This document contains the C++ core language issues that have been categorized as Defect Reports by the Committee (J16 + WG21), that is, issues with status "DR," "WP," and "TC1," along with their proposed resolutions. ONLY RESOLUTIONS FOR ISSUES WITH TC1 STATUS ARE PART OF THE INTERNATIONAL STANDARD FOR C++. The other issues are provided for informational purposes only, as an indication of the intent of the Committee. They should not be considered definitive until or unless they appear in an approved Technical Corrigendum or revised International Standard for C++.

This document is part of a group of related documents that together describe the issues that have been raised regarding the C++ Standard. The other documents in the group are:

• Active Issues List, which contains explanatory material for the entire document group and a list of the issues that have not yet been acted upon by the Committee.
• Closed Issues List, which contains the issues which the Committee has decided are not defects in the International Standard, including a brief rationale explaining the reason for the decision.
• Table of Contents, which contains a summary listing of all issues in numerical order.
• Index by Section, which contains a summary listing of all issues arranged in the order of the sections of the Standard with which they deal most directly.
• Index by Status, which contains a summary listing of all issues grouped by status.

For more information, including a description of the meaning of the issue status codes and instructions on reporting new issues, please see the Active Issues List.

Issues with "DR" Status

474. Block-scope extern declarations in namespace members

[Voted into WP at October 2005 meeting.]

Consider the following bit of code:

    namespace N {
      struct S {
        void f();
      };
    }
    using namespace N;
    void S::f() {
      extern void g();  // ::g or N::g?
    }

In 3.5  basic.link paragraph 7 the Standard says (among other things),

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.

The question then is whether N is an “enclosing namespace” for the local declaration of g()?

Proposed resolution (October 2004):

Add the following text as a new paragraph at the end of 7.3.1  namespace.def:

The enclosing namespaces of a declaration are those namespaces in which the declaration lexically appears, except for a redeclaration of a namespace member outside its original namespace (e.g., a definition as specified in 7.3.1.2  namespace.memdef). Such a redeclaration has the same enclosing namespaces as the original declaration. [Example:

  namespace Q {
    namespace V {
      void f(); // enclosing namespaces are the global namespace, Q, and Q::V
      class C { void m(); };
    }
    void V::f() { // enclosing namespaces are the global namespace, Q, and Q::V
      extern void h(); // ... so this declares Q::V::h
    }
    void V::C::m() { // enclosing namespaces are the global namespace, Q, and Q::V
    }
  }

—end example]

348. delete and user-written deallocation functions

[Voted into WP at October 2005 meeting.]

Standard is clear on behaviour of default allocation/deallocation functions. However, it is surpisingly vague on requirements to the behaviour of user-defined deallocation function and an interaction between delete-expression and deallocation function. This caused a heated argument on fido7.su.c-cpp newsgroup.

Resume:

It is not clear if user-supplied deallocation function is called from delete-expr when the operand of delete-expr is the null pointer (5.3.5  expr.delete). If it is, standard does not specify what user-supplied deallocation function shall do with the null pointer operand (18.4.1  lib.new.delete). Instead, Standard uses the term "has no effect", which meaning is too vague in context given (5.3.5  expr.delete).

Description:

Consider statements

   char* p= 0; //result of failed non-throwing ::new char[]
   ::delete[] p;

Standard does NOT specify whether user-defined deallocation function should be called in this case, or not.

Specifically, standard says in 5.3.5  expr.delete paragraph 2:

...if the value of the operand of delete is the null pointer the operation has no effect.

Furthermore, in para 4 standard says on default deallocation function:

If the delete-expression calls the implementation deallocation function (3.7.3.2  basic.stc.dynamic.deallocation), if the operand of the delete expression is not the null pointer constant, ...

If "has no effect" is a requirement to the deallocation function, then it should be stated in 3.7.3.2  basic.stc.dynamic.deallocation, or in 18.4.1.1  lib.new.delete.single and 18.4.1.2  lib.new.delete.array, and it should be stated explicitly.

Furthermore, standard does NOT specify what actions shall be performed by user-supplied deallocation function if NULL is given (18.4.1.1  lib.new.delete.single paragraph 12):

Required behaviour: accept a value of ptr that is null or that was returned by an earlier call to the default operator new(std::size_t) or operator new(std::size_t, const std::nothrow_t&).

The same corresponds to ::delete[] case.

Expected solution:

1. Make it clear that delete-expr will not call deallocation function if null pointer is given (in 5.3.5  expr.delete).
2. Specify what user deallocation function shall do when null is given (either in 3.7.3.2  basic.stc.dynamic.deallocation, or in 18.4.1.1  lib.new.delete.single, and 18.4.1.2  lib.new.delete.array).

Notes from October 2002 meeting:

We believe that study of 18.4.1.1  lib.new.delete.single paragraphs 12 and 13, 18.4.1.2  lib.new.delete.array paragraphs 11 and 12, and 3.7.3.2  basic.stc.dynamic.deallocation paragraph 3 shows that the system-provided operator delete functions must accept a null pointer and ignore it. Those sections also show that a user-written replacement for the system-provided operator delete functions must accept a null pointer. There is no requirement that such functions ignore a null pointer, which is okay -- perhaps the reason for replacing the system-provided functions is to do something special with null pointer values (e.g., log such calls and return).

