte is at least large enough to contain any member of the basic execution character set ([lex.charset]) and the eight-bit code units of the Unicode UTF-8 encoding form and is composed of a contiguous sequence of bits, the number of which is implementation-defined. The least significant bit is called the low-order bit; the most significant bit is called the high-order bit. The memory available to a C++ program consists of one or more sequences of contiguous bytes. Every byte has a unique address.

[ Note: The representation of types is described in [basic.types].  — end note ]

A memory location is either an object of scalar type or a maximal sequence of adjacent bit-fields all having non-zero width. [ Note: Various features of the language, such as references and virtual functions, might involve additional memory locations that are not accessible to programs but are managed by the implementation.  — end note ] Two or more threads of execution ([intro.multithread]) can update and access separate memory locations without interfering with each other.

[ Note: Thus a bit-field and an adjacent non-bit-field are in separate memory locations, and therefore can be concurrently updated by two threads of execution without interference. The same applies to two bit-fields, if one is declared inside a nested struct declaration and the other is not, or if the two are separated by a zero-length bit-field declaration, or if they are separated by a non-bit-field declaration. It is not safe to concurrently update two bit-fields in the same struct if all fields between them are also bit-fields of non-zero width.  — end note ]

[ Example: A structure declared as

struct {
  char a;
  int b:5,
  c:11,
  :0,
  d:8;
  struct {int ee:8;} e;
}

contains four separate memory locations: The field a and bit-fields d and e.ee are each separate memory locations, and can be modified concurrently without interfering with each other. The bit-fields b and c together constitute the fourth memory location. The bit-fields b and c cannot be concurrently modified, but b and a, for example, can be.  — end example ]

The constructs in a C++ program create, destroy, refer to, access, and manipulate objects. An object is a region of storage. [ Note: A function is not an object, regardless of whether or not it occupies storage in the way that objects do.  — end note ] An object is created by a definition ([basic.def]), by a new-expression ([expr.new]) or by the implementation ([class.temporary]) when needed. The properties of an object are determined when the object is created. An object can have a name (Clause [basic]). An object has a storage duration ([basic.stc]) which influences its lifetime ([basic.life]). An object has a type ([basic.types]). The term object type refers to the type with which the object is created. Some objects are polymorphic ([class.virtual]); the implementation generates information associated with each such object that makes it possible to determine that object's type during program execution. For other objects, the interpretation of the values found therein is determined by the type of the expressions (Clause [expr]) used to access them.

Objects can contain other objects, called subobjects. A subobject can be a member subobject ([class.mem]), a base class subobject (Clause [class.derived]), or an array element. An object that is not a subobject of any other object is called a complete object.

For every object x, there is some object called the complete object of x, determined as follows:

• (3.1)

  If x is a complete object, then x is the complete object of x.

• (3.2)

  Otherwise, the complete object of x is the complete object of the (unique) object that contains x.

If a complete object, a data member ([class.mem]), or an array element is of class type, its type is considered the most derived class, to distinguish it from the class type of any base class subobject; an object of a most derived class type or of a non-class type is called a most derived object.

Unless it is a bit-field ([class.bit]), a most derived object shall have a non-zero size and shall occupy one or more bytes of storage. Base class subobjects may have zero size. An object of trivially copyable or standard-layout type ([basic.types]) shall occupy contiguous bytes of storage.

Unless an object is a bit-field or a base class subobject of zero size, the address of that object is the address of the first byte it occupies. Two objects that are not bit-fields may have the same address if one is a subobject of the other, or if at least one is a base class subobject of zero size and they are of different types; otherwise, they shall have distinct addresses.4

[ Example:

static const char test1 = 'x';
static const char test2 = 'x';
const bool b = &test1 != &test2;      // always true

 — end example ]

[ Note: C++ provides a variety of fundamental types and several ways of composing new types from existing types ([basic.types]).  — end note ]

The semantic descriptions in this International Standard define a parameterized nondeterministic abstract machine. This International Standard places no requirement on the structure of conforming implementations. In particular, they need not copy or emulate the structure of the abstract machine. Rather, conforming implementations are required to emulate (only) the observable behavior of the abstract machine as explained below.5

Certain aspects and operations of the abstract machine are described in this International Standard as implementation-defined (for example, sizeof(int)). These constitute the parameters of the abstract machine. Each implementation shall include documentation describing its characteristics and behavior in these respects.6 Such documentation shall define the instance of the abstract machine that corresponds to that implementation (referred to as the “corresponding instance” below).

Certain other aspects and operations of the abstract machine are described in this International Standard as unspecified (for example, evaluation of expressions in a new-initializer if the allocation function fails to allocate memory ([expr.new])). Where possible, this International Standard defines a set of allowable behaviors. These define the nondeterministic aspects of the abstract machine. An instance of the abstract machine can thus have more than one possible execution for a given program and a given input.

Certain other operations are described in this International Standard as undefined (for example, the effect of attempting to modify a const object). [ Note: This International Standard imposes no requirements on the behavior of programs that contain undefined behavior.  — end note ]

A conforming implementation executing a well-formed program shall produce the same observable behavior as one of the possible executions of the corresponding instance of the abstract machine with the same program and the same input. However, if any such execution contains an undefined operation, this International Standard places no requirement on the implementation executing that program with that input (not even with regard to operations preceding the first undefined operation).

If a signal handler is executed as a result of a call to the raise function, then the execution of the handler is sequenced after the invocation of the raise function and before its return. [ Note: When a signal is received for another reason, the execution of the signal handler is usually unsequenced with respect to the rest of the program.  — end note ]

An instance of each object with automatic storage duration ([basic.stc.auto]) is associated with each entry into its block. Such an object exists and retains its last-stored value during the execution of the block and while the block is suspended (by a call of a function or receipt of a signal).

The least requirements on a conforming implementation are:

• (8.1)

  Access to volatile objects are evaluated strictly according to the rules of the abstract machine.

• (8.2)

  At program termination, all data written into files shall be identical to one of the possible results that execution of the program according to the abstract semantics would have produced.

• (8.3)

  The input and output dynamics of interactive devices shall take place in such a fashion that prompting output is actually delivered before a program waits for input. What constitutes an interactive device is implementation-defined.

These collectively are referred to as the observable behavior of the program. [ Note: More stringent correspondences between abstract and actual semantics may be defined by each implementation.  — end note ]

[ Note: Operators can be regrouped according to the usual mathematical rules only where the operators really are associative or commutative.7 For example, in the following fragment

int a, b;
/*...*/
a = a + 32760 + b + 5;

the expression statement behaves exactly the same as

a = (((a + 32760) + b) + 5);

due to the associativity and precedence of these operators. Thus, the result of the sum (a + 32760) is next added to b, and that result is then added to 5 which results in the value assigned to a. On a machine in which overflows produce an exception and in which the range of values representable by an int is [-32768,+32767], the implementation cannot rewrite this expression as

a = ((a + b) + 32765);

since if the values for a and b were, respectively, -32754 and -15, the sum a + b would produce an exception while the original expression would not; nor can the expression be rewritten either as

a = ((a + 32765) + b);

or

a = (a + (b + 32765));

since the values for a and b might have been, respectively, 4 and -8 or -17 and 12. However on a machine in which overflows do not produce an exception and in which the results of overflows are reversible, the above expression statement can be rewritten by the implementation in any of the above ways because the same result will occur.  — end note ]

A full-expression is an expression that is not a subexpression of another expression. [ Note: in some contexts, such as unevaluated operands, a syntactic subexpression is considered a full-expression (Clause [expr]).  — end note ] If a language construct is defined to produce an implicit call of a function, a use of the language construct is considered to be an expression for the purposes of this definition. A call to a destructor generated at the end of the lifetime of an object other than a temporary object is an implicit full-expression. Conversions applied to the result of an expression in order to satisfy the requirements of the language construct in which the expression appears are also considered to be part of the full-expression.

[ Example:

struct S {
  S(int i): I(i) { }
  int& v() { return I; }
private:
  int I;
};

 S s1(1);           // full-expression is call of S::S(int)
 S s2 = 2;          // full-expression is call of S::S(int)

void f() {
  if (S(3).v())     // full-expression includes lvalue-to-rvalue and
                    // int to bool conversions, performed before
                    // temporary is deleted at end of full-expression
  { }
}

 — end example ]

[ Note: The evaluation of a full-expression can include the evaluation of subexpressions that are not lexically part of the full-expression. For example, subexpressions involved in evaluating default arguments ([dcl.fct.default]) are considered to be created in the expression that calls the function, not the expression that defines the default argument.  — end note ]

Accessing an object designated by a volatile glvalue ([basic.lval]), modifying an object, calling a library I/O function, or calling a function that does any of those operations are all side effects, which are changes in the state of the execution environment. Evaluation of an expression (or a sub-expression) in general includes both value computations (including determining the identity of an object for glvalue evaluation and fetching a value previously assigned to an object for prvalue evaluation) and initiation of side effects. When a call to a library I/O function returns or an access to a volatile object is evaluated the side effect is considered complete, even though some external actions implied by the call (such as the I/O itself) or by the volatile access may not have completed yet.

Sequenced before is an asymmetric, transitive, pair-wise relation between evaluations executed by a single thread ([intro.multithread]), which induces a partial order among those evaluations. Given any two evaluations A and B, if A is sequenced before B, then the execution of A shall precede the execution of B. If A is not sequenced before B and B is not sequenced before A, then A and B are unsequenced. [ Note: The execution of unsequenced evaluations can overlap.  — end note ] Evaluations A and B are indeterminately sequenced when either A is sequenced before B or B is sequenced before A, but it is unspecified which. [ Note: Indeterminately sequenced evaluations cannot overlap, but either could be executed first.  — end note ]

Every value computation and side effect associated with a full-expression is sequenced before every value computation and side effect associated with the next full-expression to be evaluated.8

Except where noted, evaluations of operands of individual operators and of subexpressions of individual expressions are unsequenced. [ Note: In an expression that is evaluated more than once during the execution of a program, unsequenced and indeterminately sequenced evaluations of its subexpressions need not be performed consistently in different evaluations.  — end note ] The value computations of the operands of an operator are sequenced before the value computation of the result of the operator. If a side effect on a scalar object is unsequenced relative to either another side effect on the same scalar object or a value computation using the value of the same scalar object, and they are not potentially concurrent ([intro.multithread]), the behavior is undefined. [ Note: The next section imposes similar, but more complex restrictions on potentially concurrent computations.  — end note ]

[ Example:

void f(int, int);
void g(int i, int* v) {
  i = v[i++];         // the behavior is undefined
  i = 7, i++, i++;    // i becomes 9

  i = i++ + 1;        // the behavior is undefined
  i = i + 1;          // the value of i is incremented

  f(i = -1, i = -1);  // the behavior is undefined
}

 — end example ]

When calling a function (whether or not the function is inline), every value computation and side effect associated with any argument expression, or with the postfix expression designating the called function, is sequenced before execution of every expression or statement in the body of the called function. [ Note: Value computations and side effects associated with different argument expressions are unsequenced.  — end note ] Every evaluation in the calling function (including other function calls) that is not otherwise specifically sequenced before or after the execution of the body of the called function is indeterminately sequenced with respect to the execution of the called function.9 Several contexts in C++ cause evaluation of a function call, even though no corresponding function call syntax appears in the translation unit. [ Example: Evaluation of a new expression invokes one or more allocation and constructor functions; see [expr.new]. For another example, invocation of a conversion function ([class.conv.fct]) can arise in contexts in which no function call syntax appears.  — end example ] The sequencing constraints on the execution of the called function (as described above) are features of the function calls as evaluated, whatever the syntax of the expression that calls the function might be.

A thread of execution (also known as a thread) is a single flow of control within a program, including the initial invocation of a specific top-level function, and recursively including every function invocation subsequently executed by the thread. [ Note: When one thread creates another, the initial call to the top-level function of the new thread is executed by the new thread, not by the creating thread.  — end note ] Every thread in a program can potentially access every object and function in a program.10 Under a hosted implementation, a C++ program can have more than one thread running concurrently. The execution of each thread proceeds as defined by the remainder of this standard. The execution of the entire program consists of an execution of all of its threads. [ Note: Usually the execution can be viewed as an interleaving of all its threads. However, some kinds of atomic operations, for example, allow executions inconsistent with a simple interleaving, as described below.  — end note ] Under a freestanding implementation, it is implementation-defined whether a program can have more than one thread of execution.

A signal handler that is executed as a result of a call to the raise function belongs to the same thread of execution as the call to the raise function. Otherwise it is unspecified which thread of execution contains a signal handler invocation.

Implementations should ensure that all unblocked threads eventually make progress. [ Note: Standard library functions may silently block on I/O or locks. Factors in the execution environment, including externally-imposed thread priorities, may prevent an implementation from making certain guarantees of forward progress.  — end note ]

Executions of atomic functions that are either defined to be lock-free ([atomics.flag]) or indicated as lock-free ([atomics.lockfree]) are lock-free executions.

• (4.1)

  If there is only one unblocked thread, a lock-free execution in that thread shall complete. [ Note: Concurrently executing threads may prevent progress of a lock-free execution. For example, this situation can occur with load-locked store-conditional implementations. This property is sometimes termed obstruction-free.  — end note ]

• (4.2)

  When one or more lock-free executions run concurrently, at least one should complete. [ Note: It is difficult for some implementations to provide absolute guarantees to this effect, since repeated and particularly inopportune interference from other threads may prevent forward progress, e.g., by repeatedly stealing a cache line for unrelated purposes between load-locked and store-conditional instructions. Implementations should ensure that such effects cannot indefinitely delay progress under expected operating conditions, and that such anomalies can therefore safely be ignored by programmers. Outside this International Standard, this property is sometimes termed lock-free.  — end note ]

The value of an object visible to a thread T at a particular point is the initial value of the object, a value assigned to the object by T, or a value assigned to the object by another thread, according to the rules below. [ Note: In some cases, there may instead be undefined behavior. Much of this section is motivated by the desire to support atomic operations with explicit and detailed visibility constraints. However, it also implicitly supports a simpler view for more restricted programs.  — end note ]

Two expression evaluations conflict if one of them modifies a memory location ([intro.memory]) and the other one accesses or modifies the same memory location.