We believe that the standard should not require an implementation to call a delete function with a null pointer, but it must allow that. For the system-provided delete functions or replacements thereof, the standard already makes it clear that the delete function must accept a null pointer. For class-specific delete functions, we believe the standard should require that such functions accept a null pointer, though it should not mandate what they do with null pointers.

5.3.5  expr.delete needs to be updated to say that it is unspecified whether or not the operator delete function is called with a null pointer, and 3.7.3.2  basic.stc.dynamic.deallocation needs to be updated to say that any deallocation function must accept a null pointer.

Proposed resolution (October, 2004):

1. Change 5.3.5  expr.delete paragraph 2 as indicated:

   If the operand has a class type, the operand is converted to a pointer type by calling the above-mentioned conversion function, and the converted operand is used in place of the original operand for the remainder of this section. In either alternative, if the value of the operand of delete is the null pointer the operation has no effect may be a null pointer value. If it is not a null pointer value, in In the first alternative (delete object), the value of the operand of delete shall be a pointer to a non-array object or a pointer to a sub-object (1.8  intro.object) representing a base class of such an object (clause 10  class.derived)...

2. Change 5.3.5  expr.delete paragraph 4 as follows (note that the old wording reflects the changes proposed by issue 442:

   The cast-expression in a delete-expression shall be evaluated exactly once. If the delete-expression calls the implementation deallocation function (3.7.3.2  basic.stc.dynamic.deallocation), and if the value of the operand of the delete expression is not a null pointer, the deallocation function will deallocate the storage referenced by the pointer thus rendering the pointer invalid. [Note: the value of a pointer that refers to deallocated storage is indeterminate. —end note]

3. Change 5.3.5  expr.delete paragraphs 6-7 as follows:

   The If the value of the operand of the delete-expression is not a null pointer value, the delete-expression will invoke the destructor (if any) for the object or the elements of the array being deleted. In the case of an array, the elements will be destroyed in order of decreasing address (that is, in reverse order of the completion of their constructor; see 12.6.2  class.base.init).

   The If the value of the operand of the delete-expression is not a null pointer value, the delete-expression will call a deallocation function (3.7.3.2  basic.stc.dynamic.deallocation). Otherwise, it is unspecified whether the deallocation function will be called. [Note: The deallocation function is called regardless of whether the destructor for the object or some element of the array throws an exception. —end note]

4. Change 3.7.3.2  basic.stc.dynamic.deallocation paragraph 3 as indicated:

   The value of the first argument supplied to one of the a deallocation functions provided in the standard library may be a null pointer value; if so, and if the deallocation function is one supplied in the standard library, the call to the deallocation function has no effect. Otherwise, the value supplied to operator delete(void*) in the standard library shall be one of the values returned by a previous invocation of either operator new(std::size_t) or operator new(std::size_t, const std::nothrow_t&) in the standard library, and the value supplied to operator delete[](void*) in the standard library shall be one of the values returned by a previous invocation of either operator new[](std::size_t) or operator new[](std::size_t, const std::nothrow_t&) in the standard library.

[Note: this resolution also resolves issue 442.]

451. Expressions with invalid results and ill-formedness

[Voted into WP at October 2005 meeting.]

Clause 5  expr par. 5 of the standard says:

If during the evaluation of an expression, the result is not mathematically defined or not in the range of representable values for its type, the behavior is undefined, unless such an expression is a constant expression (5.19), in which case the program is ill-formed.

Well, we do know that except in some contexts (e.g. controlling expression of a #if, array bounds), a compiler is not required to evaluate constant-expressions in compile time, right?

Now, let us consider, the following simple snippet:

  if (a && 1/0)
      ...

I think the intent is actually the latter, so I propose the following rewording of the quoted section:

If an expression is evaluated but its result is not mathematically defined or not in the range of representable values for its type the behavior is undefined, unless such an expression is a constant expression (5.19) that shall be evaluated during program translation, in which case the program is ill-formed.

Rationale (March, 2004):

We feel the standard is clear enough. The quoted sentence does begin "If during the evaluation of an expression, ..." so the rest of the sentence does not apply to an expression that is not evaluated.

Note (September, 2004):

Gennaro Prota feels that the CWG missed the point of his original comment: unless a constant expression appears in a context that requires a constant expression, an implementation is permitted to defer its evaluation to runtime. An evaluation that fails at runtime cannot affect the well-formedness of the program; only expressions that are evaluated at compile time can make a program ill-formed.

The status has been reset to “open” to allow further discussion.

Proposed resolution (October, 2004):

Change paragraph 5 of 5  expr as indicated:

If during the evaluation of an expression, the result is not mathematically defined or not in the range of representable values for its type, the behavior is undefined, unless such an expression is a constant expression appears where an integral constant expression is required (5.19  expr.const), in which case the program is ill-formed.

299. Conversion on array bound expression in new

[Voted into WP at October 2005 meeting.]