The library defines a number of atomic operations (Clause [atomics]) and operations on mutexes (Clause [thread]) that are specially identified as synchronization operations. These operations play a special role in making assignments in one thread visible to another. A synchronization operation on one or more memory locations is either a consume operation, an acquire operation, a release operation, or both an acquire and release operation. A synchronization operation without an associated memory location is a fence and can be either an acquire fence, a release fence, or both an acquire and release fence. In addition, there are relaxed atomic operations, which are not synchronization operations, and atomic read-modify-write operations, which have special characteristics. [ Note: For example, a call that acquires a mutex will perform an acquire operation on the locations comprising the mutex. Correspondingly, a call that releases the same mutex will perform a release operation on those same locations. Informally, performing a release operation on A forces prior side effects on other memory locations to become visible to other threads that later perform a consume or an acquire operation on A. “Relaxed” atomic operations are not synchronization operations even though, like synchronization operations, they cannot contribute to data races.  — end note ]

All modifications to a particular atomic object M occur in some particular total order, called the modification order of M. If A and B are modifications of an atomic object M and A happens before (as defined below) B, then A shall precede B in the modification order of M, which is defined below. [ Note: This states that the modification orders must respect the “happens before” relationship.  — end note ] [ Note: There is a separate order for each atomic object. There is no requirement that these can be combined into a single total order for all objects. In general this will be impossible since different threads may observe modifications to different objects in inconsistent orders.  — end note ]

A release sequence headed by a release operation A on an atomic object M is a maximal contiguous sub-sequence of side effects in the modification order of M, where the first operation is A, and every subsequent operation

• (9.1)

  is performed by the same thread that performed A, or

• (9.2)

  is an atomic read-modify-write operation.

Certain library calls synchronize with other library calls performed by another thread. For example, an atomic store-release synchronizes with a load-acquire that takes its value from the store ([atomics.order]). [ Note: Except in the specified cases, reading a later value does not necessarily ensure visibility as described below. Such a requirement would sometimes interfere with efficient implementation.  — end note ] [ Note: The specifications of the synchronization operations define when one reads the value written by another. For atomic objects, the definition is clear. All operations on a given mutex occur in a single total order. Each mutex acquisition “reads the value written” by the last mutex release.  — end note ]

An evaluation A carries a dependency to an evaluation B if

• (11.1)

  the value of A is used as an operand of B, unless:

  • (11.1.1)

    B is an invocation of any specialization of std::kill_dependency ([atomics.order]), or

  • (11.1.2)

    A is the left operand of a built-in logical AND (&&, see [expr.log.and]) or logical OR (||, see [expr.log.or]) operator, or

  • (11.1.3)

    A is the left operand of a conditional (?:, see [expr.cond]) operator, or

  • (11.1.4)

    A is the left operand of the built-in comma (,) operator ([expr.comma]);

  or

• (11.2)

  A writes a scalar object or bit-field M, B reads the value written by A from M, and A is sequenced before B, or

• (11.3)

  for some evaluation X, A carries a dependency to X, and X carries a dependency to B.

[ Note: “Carries a dependency to” is a subset of “is sequenced before”, and is similarly strictly intra-thread.  — end note ]

An evaluation A is dependency-ordered before an evaluation B if

• (12.1)

  A performs a release operation on an atomic object M, and, in another thread, B performs a consume operation on M and reads a value written by any side effect in the release sequence headed by A, or

• (12.2)

  for some evaluation X, A is dependency-ordered before X and X carries a dependency to B.

[ Note: The relation “is dependency-ordered before” is analogous to “synchronizes with”, but uses release/consume in place of release/acquire.  — end note ]

An evaluation A inter-thread happens before an evaluation B if

• (13.1)

  A synchronizes with B, or

• (13.2)

  A is dependency-ordered before B, or

• (13.3)

  for some evaluation X

  • (13.3.1)

    A synchronizes with X and X is sequenced before B, or

  • (13.3.2)

    A is sequenced before X and X inter-thread happens before B, or

  • (13.3.3)

    A inter-thread happens before X and X inter-thread happens before B.

[ Note: The “inter-thread happens before” relation describes arbitrary concatenations of “sequenced before”, “synchronizes with” and “dependency-ordered before” relationships, with two exceptions. The first exception is that a concatenation is not permitted to end with “dependency-ordered before” followed by “sequenced before”. The reason for this limitation is that a consume operation participating in a “dependency-ordered before” relationship provides ordering only with respect to operations to which this consume operation actually carries a dependency. The reason that this limitation applies only to the end of such a concatenation is that any subsequent release operation will provide the required ordering for a prior consume operation. The second exception is that a concatenation is not permitted to consist entirely of “sequenced before”. The reasons for this limitation are (1) to permit “inter-thread happens before” to be transitively closed and (2) the “happens before” relation, defined below, provides for relationships consisting entirely of “sequenced before”.  — end note ]

An evaluation A happens before an evaluation B if:

• (14.1)

  A is sequenced before B, or

• (14.2)

  A inter-thread happens before B.

The implementation shall ensure that no program execution demonstrates a cycle in the “happens before” relation. [ Note: This cycle would otherwise be possible only through the use of consume operations.  — end note ]

A visible side effect A on a scalar object or bit-field M with respect to a value computation B of M satisfies the conditions:

• (15.1)

  A happens before B and

• (15.2)

  there is no other side effect X to M such that A happens before X and X happens before B.

The value of a non-atomic scalar object or bit-field M, as determined by evaluation B, shall be the value stored by the visible side effect A. [ Note: If there is ambiguity about which side effect to a non-atomic object or bit-field is visible, then the behavior is either unspecified or undefined.  — end note ] [ Note: This states that operations on ordinary objects are not visibly reordered. This is not actually detectable without data races, but it is necessary to ensure that data races, as defined below, and with suitable restrictions on the use of atomics, correspond to data races in a simple interleaved (sequentially consistent) execution.  — end note ]

The value of an atomic object M, as determined by evaluation B, shall be the value stored by some side effect A that modifies M, where B does not happen before A. [ Note: The set of such side effects is also restricted by the rest of the rules described here, and in particular, by the coherence requirements below.  — end note ]

If an operation A that modifies an atomic object M happens before an operation B that modifies M, then A shall be earlier than B in the modification order of M. [ Note: This requirement is known as write-write coherence.  — end note ]

If a value computation A of an atomic object M happens before a value computation B of M, and A takes its value from a side effect X on M, then the value computed by B shall either be the value stored by X or the value stored by a side effect Y on M, where Y follows X in the modification order of M. [ Note: This requirement is known as read-read coherence.  — end note ]

If a value computation A of an atomic object M happens before an operation B that modifies M, then A shall take its value from a side effect X on M, where X precedes B in the modification order of M. [ Note: This requirement is known as read-write coherence.  — end note ]

If a side effect X on an atomic object M happens before a value computation B of M, then the evaluation B shall take its value from X or from a side effect Y that follows X in the modification order of M. [ Note: This requirement is known as write-read coherence.  — end note ]

[ Note: The four preceding coherence requirements effectively disallow compiler reordering of atomic operations to a single object, even if both operations are relaxed loads. This effectively makes the cache coherence guarantee provided by most hardware available to C++ atomic operations.  — end note ]

[ Note: The value observed by a load of an atomic depends on the “happens before” relation, which depends on the values observed by loads of atomics. The intended reading is that there must exist an association of atomic loads with modifications they observe that, together with suitably chosen modification orders and the “happens before” relation derived as described above, satisfy the resulting constraints as imposed here.  — end note ]

Two actions are potentially concurrent if

• (23.1)

  they are performed by different threads, or

• (23.2)

  they are unsequenced, and at least one is performed by a signal handler.