In 5.3.4  expr.new, the standard says that the expression in an array-new has to have integral type. There's already a DR (issue 74) that says it should also be allowed to have enumeration type. But, it should probably also say that it can have a class type with a single conversion to integral type; in other words the same thing as in 6.4.2  stmt.switch paragraph 2.

Suggested resolution:

In 5.3.4  expr.new paragraph 6, replace "integral or enumeration type (3.9.1  basic.fundamental)" with "integral or enumeration type (3.9.1  basic.fundamental), or a class type for which a single conversion function to integral or enumeration type exists".

Proposed resolution (October, 2004):

Change 5.3.4  expr.new paragraph 6 as follows:

Every constant-expression in a direct-new-declarator shall be an integral constant expression (5.19  expr.const) and evaluate to a strictly positive value. The expression in a direct-new-declarator shall have be of integral type, or enumeration type (3.9.1), or a class type for which a single conversion function to integral or enumeration type exists (12.3  class.conv). If the expression is of class type, the expression is converted by calling the conversion function, and the result of the conversion is used in place of the original expression. The value of the expression shall bewith a non-negative value. [Example: ...

Proposed resolution (April, 2005):

Change 5.3.4  expr.new paragraph 6 as follows:

Every constant-expression in a direct-new-declarator shall be an integral constant expression (5.19  expr.const) and evaluate to a strictly positive value. The expression in a direct-new-declarator shall have integral or enumeration type (3.9.1  basic.fundamental) with a non-negative value be of integral type, enumeration type, or a class type for which a single conversion function to integral or enumeration type exists (12.3  class.conv). If the expression is of class type, the expression is converted by calling that conversion function, and the result of the conversion is used in place of the original expression. If the value of the expression is negative, the behavior is undefined. [Example: ...

442. Incorrect use of null pointer constant in description of delete operator

[Voted into WP at October 2005 meeting.]

After some discussion in comp.lang.c++.moderated we came to the conclusion that there seems to be a defect in 5.3.5  expr.delete/4, which says:

The cast-expression in a delete-expression shall be evaluated exactly once. If the delete-expression calls the implementation deallocation function (3.7.3.2), and if the operand of the delete expression is not the null pointer constant, the deallocation function will deallocate the storage referenced by the pointer thus rendering the pointer invalid. [Note: the value of a pointer that refers to deallocated storage is indeterminate. ]

In the second sentence, the term "null pointer constant" should be changed to "null pointer". In its present form, the passage claims that the deallocation function will deallocate the storage refered to by a null pointer that did not come from a null pointer constant in the delete expression. Besides, how can the null pointer constant be the operand of a delete expression, as "delete 0" is an error because delete requires a pointer type or a class type having a single conversion function to a pointer type?

See also issue 348.

Proposed resolution:

Change the indicated sentence of 5.3.5  expr.delete paragraph 4 as follows:

If the delete-expression calls the implementation deallocation function (3.7.3.2  basic.stc.dynamic.deallocation), and if the value of the operand of the delete expression is not the a null pointer constant, the deallocation function will deallocate the storage referenced by the pointer thus rendering the pointer invalid.

Notes from October 2004 meeting:

This wording is superseded by, and this issue will be resolved by, the resolution of issue 348.

Proposed resolution (April, 2005):

This issue is resolved by the resolution of issue 348.

497. Missing required initialization in example

[Voted into WP at October 2005 meeting.]

5.5  expr.mptr.oper paragraph 5 contains the following example:

    struct S {
        mutable int i;
    };
    const S cs;
    int S::* pm = &S::i;   // pm refers to mutable member S::i
    cs.*pm = 88;           // ill-formed: cs is a const object

The const object cs is not explicitly initialized, and class S does not have a user-declared default constructor. This makes the code ill-formed as per 8.5  dcl.init paragraph 9.

Proposed resolution (April, 2005):

Change the example in 5.5  expr.mptr.oper paragraph 5 to read as follows:

    struct S {
        S() : i(0) { }
        mutable int i;
    };
    void f()
    {
        const S cs;
        int S::* pm = &S::i;   // pm refers to mutable member S::i
        cs.*pm = 88;           // ill-formed: cs is a const object
    }

446. Does an lvalue-to-rvalue conversion on the "?" operator produce a temporary?

[Voted into WP at October 2005 meeting.]

The problem occurs when the value of the operator is determined to be an rvalue, the selected argument is an lvalue, the type is a class type and a non-const member is invoked on the modifiable rvalue result.

    struct B {
        int v;
        B (int v) : v(v) { }
        void inc () { ++ v; }
        };
    struct D : B {
        D (int v) : B(v) { }
        };

    B b1(42);
    (0 ? B(13) : b1).inc();
    assert(b1.v == 42);

The types of the second and third operands are the same and one is an rvalue. Nothing changes until p6 where an lvalue to rvalue conversion is performed on the third operand. 12.2  class.temporary states that an lvalue to rvalue conversion produces a temporary and there is nothing to remove it. It seems clear that the assertion must pass, yet most implementations fail.

There seems to be a defect in p3 b2 b1. First, the conditions to get here and pass the test.

If E1 and E2 have class type, and the underlying class types are the same or one is a base class of the other: E1 can be converted to match E2 if the class of T2 is the same type as, or a base class of, the class of T1, and the cv-qualification of T2 is the same cv-qualification as, or a greater cv-qualification than, the cv-qualification of T1.

If both E1 and E2 are lvalues, passing the conditions here also passes the conditions for p3 b1. Thus, at least one is an rvalue. The case of two rvalues is not interesting and the action is covered by the case when E1 is an rvalue.

    (0 ? D(13) : b1).inc();
    assert(b1.v == 42);

E1 is changed to an rvalue of type T2 that still refers to the original source class object (or the appropriate subobject thereof). [Note: that is, no copy is made. ]

Having changed the rvalue to base type, we are back to the above case where an lvalue to rvalue conversion is required on the third operand at p6. Again, most implementations fail.

The remaining case, E1 an lvalue and E2 an rvalue, is the defect.

    D d1(42);
    (0 ? B(13) : d1).inc();
    assert(d1.v == 42);

The above quote states that an lvalue of type T1 is changed to an rvalue of type T2 without creating a temporary. This is in contradiction to everything else in the standard about lvalue to rvalue conversions. Most implementations pass in spite of the defect.

The usual accessible and unambiguous is missing from the base class.

There seems to be two possible solutions. Following other temporary creations would produce a temporary rvalue of type T1 and change it to an rvalue of type T2. Keeping the no copy aspect of this bullet intact would change the lvalue of type T1 to an lvalue of type T2. In this case the lvalue to rvalue conversion would happen in p6 as usual.

Suggested wording for p3 b2 b1

The base part:

If E1 and E2 have class type, and the underlying class types are the same or one is a base class of the other: E1 can be converted to match E2 if the class of T2 is the same type as, or an accessible and unambiguous base class of, the class of T1, and the cv-qualification of T2 is the same cv-qualification as, or a greater cv-qualification than, the cv-qualification of T1. If the conversion is applied:

The same type temporary version:

If E1 is an lvalue, an lvalue to rvalue conversion is applied. The resulting or original rvalue is changed to an rvalue of type T2 that refers to the same class object (or the appropriate subobject thereof). [Note: that is, no copy is made in changing the type of the rvalue. ]

The never copy version:

The lvalue(rvalue) E1 is changed to an lvalue(rvalue) of type T2 that refers to the original class object (or the appropriate subobject thereof). [Note: that is, no copy is made. ]

The test case was posted to clc++m and results for implementations were reported.

#include <cassert>
struct B {
    int v;
    B (int v) : v(v) { }
    void inc () { ++ v; }
    };
struct D : B {
    D (int v) : B(v) { }
    };
int main () {
    B b1(42);
    D d1(42);
    (0 ? B(13) : b1).inc();
    assert(b1.v == 42);
    (0 ? D(13) : b1).inc();
    assert(b1.v == 42);
    (0 ? B(13) : d1).inc();
    assert(d1.v == 42);
    }

// CbuilderX(EDG301) FFF  Rob Williscroft
// ICC-8.0           FFF  Alexander Stippler
// COMO-4.301        FFF  Alexander Stippler

// BCC-5.4           FFP  Rob Williscroft
// BCC32-5.5         FFP  John Potter
// BCC32-5.65        FFP  Rob Williscroft
// VC-6.0            FFP  Stephen Howe
// VC-7.0            FFP  Ben Hutchings
// VC-7.1            FFP  Stephen Howe
// OpenWatcom-1.1    FFP  Stephen Howe

// Sun C++-6.2       PFF  Ron Natalie

// GCC-3.2           PFP  John Potter
// GCC-3.3           PFP  Alexander Stippler

// GCC-2.95          PPP  Ben Hutchings
// GCC-3.4           PPP  Florian Weimer

I see no defect with regards to lvalue to rvalue conversions; however, there seems to be disagreement about what it means by implementers. It may not be surprising because 5.16 and passing a POD struct to an ellipsis are the only places where an lvalue to rvalue conversion applies to a class type. Most lvalue to rvalue conversions are on basic types as operands of builtin operators.

Notes from the March 2004 meeting:

We decided all "?" operators that return a class rvalue should copy the second or third operand to a temporary. See issue 86.

Proposed resolution (October 2004):

1. Change 5.16  expr.cond paragraph 3 bullet 2 sub-bullet 1 as follows:

   if E1 and E2 have class type, and the underlying class types are the same or one is a base class of the other: E1 can be converted to match E2 if the class of T2 is the same type as, or a base class of, the class of T1, and the cv-qualification of T2 is the same cv-qualification as, or a greater cv-qualification than, the cv-qualification of T1. If the conversion is applied, E1 is changed to an rvalue of type T2 that still refers to the original source class object (or the appropriate subobject thereof). [Note: that is, no copy is made. —end note] by copy-initializing a temporary of type T2 from E1 and using that temporary as the converted operand.

2. Change 5.16  expr.cond paragraph 6 bullet 1 as follows:

   The second and third operands have the same type; the result is of that type. If the operands have class type, the result is an rvalue temporary of the result type, which is copy-initialized from either the second operand or the third operand depending on the value of the first operand.