The execution of a program contains a data race if it contains two potentially concurrent conflicting actions, at least one of which is not atomic, and neither happens before the other, except for the special case for signal handlers described below. Any such data race results in undefined behavior. [ Note: It can be shown that programs that correctly use mutexes and memory_order_seq_cst operations to prevent all data races and use no other synchronization operations behave as if the operations executed by their constituent threads were simply interleaved, with each value computation of an object being taken from the last side effect on that object in that interleaving. This is normally referred to as “sequential consistency”. However, this applies only to data-race-free programs, and data-race-free programs cannot observe most program transformations that do not change single-threaded program semantics. In fact, most single-threaded program transformations continue to be allowed, since any program that behaves differently as a result must perform an undefined operation.  — end note ]

Two accesses to the same object of type volatile sig_atomic_t do not result in a data race if both occur in the same thread, even if one or more occurs in a signal handler. For each signal handler invocation, evaluations performed by the thread invoking a signal handler can be divided into two groups A and B, such that no evaluations in B happen before evaluations in A, and the evaluations of such volatile sig_atomic_t objects take values as though all evaluations in A happened before the execution of the signal handler and the execution of the signal handler happened before all evaluations in B.

[ Note: Compiler transformations that introduce assignments to a potentially shared memory location that would not be modified by the abstract machine are generally precluded by this standard, since such an assignment might overwrite another assignment by a different thread in cases in which an abstract machine execution would not have encountered a data race. This includes implementations of data member assignment that overwrite adjacent members in separate memory locations. Reordering of atomic loads in cases in which the atomics in question may alias is also generally precluded, since this may violate the coherence rules.  — end note ]

[ Note: Transformations that introduce a speculative read of a potentially shared memory location may not preserve the semantics of the C++ program as defined in this standard, since they potentially introduce a data race. However, they are typically valid in the context of an optimizing compiler that targets a specific machine with well-defined semantics for data races. They would be invalid for a hypothetical machine that is not tolerant of races or provides hardware race detection.  — end note ]

The implementation may assume that any thread will eventually do one of the following:

• (27.1)

  terminate,

• (27.2)

  make a call to a library I/O function,

• (27.3)

  access or modify a volatile object, or

• (27.4)

  perform a synchronization operation or an atomic operation.

[ Note: This is intended to allow compiler transformations such as removal of empty loops, even when termination cannot be proven.  — end note ]

An implementation should ensure that the last value (in modification order) assigned by an atomic or synchronization operation will become visible to all other threads in a finite period of time.

The C++ programming language as described in this International Standard is based on the language as described in Chapter R (Reference Manual) of Stroustrup: The C++ Programming Language (second edition, Addison-Wesley Publishing Company, ISBN 0-201-53992-6, copyright © 1991 AT&T). That, in turn, is based on the C programming language as described in Appendix A of Kernighan and Ritchie: The C Programming Language (Prentice-Hall, 1978, ISBN 0-13-110163-3, copyright © 1978 AT&T).

Portions of the library Clauses of this International Standard are based on work by P.J. Plauger, which was published as The Draft Standard C++ Library (Prentice-Hall, ISBN 0-13-117003-1, copyright © 1995 P.J. Plauger).

POSIX® is a registered trademark of the Institute of Electrical and Electronic Engineers, Inc.

All rights in these originals are reserved.

The text of the program is kept in units called source files in this International Standard. A source file together with all the headers ([headers]) and source files included ([cpp.include]) via the preprocessing directive #include, less any source lines skipped by any of the conditional inclusion ([cpp.cond]) preprocessing directives, is called a translation unit. [ Note: A C++ program need not all be translated at the same time.  — end note ]

[ Note: Previously translated translation units and instantiation units can be preserved individually or in libraries. The separate translation units of a program communicate ([basic.link]) by (for example) calls to functions whose identifiers have external linkage, manipulation of objects whose identifiers have external linkage, or manipulation of data files. Translation units can be separately translated and then later linked to produce an executable program ([basic.link]).  — end note ]

The precedence among the syntax rules of translation is specified by the following phases.11

1. 1.

   Physical source file characters are mapped, in an implementation-defined manner, to the basic source character set (introducing new-line characters for end-of-line indicators) if necessary. The set of physical source file characters accepted is implementation-defined. Trigraph sequences ([lex.trigraph]) are replaced by corresponding single-character internal representations. Any source file character not in the basic source character set ([lex.charset]) is replaced by the universal-character-name that designates that character. (An implementation may use any internal encoding, so long as an actual extended character encountered in the source file, and the same extended character expressed in the source file as a universal-character-name (i.e., using the \uXXXX notation), are handled equivalently except where this replacement is reverted in a raw string literal.)

2. 2.

   Each instance of a backslash character (\) immediately followed by a new-line character is deleted, splicing physical source lines to form logical source lines. Only the last backslash on any physical source line shall be eligible for being part of such a splice. Except for splices reverted in a raw string literal, if a splice results in a character sequence that matches the syntax of a universal-character-name, the behavior is undefined. A source file that is not empty and that does not end in a new-line character, or that ends in a new-line character immediately preceded by a backslash character before any such splicing takes place, shall be processed as if an additional new-line character were appended to the file.

3. 3.

   The source file is decomposed into preprocessing tokens ([lex.pptoken]) and sequences of white-space characters (including comments). A source file shall not end in a partial preprocessing token or in a partial comment.12 Each comment is replaced by one space character. New-line characters are retained. Whether each nonempty sequence of white-space characters other than new-line is retained or replaced by one space character is unspecified. The process of dividing a source file's characters into preprocessing tokens is context-dependent. [ Example: see the handling of < within a #include preprocessing directive.  — end example ]

4. 4.

   Preprocessing directives are executed, macro invocations are expanded, and _Pragma unary operator expressions are executed. If a character sequence that matches the syntax of a universal-character-name is produced by token concatenation ([cpp.concat]), the behavior is undefined. A #include preprocessing directive causes the named header or source file to be processed from phase 1 through phase 4, recursively. All preprocessing directives are then deleted.

5. 5.

   Each source character set member in a character literal or a string literal, as well as each escape sequence and universal-character-name in a character literal or a non-raw string literal, is converted to the corresponding member of the execution character set ([lex.ccon], [lex.string]); if there is no corresponding member, it is converted to an implementation-defined member other than the null (wide) character.13

6. 6.

   Adjacent string literal tokens are concatenated.

7. 7.

   White-space characters separating tokens are no longer significant. Each preprocessing token is converted into a token. ([lex.token]). The resulting tokens are syntactically and semantically analyzed and translated as a translation unit. [ Note: The process of analyzing and translating the tokens may occasionally result in one token being replaced by a sequence of other tokens ([temp.names]). — end note ] [ Note: Source files, translation units and translated translation units need not necessarily be stored as files, nor need there be any one-to-one correspondence between these entities and any external representation. The description is conceptual only, and does not specify any particular implementation.  — end note ]

8. 8.

   Translated translation units and instantiation units are combined as follows: [ Note: Some or all of these may be supplied from a library.  — end note ] Each translated translation unit is examined to produce a list of required instantiations. [ Note: This may include instantiations which have been explicitly requested ([temp.explicit]).  — end note ] The definitions of the required templates are located. It is implementation-defined whether the source of the translation units containing these definitions is required to be available. [ Note: An implementation could encode sufficient information into the translated translation unit so as to ensure the source is not required here.  — end note ] All the required instantiations are performed to produce instantiation units. [ Note: These are similar to translated translation units, but contain no references to uninstantiated templates and no template definitions.  — end note ] The program is ill-formed if any instantiation fails.

9. 9.

   All external entity references are resolved. Library components are linked to satisfy external references to entities not defined in the current translation. All such translator output is collected into a program image which contains information needed for execution in its execution environment.

The basic source character set consists of 96 characters: the space character, the control characters representing horizontal tab, vertical tab, form feed, and new-line, plus the following 91 graphical characters:14

a b c d e f g h i j k l m n o p q r s t u v w x y z

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

0 1 2 3 4 5 6 7 8 9

_ { } [ ] # ( ) < > % : ; . ? * + - / ^ & | ~ ! = , \ " '

The universal-character-name construct provides a way to name other characters.

hex-quad:
    hexadecimal-digit hexadecimal-digit hexadecimal-digit hexadecimal-digit

universal-character-name:
    \u hex-quad
    \U hex-quad hex-quad

The character designated by the universal-character-name \UNNNNNNNN is that character whose character short name in ISO/IEC 10646 is NNNNNNNN; the character designated by the universal-character-name \uNNNN is that character whose character short name in ISO/IEC 10646 is 0000NNNN. If the hexadecimal value for a universal-character-name corresponds to a surrogate code point (in the range 0xD800–0xDFFF, inclusive), the program is ill-formed. Additionally, if the hexadecimal value for a universal-character-name outside the c-char-sequence, s-char-sequence, or r-char-sequence of a character or string literal corresponds to a control character (in either of the ranges 0x00–0x1F or 0x7F–0x9F, both inclusive) or to a character in the basic source character set, the program is ill-formed.15

The basic execution character set and the basic execution wide-character set shall each contain all the members of the basic source character set, plus control characters representing alert, backspace, and carriage return, plus a null character (respectively, null wide character), whose representation has all zero bits. For each basic execution character set, the values of the members shall be non-negative and distinct from one another. In both the source and execution basic character sets, the value of each character after 0 in the above list of decimal digits shall be one greater than the value of the previous. The execution character set and the execution wide-character set are implementation-defined supersets of the basic execution character set and the basic execution wide-character set, respectively. The values of the members of the execution character sets and the sets of additional members are locale-specific.

Before any other processing takes place, each occurrence of one of the following sequences of three characters (“trigraph sequences”) is replaced by the single character indicated in Table [tab:trigraph.sequences].

[ Example:

??=define arraycheck(a,b) a??(b??) ??!??! b??(a??)

becomes

#define arraycheck(a,b) a[b] || b[a]

 — end example ]

No other trigraph sequence exists. Each ? that does not begin one of the trigraphs listed above is not changed.

Each preprocessing token that is converted to a token ([lex.token]) shall have the lexical form of a keyword, an identifier, a literal, an operator, or a punctuator.

A preprocessing token is the minimal lexical element of the language in translation phases 3 through 6. The categories of preprocessing token are: header names, identifiers, preprocessing numbers, character literals (including user-defined character literals), string literals (including user-defined string literals), preprocessing operators and punctuators, and single non-white-space characters that do not lexically match the other preprocessing token categories. If a ' or a " character matches the last category, the behavior is undefined. Preprocessing tokens can be separated by white space; this consists of comments ([lex.comment]), or white-space characters (space, horizontal tab, new-line, vertical tab, and form-feed), or both. As described in Clause [cpp], in certain circumstances during translation phase 4, white space (or the absence thereof) serves as more than preprocessing token separation. White space can appear within a preprocessing token only as part of a header name or between the quotation characters in a character literal or string literal.

If the input stream has been parsed into preprocessing tokens up to a given character:

• (3.1)

  If the next character begins a sequence of characters that could be the prefix and initial double quote of a raw string literal, such as R", the next preprocessing token shall be a raw string literal. Between the initial and final double quote characters of the raw string, any transformations performed in phases 1 and 2 (trigraphs, universal-character-names, and line splicing) are reverted; this reversion shall apply before any d-char, r-char, or delimiting parenthesis is identified. The raw string literal is defined as the shortest sequence of characters that matches the raw-string pattern

  encoding-prefixopt R raw-string

• (3.2)

  Otherwise, if the next three characters are <:: and the subsequent character is neither : nor >, the < is treated as a preprocessor token by itself and not as the first character of the alternative token <:.

• (3.3)

  Otherwise, the next preprocessing token is the longest sequence of characters that could constitute a preprocessing token, even if that would cause further lexical analysis to fail.

[ Example:

#define R "x"
const char* s = R"y";           // ill-formed raw string, not "x" "y"

 — end example ]

[ Example: The program fragment 1Ex is parsed as a preprocessing number token (one that is not a valid floating or integer literal token), even though a parse as the pair of preprocessing tokens 1 and Ex might produce a valid expression (for example, if Ex were a macro defined as +1). Similarly, the program fragment 1E1 is parsed as a preprocessing number (one that is a valid floating literal token), whether or not E is a macro name.  — end example ]

[ Example: The program fragment x+++++y is parsed as x ++ ++ + y, which, if x and y have integral types, violates a constraint on increment operators, even though the parse x ++ + ++ y might yield a correct expression.  — end example ]

Alternative token representations are provided for some operators and punctuators.16

In all respects of the language, each alternative token behaves the same, respectively, as its primary token, except for its spelling.17 The set of alternative tokens is defined in Table [tab:alternative.tokens].

There are five kinds of tokens: identifiers, keywords, literals,18 operators, and other separators. Blanks, horizontal and vertical tabs, newlines, formfeeds, and comments (collectively, “white space”), as described below, are ignored except as they serve to separate tokens. [ Note: Some white space is required to separate otherwise adjacent identifiers, keywords, numeric literals, and alternative tokens containing alphabetic characters.  — end note ]

The characters /* start a comment, which terminates with the characters */. These comments do not nest. The characters // start a comment, which terminates immediately before the next new-line character. If there is a form-feed or a vertical-tab character in such a comment, only white-space characters shall appear between it and the new-line that terminates the comment; no diagnostic is required. [ Note: The comment characters //, /*, and */ have no special meaning within a // comment and are treated just like other characters. Similarly, the comment characters // and /* have no special meaning within a /* comment.  — end note ]

Header name preprocessing tokens shall only appear within a #include preprocessing directive ([cpp.include]). The sequences in both forms of header-names are mapped in an implementation-defined manner to headers or to external source file names as specified in [cpp.include].