3. Change 4.1  conv.lval paragraph 2 as follows:

   The value contained in the object indicated by the lvalue is the rvalue result. When an lvalue-to-rvalue conversion occurs within the operand of sizeof (5.3.3  expr.sizeof) the value contained in the referenced object is not accessed, since that operator does not evaluate its operand. Otherwise, if the lvalue has a class type, the conversion copy-initializes a temporary of type T from the lvalue and the result of the conversion is an rvalue for the temporary. Otherwise, the value contained in the object indicated by the lvalue is the rvalue result.

[Note: this wording partially resolves issue 86. See also issue 462.]

317. Can a function be declared inline after it has been called?

[Voted into WP at October 2005 meeting.]

Steve Clamage: Consider this sequence of declarations:

  void foo() { ... }
  inline void foo();

Bjarne Stroustrup: Neither could I, so I looked in the ARM, which addressed this case (apparently for member function only) in some detail in 7.1.2 (pp103-104).

The ARM allows declaring a function inline after its initial declaration, as long as it has not been called.

Steve Clamage: If the above code is valid, how about this:

  void foo() { ... }    // define foo
  void bar() { foo(); } // use foo
  inline void foo();    // declare foo inline

Bjarne Stroustrup: ... and [the ARM] disallows declaring a function inline after it has been called.

This may still be a good resolution.

Steve Clamage: But the situation in the ARM is the reverse: Declare a function inline, and define it later (with no intervening call). That's a long-standing rule in C++, and allows you to write member function definitions outside the class.

In my example, the compiler could reasonably process the entire function as out-of-line, and not discover the inline declaration until it was too late to save the information necessary for inline generation. The equivalent of another compiler pass would be needed to handle this situation.

Bjarne Stroustrup: I see, and I think your argument it conclusive.

Steve Clamage: I'd like to open a core issue on this point, and I recommend wording along the lines of: "A function defined without an inline specifier shall not be followed by a declaration having an inline specifier."

I'd still like to allow the common idiom

  class T {
    int f();
  };
  inline int T::f() { ... }

Martin Sebor: Since the inline keyword is just a hint to the compiler, I don't see any harm in allowing the construct. Your hypothetical compiler can simply ignore the inline on the second declaration. On the other hand, I feel that adding another special rule will unnecessarily complicate the language.

Steve Clamage: The inline specifier is more than a hint. You can have multiple definitions of inline functions, but only one definition of a function not declared inline. In particular, suppose the above example were in a header file, and included multiple times in a program.

Proposed resolution (October, 2004):

Add the indicated words to 7.1.2  dcl.fct.spec paragraph 4:

An inline function shall be defined in every translation unit in which it is used and shall have exactly the same definition in every case (3.2  basic.def.odr). [Note: a call to the inline function may be encountered before its definition appears in the translation unit. —end note] If the definition of a function appears in a translation unit before its first declaration as inline, the program is ill-formed. If a function with external linkage is declared inline in one translation unit...

291. Overload resolution needed when binding reference to class rvalue

[Voted into WP at October 2005 meeting.]

There is a place in the Standard where overload resolution is implied but the way that a set of candidate functions is to be formed is omitted. See below.

According to the Standard, when initializing a reference to a non-volatile const class type (cv1 T1) with an rvalue expression (cv2 T2) where cv1 T1 is reference compatible with cv2 T2, the implementation shall proceed in one of the following ways (except when initializing the implicit object parameter of a copy constructor) 8.5.3  dcl.init.ref paragraph 5 bullet 2 sub-bullet 1:

• The reference is bound to the object represented by the rvalue (see 3.10  basic.lval) or to a sub-object within that object.
• A temporary of type "cv1 T2" [sic] is created, and a constructor is called to copy the entire rvalue object into the temporary...

While the first case is quite obvious, the second one is a bit unclear as it says "a constructor is called to copy the entire rvalue object into the temporary" without specifying how the temporary is created -- by direct-initialization or by copy-initialization? As stated in DR 152, this can make a difference when the copy constructor is declared as explicit. How should the set of candidate functions be formed? The most appropriate guess is that it shall proceed as per 13.3.1.3  over.match.ctor.

Another detail worth of note is that in the draft version of the Standard as of 2 December 1996 the second bullet read:

• A temporary of type "cv1 T2" [sic] is created, and a copy constructor is called to copy the entire rvalue object into the temporary...

J. Stephen Adamczyk replied that the reason for changing "a copy constructor" to "a constructor" was to allow for member template converting constructors.

However, the new wording is somewhat in conflict with the footnote #93 that says that when initializing the implicit object parameter of a copy constructor an implementation must eventually choose the first alternative (binding without copying) to avoid infinite recursion. This seems to suggest that a copy constructor is always used for initializing the temporary of type "cv1 T2".