The appearance of either of the characters ' or \ or of either of the character sequences /* or // in a q-char-sequence or an h-char-sequence is conditionally-supported with implementation-defined semantics, as is the appearance of the character " in an h-char-sequence.19

Preprocessing number tokens lexically include all integer literal tokens ([lex.icon]) and all floating literal tokens ([lex.fcon]).

A preprocessing number does not have a type or a value; it acquires both after a successful conversion to an integer literal token or a floating literal token.

An identifier is an arbitrarily long sequence of letters and digits. Each universal-character-name in an identifier shall designate a character whose encoding in ISO 10646 falls into one of the ranges specified in [charname.allowed]. The initial element shall not be a universal-character-name designating a character whose encoding falls into one of the ranges specified in [charname.disallowed]. Upper- and lower-case letters are different. All characters are significant.20

The identifiers in Table [tab:identifiers.special] have a special meaning when appearing in a certain context. When referred to in the grammar, these identifiers are used explicitly rather than using the identifier grammar production. Unless otherwise specified, any ambiguity as to whether a given identifier has a special meaning is resolved to interpret the token as a regular identifier.

In addition, some identifiers are reserved for use by C++ implementations and standard libraries ([global.names]) and shall not be used otherwise; no diagnostic is required.

The identifiers shown in Table [tab:keywords] are reserved for use as keywords (that is, they are unconditionally treated as keywords in phase 7) except in an attribute-token ([dcl.attr.grammar]) [ Note: The export keyword is unused but is reserved for future use. — end note ]:

Furthermore, the alternative representations shown in Table [tab:alternative.representations] for certain operators and punctuators ([lex.digraph]) are reserved and shall not be used otherwise:

The lexical representation of C++ programs includes a number of preprocessing tokens which are used in the syntax of the preprocessor or are converted into tokens for operators and punctuators:

preprocessing-op-or-punc: one of
	{ 	} 	[ 	] 	# 	## 	( 	)
	<: 	:> 	<% 	%> 	%: 	%:%: 	; 	: 	...
	new 	delete 	? 	:: 	. 	.*
	+ 	- 	* 	/ 	% 	^ 	& 	| 	~
	! 	= 	< 	> 	+= 	-= 	*= 	/= 	%=
	^= 	&= 	|= 	<< 	>> 	>>= 	<<= 	== 	!=
	<= 	>= 	&& 	|| 	++ 	-- 	, 	->* 	->
	and 	and_eq 	bitand 	bitor 	compl 	not 	not_eq
	or 	or_eq 	xor 	xor_eq

Each preprocessing-op-or-punc is converted to a single token in translation phase 7 ([lex.phases]).

A declaration (Clause [dcl.dcl]) may introduce one or more names into a translation unit or redeclare names introduced by previous declarations. If so, the declaration specifies the interpretation and attributes of these names. A declaration may also have effects including:

• (1.1)

  a static assertion (Clause [dcl.dcl]),

• (1.2)

  controlling template instantiation ([temp.explicit]),

• (1.3)

  use of attributes (Clause [dcl.dcl]), and

• (1.4)

  nothing (in the case of an empty-declaration).

A declaration is a definition unless it declares a function without specifying the function's body ([dcl.fct.def]), it contains the extern specifier ([dcl.stc]) or a linkage-specification25 ([dcl.link]) and neither an initializer nor a function-body, it declares a static data member in a class definition ([class.mem], [class.static]), it is a class name declaration ([class.name]), it is an opaque-enum-declaration ([dcl.enum]), it is a template-parameter ([temp.param]), it is a parameter-declaration ([dcl.fct]) in a function declarator that is not the declarator of a function-definition, or it is a typedef declaration ([dcl.typedef]), an alias-declaration ([dcl.typedef]), a using-declaration ([namespace.udecl]), a static_assert-declaration (Clause [dcl.dcl]), an attribute-declaration (Clause [dcl.dcl]), an empty-declaration (Clause [dcl.dcl]), or a using-directive ([namespace.udir]).

[ Example: all but one of the following are definitions:

int a;                          // defines a
extern const int c = 1;         // defines c
int f(int x) { return x+a; }    // defines f and defines x
struct S { int a; int b; };     // defines S, S::a, and S::b
struct X {                      // defines X
  int x;                        // defines non-static data member x
  static int y;                 // declares static data member y
  X(): x(0) { }                 // defines a constructor of X
};
int X::y = 1;                   // defines X::y
enum { up, down };              // defines up and down
namespace N { int d; }          // defines N and N::d
namespace N1 = N;               // defines N1
X anX;                          // defines anX

whereas these are just declarations:

extern int a;                   // declares a
extern const int c;             // declares c
int f(int);                     // declares f
struct S;                       // declares S
typedef int Int;                // declares Int
extern X anotherX;              // declares anotherX
using N::d;                     // declares d

 — end example ]

[ Note: In some circumstances, C++ implementations implicitly define the default constructor ([class.ctor]), copy constructor ([class.copy]), move constructor ([class.copy]), copy assignment operator ([class.copy]), move assignment operator ([class.copy]), or destructor ([class.dtor]) member functions.  — end note ] [ Example: given

#include <string>

struct C {
  std::string s;              // std::string is the standard library class (Clause [strings])
};

int main() {
  C a;
  C b = a;
  b = a;
}

the implementation will implicitly define functions to make the definition of C equivalent to

struct C {
  std::string s;
  C() : s() { }
  C(const C& x): s(x.s) { }
  C(C&& x): s(static_cast<std::string&&>(x.s)) { }
    // : s(std::move(x.s)) { }
  C& operator=(const C& x) { s = x.s; return *this; }
  C& operator=(C&& x) { s = static_cast<std::string&&>(x.s); return *this; }
    // { s = std::move(x.s); return *this; }
  ~C() { }
};

 — end example ]

[ Note: A class name can also be implicitly declared by an elaborated-type-specifier ([dcl.type.elab]).  — end note ]

A program is ill-formed if the definition of any object gives the object an incomplete type ([basic.types]).

No translation unit shall contain more than one definition of any variable, function, class type, enumeration type, or template.

An expression is potentially evaluated unless it is an unevaluated operand (Clause [expr]) or a subexpression thereof. The set of potential results of an expression e is defined as follows:

• (2.1)

  If e is an id-expression ([expr.prim.general]), the set contains only e.

• (2.2)

  If e is a class member access expression ([expr.ref]), the set contains the potential results of the object expression.

• (2.3)

  If e is a pointer-to-member expression ([expr.mptr.oper]) whose second operand is a constant expression, the set contains the potential results of the object expression.

• (2.4)

  If e has the form (e1), the set contains the potential results of e1.

• (2.5)

  If e is a glvalue conditional expression ([expr.cond]), the set is the union of the sets of potential results of the second and third operands.

• (2.6)

  If e is a comma expression ([expr.comma]), the set contains the potential results of the right operand.

• (2.7)

  Otherwise, the set is empty.

[ Note: This set is a (possibly-empty) set of id-expressions, each of which is either e or a subexpression of e. [ Example: In the following example, the set of potential results of the initializer of n contains the first S::x subexpression, but not the second S::x subexpression.

struct S { static const int x = 0; };
const int &f(const int &r);
int n = b ? (1, S::x)  // S::x is not odr-used here
          : f(S::x);   // S::x is odr-used here, so
                       // a definition is required

 — end example ]  — end note ]

A variable x whose name appears as a potentially-evaluated expression ex is odr-used by ex unless applying the lvalue-to-rvalue conversion ([conv.lval]) to x yields a constant expression ([expr.const]) that does not invoke any non-trivial functions and, if x is an object, ex is an element of the set of potential results of an expression e, where either the lvalue-to-rvalue conversion ([conv.lval]) is applied to e, or e is a discarded-value expression (Clause [expr]). this is odr-used if it appears as a potentially-evaluated expression (including as the result of the implicit transformation in the body of a non-static member function ([class.mfct.non-static])). A virtual member function is odr-used if it is not pure. A function whose name appears as a potentially-evaluated expression is odr-used if it is the unique lookup result or the selected member of a set of overloaded functions ([basic.lookup], [over.match], [over.over]), unless it is a pure virtual function and its name is not explicitly qualified. [ Note: This covers calls to named functions ([expr.call]), operator overloading (Clause [over]), user-defined conversions ([class.conv.fct]), allocation function for placement new ([expr.new]), as well as non-default initialization ([dcl.init]). A constructor selected to copy or move an object of class type is odr-used even if the call is actually elided by the implementation ([class.copy]).  — end note ] An allocation or deallocation function for a class is odr-used by a new expression appearing in a potentially-evaluated expression as specified in [expr.new] and [class.free]. A deallocation function for a class is odr-used by a delete expression appearing in a potentially-evaluated expression as specified in [expr.delete] and [class.free]. A non-placement allocation or deallocation function for a class is odr-used by the definition of a constructor of that class. A non-placement deallocation function for a class is odr-used by the definition of the destructor of that class, or by being selected by the lookup at the point of definition of a virtual destructor ([class.dtor]).26 An assignment operator function in a class is odr-used by an implicitly-defined copy-assignment or move-assignment function for another class as specified in [class.copy]. A default constructor for a class is odr-used by default initialization or value initialization as specified in [dcl.init]. A constructor for a class is odr-used as specified in [dcl.init]. A destructor for a class is odr-used if it is potentially invoked ([class.dtor]).

Every program shall contain exactly one definition of every non-inline function or variable that is odr-used in that program; no diagnostic required. The definition can appear explicitly in the program, it can be found in the standard or a user-defined library, or (when appropriate) it is implicitly defined (see [class.ctor], [class.dtor] and [class.copy]). An inline function shall be defined in every translation unit in which it is odr-used.

Exactly one definition of a class is required in a translation unit if the class is used in a way that requires the class type to be complete. [ Example: the following complete translation unit is well-formed, even though it never defines X:

struct X;                       // declare X as a struct type
struct X* x1;                   // use X in pointer formation
X* x2;                          // use X in pointer formation

 — end example ] [ Note: The rules for declarations and expressions describe in which contexts complete class types are required. A class type T must be complete if:

• (5.1)

  an object of type T is defined ([basic.def]), or

• (5.2)

  a non-static class data member of type T is declared ([class.mem]), or

• (5.3)

  T is used as the object type or array element type in a new-expression ([expr.new]), or

• (5.4)

  an lvalue-to-rvalue conversion is applied to a glvalue referring to an object of type T ([conv.lval]), or

• (5.5)

  an expression is converted (either implicitly or explicitly) to type T (Clause [conv], [expr.type.conv], [expr.dynamic.cast], [expr.static.cast], [expr.cast]), or

• (5.6)

  an expression that is not a null pointer constant, and has type other than cv void*, is converted to the type pointer to T or reference to T using a standard conversion (Clause [conv]), a dynamic_cast ([expr.dynamic.cast]) or a static_cast ([expr.static.cast]), or

• (5.7)

  a class member access operator is applied to an expression of type T ([expr.ref]), or

• (5.8)

  the typeid operator ([expr.typeid]) or the sizeof operator ([expr.sizeof]) is applied to an operand of type T, or

• (5.9)

  a function with a return type or argument type of type T is defined ([basic.def]) or called ([expr.call]), or

• (5.10)

  a class with a base class of type T is defined (Clause [class.derived]), or

• (5.11)

  an lvalue of type T is assigned to ([expr.ass]), or

• (5.12)

  the type T is the subject of an alignof expression ([expr.alignof]), or

• (5.13)

  an exception-declaration has type T, reference to T, or pointer to T ([except.handle]).

 — end note ]

There can be more than one definition of a class type (Clause [class]), enumeration type ([dcl.enum]), inline function with external linkage ([dcl.fct.spec]), class template (Clause [temp]), non-static function template ([temp.fct]), static data member of a class template ([temp.static]), member function of a class template ([temp.mem.func]), or template specialization for which some template parameters are not specified ([temp.spec], [temp.class.spec]) in a program provided that each definition appears in a different translation unit, and provided the definitions satisfy the following requirements. Given such an entity named D defined in more than one translation unit, then

• (6.1)

  each definition of D shall consist of the same sequence of tokens; and

• (6.2)

  in each definition of D, corresponding names, looked up according to [basic.lookup], shall refer to an entity defined within the definition of D, or shall refer to the same entity, after overload resolution ([over.match]) and after matching of partial template specialization ([temp.over]), except that a name can refer to a non-volatile const object with internal or no linkage if the object has the same literal type in all definitions of D, and the object is initialized with a constant expression ([expr.const]), and the object is not odr-used, and the object has the same value in all definitions of D; and

• (6.3)

  in each definition of D, corresponding entities shall have the same language linkage; and

• (6.4)

  in each definition of D, the overloaded operators referred to, the implicit calls to conversion functions, constructors, operator new functions and operator delete functions, shall refer to the same function, or to a function defined within the definition of D; and

• (6.5)

  in each definition of D, a default argument used by an (implicit or explicit) function call is treated as if its token sequence were present in the definition of D; that is, the default argument is subject to the three requirements described above (and, if the default argument has sub-expressions with default arguments, this requirement applies recursively).27

• (6.6)

  if D is a class with an implicitly-declared constructor ([class.ctor]), it is as if the constructor was implicitly defined in every translation unit where it is odr-used, and the implicit definition in every translation unit shall call the same constructor for a base class or a class member of D. [ Example:

  //translation unit 1:
  struct X {
    X(int);
    X(int, int);
  };
  X::X(int = 0) { }
  class D: public X { };
  D d2;                           // X(int) called by D()
  
  //translation unit 2:
  struct X {
    X(int);
    X(int, int);
  };
  X::X(int = 0, int = 0) { }
  class D: public X { };          // X(int, int) called by D();
                                  // D()'s implicit definition
                                  // violates the ODR
  

   — end example ]

If D is a template and is defined in more than one translation unit, then the preceding requirements shall apply both to names from the template's enclosing scope used in the template definition ([temp.nondep]), and also to dependent names at the point of instantiation ([temp.dep]). If the definitions of D satisfy all these requirements, then the behavior is as if there were a single definition of D. If the definitions of D do not satisfy these requirements, then the behavior is undefined.