Furthermore, now that the set of candidate functions is not limited to only the copy constructors of T2, there might be some unpleasant consequences. Consider a rather contrived sample below:

    int   * pi = ::new(std::nothrow) int;
    const std::auto_ptr<int>   & ri = std::auto_ptr<int>(pi);

In this example the initialization of the temporary of type '<TT>const std::auto_ptr<int>' (to which 'ri' is meant to be subsequently bound) doesn't fail, as it would had the approach with copy constructors been retained, instead, a yet another temporary gets created as the well-known sequence:

    std::auto_ptr<int>::operator std::auto_ptr_ref<int>()
    std::auto_ptr<int>(std::auto_ptr_ref<int>)

is called (assuming, of course, that the set of candidate functions is formed as per 13.3.1.3  over.match.ctor). The second temporary is transient and gets destroyed at the end of the initialization. I doubt that this is the way that the committee wanted this kind of reference binding to go.

Besides, even if the approach restricting the set of candidates to copy constructors is restored, it is still not clear how the initialization of the temporary (to which the reference is intended to be bound) is to be performed -- using direct-initialization or copy-initialization.

Another place in the Standard that would benefit from a similar clarification is the creation of an exception object, which is delineated in 15.1  except.throw.

David Abrahams (February 2004): It appears, looking at core 291, that there may be a need to tighten up 8.5.3  dcl.init.ref/5.

Please see the attached example file, which demonstrates "move semantics" in C++98. Many compilers fail to compile test 10 because of the way 8.5.3/5 is interpreted. My problem with that interpretation is that test 20:

    typedef X const XC;
    sink2(XC(X()));

8.5.3/5 doesn't demand that a "copy constructor" is used to copy the temporary, only that a constructor is used "to copy the temporary". I hope that when the language is tightened up to specify direct (or copy initialization), that it also unambiguously allows the enclosed test to compile. Not only is it, I believe, within the scope of reasonable interpretation of the current standard, but it's an incredibly important piece of functionality for library writers and users alike.

#include <iostream>
#include <cassert>

template <class T, class X>
struct enable_if_same
{
};

template <class X>
struct enable_if_same<X, X>
{
    typedef char type;
};

struct X
{
    static int cnt;  // count the number of Xs
    
    X()
      : id(++cnt)
      , owner(true)
    {
        std::cout << "X() #" << id << std::endl;
    }
    
    // non-const lvalue - copy ctor
    X(X& rhs)
      : id(++cnt)
      , owner(true)
    {
        std::cout << "copy #" << id << " <- #" << rhs.id << std::endl;
    }

    // const lvalue - T will be deduced as X const
    template <class T>
    X(T& rhs, typename enable_if_same<X const,T>::type = 0)
      : id(++cnt)
      , owner(true)
    {
        std::cout << "copy #" << id << " <- #" << rhs.id << " (const)" << std::endl;
    }

    ~X()
    {
        std::cout << "destroy #" << id << (owner?"":" (EMPTY)") << std::endl;
    }
    
 private:    // Move stuff
    struct ref { ref(X*p) : p(p) {} X* p; };

 public:    // Move stuff
    operator ref() {
        return ref(this);
    }

    // non-const rvalue
    X(ref rhs)
      : id(++cnt)
      , owner(rhs.p->owner)
    {
        std::cout << "MOVE #" << id << " <== #" << rhs.p->id << std::endl;
        rhs.p->owner = false;
        assert(owner);
    }
    
 private:   // Data members
    int id;
    bool owner;
};

int X::cnt;


X source()
{
    return X();
}

X const csource()
{
    return X();
}

void sink(X)
{
    std::cout << "in rvalue sink" << std::endl;
}

void sink2(X&)
{
    std::cout << "in non-const lvalue sink2" << std::endl;
}

void sink2(X const&)
{
    std::cout << "in const lvalue sink2" << std::endl;
}

void sink3(X&)
{
    std::cout << "in non-const lvalue sink3" << std::endl;
}

template <class T>
void tsink(T)
{
    std::cout << "in templated rvalue sink" << std::endl;
}

int main()
{
    std::cout << " ------ test 1, direct init from rvalue ------- " << std::endl;
#ifdef __GNUC__ // GCC having trouble parsing the extra parens
    X z2((0, X() ));
#else
    X z2((X()));
#endif 

    std::cout << " ------ test 2, copy init from rvalue ------- " << std::endl;
    X z4 = X();

    std::cout << " ------ test 3, copy init from lvalue ------- " << std::endl;
    X z5 = z4;

    std::cout << " ------ test 4, direct init from lvalue ------- " << std::endl;
    X z6(z4);
    
    std::cout << " ------ test 5, construct const ------- " << std::endl;
    X const z7;
    
    std::cout << " ------ test 6, copy init from lvalue ------- " << std::endl;
    X z8 = z7;

    std::cout << " ------ test 7, direct init from lvalue ------- " << std::endl;
    X z9(z7);
    
    std::cout << " ------ test 8, pass rvalue by-value ------- " << std::endl;
    sink(source());
    
    std::cout << " ------ test 9, pass const rvalue by-value ------- " << std::endl;
    sink(csource());

    std::cout << " ------ test 10, pass rvalue by overloaded reference ------- " << std::endl;
    // This one fails in Comeau's strict mode due to 8.5.3/5.  GCC 3.3.1 passes it.
    sink2(source());

    std::cout << " ------ test 11, pass const rvalue by overloaded reference ------- " << std::endl;
    sink2(csource());