The name lookup rules apply uniformly to all names (including typedef-names ([dcl.typedef]), namespace-names ([basic.namespace]), and class-names ([class.name])) wherever the grammar allows such names in the context discussed by a particular rule. Name lookup associates the use of a name with a declaration ([basic.def]) of that name. Name lookup shall find an unambiguous declaration for the name (see [class.member.lookup]). Name lookup may associate more than one declaration with a name if it finds the name to be a function name; the declarations are said to form a set of overloaded functions ([over.load]). Overload resolution ([over.match]) takes place after name lookup has succeeded. The access rules (Clause [class.access]) are considered only once name lookup and function overload resolution (if applicable) have succeeded. Only after name lookup, function overload resolution (if applicable) and access checking have succeeded are the attributes introduced by the name's declaration used further in expression processing (Clause [expr]).

A name “looked up in the context of an expression” is looked up as an unqualified name in the scope where the expression is found.

The injected-class-name of a class (Clause [class]) is also considered to be a member of that class for the purposes of name hiding and lookup.

[ Note: [basic.link] discusses linkage issues. The notions of scope, point of declaration and name hiding are discussed in [basic.scope].  — end note ]

A program consists of one or more translation units (Clause [lex]) linked together. A translation unit consists of a sequence of declarations.

translation-unit:
    declaration-seqopt

A name is said to have linkage when it might denote the same object, reference, function, type, template, namespace or value as a name introduced by a declaration in another scope:

• (2.1)

  When a name has external linkage, the entity it denotes can be referred to by names from scopes of other translation units or from other scopes of the same translation unit.

• (2.2)

  When a name has internal linkage, the entity it denotes can be referred to by names from other scopes in the same translation unit.

• (2.3)

  When a name has no linkage, the entity it denotes cannot be referred to by names from other scopes.

A name having namespace scope ([basic.scope.namespace]) has internal linkage if it is the name of

• (3.1)

  a variable, function or function template that is explicitly declared static; or,

• (3.2)

  a non-volatile variable that is explicitly declared const or constexpr and neither explicitly declared extern nor previously declared to have external linkage; or

• (3.3)

  a data member of an anonymous union.

An unnamed namespace or a namespace declared directly or indirectly within an unnamed namespace has internal linkage. All other namespaces have external linkage. A name having namespace scope that has not been given internal linkage above has the same linkage as the enclosing namespace if it is the name of

• (4.1)

  a variable; or

• (4.2)

  a function; or

• (4.3)

  a named class (Clause [class]), or an unnamed class defined in a typedef declaration in which the class has the typedef name for linkage purposes ([dcl.typedef]); or

• (4.4)

  a named enumeration ([dcl.enum]), or an unnamed enumeration defined in a typedef declaration in which the enumeration has the typedef name for linkage purposes ([dcl.typedef]); or

• (4.5)

  an enumerator belonging to an enumeration with linkage; or

• (4.6)

  a template.

In addition, a member function, static data member, a named class or enumeration of class scope, or an unnamed class or enumeration defined in a class-scope typedef declaration such that the class or enumeration has the typedef name for linkage purposes ([dcl.typedef]), has external linkage if the name of the class has external linkage.

The name of a function declared in block scope and the name of a variable declared by a block scope extern declaration have linkage. If there is a visible declaration of an entity with linkage having the same name and type, ignoring entities declared outside the innermost enclosing namespace scope, the block scope declaration declares that same entity and receives the linkage of the previous declaration. If there is more than one such matching entity, the program is ill-formed. Otherwise, if no matching entity is found, the block scope entity receives external linkage.[ Example:

static void f();
static int i = 0;               // #1
void g() {
  extern void f();              // internal linkage
  int i;                        // #2 i has no linkage
  {
    extern void f();            // internal linkage
    extern int i;               // #3 external linkage
  }
}

There are three objects named i in this program. The object with internal linkage introduced by the declaration in global scope (line #1 ), the object with automatic storage duration and no linkage introduced by the declaration on line #2, and the object with static storage duration and external linkage introduced by the declaration on line #3.  — end example ]

When a block scope declaration of an entity with linkage is not found to refer to some other declaration, then that entity is a member of the innermost enclosing namespace. However such a declaration does not introduce the member name in its namespace scope. [ Example:

namespace X {
  void p() {
    q();                        // error: q not yet declared
    extern void q();            // q is a member of namespace X
  }

  void middle() {
    q();                        // error: q not yet declared
  }

  void q() { /* ... */ }       // definition of X::q        
}

void q() { /* ... */ }         // some other, unrelated q

 — end example ]

Names not covered by these rules have no linkage. Moreover, except as noted, a name declared at block scope ([basic.scope.block]) has no linkage. A type is said to have linkage if and only if:

• (8.1)

  it is a class or enumeration type that is named (or has a name for linkage purposes ([dcl.typedef])) and the name has linkage; or

• (8.2)

  it is an unnamed class or enumeration member of a class with linkage; or

• (8.3)

  it is a specialization of a class template (Clause [temp])34; or

• (8.4)

  it is a fundamental type ([basic.fundamental]); or

• (8.5)

  it is a compound type ([basic.compound]) other than a class or enumeration, compounded exclusively from types that have linkage; or

• (8.6)

  it is a cv-qualified ([basic.type.qualifier]) version of a type that has linkage.

A type without linkage shall not be used as the type of a variable or function with external linkage unless

• (8.7)

  the entity has C language linkage ([dcl.link]), or

• (8.8)

  the entity is declared within an unnamed namespace ([namespace.def]), or

• (8.9)

  the entity is not odr-used ([basic.def.odr]) or is defined in the same translation unit.

[ Note: In other words, a type without linkage contains a class or enumeration that cannot be named outside its translation unit. An entity with external linkage declared using such a type could not correspond to any other entity in another translation unit of the program and thus must be defined in the translation unit if it is odr-used. Also note that classes with linkage may contain members whose types do not have linkage, and that typedef names are ignored in the determination of whether a type has linkage.  — end note ]

[ Example:

template <class T> struct B {
  void g(T) { }
  void h(T);
  friend void i(B, T) { }
};

void f() {
  struct A { int x; };  // no linkage
  A a = { 1 };
  B<A> ba;              // declares B<A>::g(A) and B<A>::h(A)
  ba.g(a);              // OK
  ba.h(a);              // error: B<A>::h(A) not defined in the translation unit
  i(ba, a);             // OK
}

 — end example ]

Two names that are the same (Clause [basic]) and that are declared in different scopes shall denote the same variable, function, type, enumerator, template or namespace if

• (9.1)

  both names have external linkage or else both names have internal linkage and are declared in the same translation unit; and

• (9.2)

  both names refer to members of the same namespace or to members, not by inheritance, of the same class; and

• (9.3)

  when both names denote functions, the parameter-type-lists of the functions ([dcl.fct]) are identical; and

• (9.4)

  when both names denote function templates, the signatures ([temp.over.link]) are the same.

After all adjustments of types (during which typedefs ([dcl.typedef]) are replaced by their definitions), the types specified by all declarations referring to a given variable or function shall be identical, except that declarations for an array object can specify array types that differ by the presence or absence of a major array bound ([dcl.array]). A violation of this rule on type identity does not require a diagnostic.

[ Note: Linkage to non-C++ declarations can be achieved using a linkage-specification ([dcl.link]).  — end note ]