#if 0    // These two correctly fail to compile, just as desired
    std::cout << " ------ test 12, pass rvalue by non-const reference ------- " << std::endl;
    sink3(source());

    std::cout << " ------ test 13, pass const rvalue by non-const reference ------- " << std::endl;
    sink3(csource());
#endif 

    std::cout << " ------ test 14, pass lvalue by-value ------- " << std::endl;
    sink(z5);
    
    std::cout << " ------ test 15, pass const lvalue by-value ------- " << std::endl;
    sink(z7);

    std::cout << " ------ test 16, pass lvalue by-reference ------- " << std::endl;
    sink2(z4);

    std::cout << " ------ test 17, pass const lvalue by const reference ------- " << std::endl;
    sink2(z7);

    std::cout << " ------ test 18, pass const lvalue by-reference ------- " << std::endl;
#if 0   // correctly fails to compile, just as desired
    sink3(z7);
#endif 

    std::cout << " ------ test 19, pass rvalue by value to template param ------- " << std::endl;
    tsink(source());

    std::cout << " ------ test 20, direct initialize a const A with an A ------- " << std::endl;
    typedef X const XC;
    sink2(XC(X()));
}

Proposed Resolution:

(As proposed by N1610 section 5, with editing.)

Change paragraph 5, second bullet, first sub-bullet, second sub-sub-bullet as follows:

A temporary of type "cv1 T2" [sic] is created, and a constructor is called to copy the entire rvalue object into the temporary via copy-initialization from the entire rvalue object. The reference is bound to the temporary or to a sub-object within the temporary.

The text immediately following that is changed as follows:

The constructor that would be used to make the copy shall be callable whether or not the copy is actually done. The constructor and any conversion function that would be used in the initialization shall be callable whether or not the temporary is actually created.

Note, however, that the way the core working group is leaning on issue 391 (i.e., requiring direct binding) would make this change unnecessary.

Proposed resolution (April, 2005):

This issue is resolved by the resolution of issue 391.

391. Require direct binding of short-lived references to rvalues

[Voted into WP at October 2005 meeting.]

After some email exchanges with Rani Sharoni, I've come up with the following proposal to allow reference binding to non-copyable rvalues in some cases. Rationale and some background appear afterwards.

---- proposal ----

Replace the section of 8.5.3  dcl.init.ref paragraph 5 that begins "If the initializer expression is an rvalue" with the following:

• If the initializer expression is an rvalue, with T2 a class type, and ``cv1 T1'' is reference-compatible with ``cv2 T2,'' the reference is bound as follows:
  • If the lifetime of the reference does not extend beyond the end of the full expression containing the initializer expression, the reference is bound to the object represented by the rvalue (see 3.10  basic.lval) or to a sub-object within that object.
  • otherwise, the reference is bound in one of the following ways (the choice is implementation-defined):
    • [... continues as before - the original wording applies unchanged to longer-lived references]

---- rationale ----

1. The intention of the current wording is to provide the implementation freedom to construct an rvalue of class type at an arbitrary location and copy it zero or more times before binding any reference to it.
2. The standard allows code to call a member function on an rvalue of class type (in 5.2.5  expr.ref, I guess). This means that the implementation can be forced to bind the reference directly, with no freedom to create any temporary copies. e.g.

      class nc {
        nc (nc const &);  // private, nowhere defined
      public:
        nc ();
        nc const &by_ref () const { return *this; }
      };
   
      void f () {
        void g (nc const &);
   
        g (nc());          // Ill-formed
        g (nc().by_ref()); // Ok - binds directly to rvalue
      }
   

   Forcing a direct binding in this way is possible wherever the lifetime of the reference does not extend beyond the containing full expression, since the reference returned by the member function remains valid for this long.
3. As demonstrated above, existing implementations must already be capable of constructing an rvalue of class type in the "right" place the first time. Some compilers already silently allow the direct binding of references to non-copyable rvalues.
4. The change will not break any portable user code. It would break any platform-specific user code that relies on copies being performed by the particular implementation.

---- background ----

The proposal is based on a recent discussion in this group. I originally wanted to leave the implementation free to copy the rvalue if there was a callable copy constructor, and only have to bind directly if none was callable. Unfortunately, a traditional compiler can't always tell whether a function is callable or not, e.g. if the copy constructor is declared but not defined. Rani pointed this out in an example, and suggested that maybe trivial copy constructors should still be allowed (by extension, maybe wherever the compiler can determine callability). I've gone with this version because it's simpler, and I also figure the "as if" rule gives the compiler some freedom with POD types anyway.

Notes from April 2003 meeting:

We agreed generally with the proposal. We were unsure about the need for the restriction regarding long-lived references. We will check with the proposer about that.

Jason Merrill points out that the test case in issue 86 may be a case where we do not want to require direct binding.

Further information from Rani Sharoni (April 2003):

I wasn't aware about the latest suggestion of Raoul as it appears in core issue 391. In our discussions we tried to formulate a different proposal.

The rational, as we understood, behind the implementation freedom to make an extra copying (8.5.3/5/2/12) of the rvalue is to allow return values in registers which on some architectures are not addressable. The example that Raoul and I presented shows that this implementation freedom is not always possible since we can "force" the rvalue to be addressable using additional member function (by_ref). The example only works for short lived rvalues and this is probably why Raoul narrow the suggestion.

I had different rational which was related to the implementation of conditional operator in VC. It seems that when conditional operator is involved VC does use an extra copying when the lifetime of the temporary is extended:

  struct A { /* ctor with side effect */};

  void f(A& x) {
    A const& r = cond ? A(1) : x; // VC actually make an extra copy of
                                  // the rvalue A(1)
  }

I don't know what the consideration behind the VC implementation was (I saw open bug on this issue) but it convinced me to narrow the suggestion.

IMHO such limitation seems to be too strict because it might limit the optimizer since returning class rvalues in registers might be useful (although I'm not aware about any implementation that actually does it). My suggestion was to forbid the extra copying if the ctor is not viable (e.g. A::A(A&) ). In this case the implementation "freedom" doesn't exist (since the code might not compile) and only limits the programmer freedom (e.g. Move Constructors - http://www.cuj.com/experts/2102/alexandr.htm).

Core issue 291 is strongly related to the above issue and I personally prefer to see it resolved first. It seems that VC already supports the resolution I prefer.

Notes from October 2003 meeting:

We ended up feeling that this is just one of a number of cases of optimizations that are widely done by compilers and allowed but not required by the standard. We don't see any strong reason to require compilers to do this particular optimization.

Notes from the March 2004 meeting:

After discussing issue 450, we found ourselves reconsidering this, and we are now inclined to make a change to require the direct binding in all cases, with no restriction on long-lived references. Note that such a change would eliminate the need for a change for issue 291.

Proposed resolution (October, 2004):

Change 8.5.3  dcl.init.ref paragraph 5 bullet 2 sub-bullet 1 as follows:

If the initializer expression is an rvalue, with T2 a class type, and "cv1 T1" is reference-compatible with "cv2 T2", the reference is bound to the object represented by the rvalue (see 3.10  basic.lval) or to a sub-object within that object. in one of the following ways (the choice is implementation-defined):

• The reference is bound to the object represented by the rvalue (see 3.10  basic.lval) or to a sub-object within that object.
• A temporary of type "cv1 T2" [sic] is created, and a constructor is called to copy the entire rvalue object into the temporary. The reference is bound to the temporary or to a sub-object within the temporary.

The constructor that would be used to make the copy shall be callable whether or not the copy is actually done. [Example:

  struct A { };
  struct B : public A { } b;
  extern B f();
  const A& rca = f ();  // Bound Either bound to the A sub-object of the B rvalue,
                        // or the entire B object is copied and the reference
                        // is bound to the A sub-object of the copy

—end example]

[This resolution also resolves issue 291.]

450. Binding a reference to const to a cv-qualified array rvalue

[Voted into WP at October 2005 meeting.]

It's unclear whether the following is valid:

const int N = 10;
const int M = 20;
typedef int T;
void f(T const (&x)[N][M]){}

struct X {
	int i[10][20];
};

X g();

int main()
{
	f(g().i);
}

When you run this through 8.5.3  dcl.init.ref, you sort of end up falling off the end of the standard's description of reference binding. The standard says in the final bullet of paragraph 5 that an array temporary should be created and copy-initialized from the rvalue array, which seems implausible.

I'm not sure what the right answer is. I think I'd be happy with allowing the binding in this case. We would have to introduce a special case like the one for class rvalues.

Notes from the March 2004 meeting:

g++ and EDG give an error. Microsoft (8.0 beta) and Sun accept the example. Our preference is to allow the direct binding (no copy). See the similar issue with class rvalues in issue 391.

Proposed resolution (October, 2004):

1. Insert a new bullet in 8.5.3  dcl.init.ref paragraph 5 bullet 2 before sub-bullet 2 (which begins, “Otherwise, a temporary of type ‘cv1 T1’ is created...”):

   If the initializer expression is an rvalue, with T2 an array type, and “cv1 T1” is reference-compatible with “cv2 T2”, the reference is bound to the object represented by the rvalue (see 3.10  basic.lval).

2. Change 3.10  basic.lval paragraph 2 as follows:

   An lvalue refers to an object or function. Some rvalue expressions — those of (possibly cv-qualified) class or array type or cv-qualified class type — also refer to objects.

401. When is access for template parameter default arguments checked?

[Voted into WP at October 2005 meeting.]

Is the following well-formed?

  class policy {};
  class policy_interface {};
  template <class POLICY_INTERFACE>
  class aph {
  protected:
    typedef POLICY_INTERFACE PI;
  };
  template <class POLICY, class BASE, class PI = typename BASE::PI>
  class ConcretePolicyHolder : public BASE, protected POLICY
  {};
  ConcretePolicyHolder < policy , aph < policy_interface > > foo;
  void xx() { }

The issue is whether the access to the default argument type BASE::PI is checked before or after it is known that BASE is a base class of the template. To some extent, one needs to develop the list of template arguments (and therefore evaluate the default argument) before one can instantiate the template, and one does not know what base classes the template has until it has been instantiated.

Notes from April 2003 meeting:

Shortened example:

  class B {
  protected:
    typedef int A;
  };
  template<class T, class U = typename T::A>
  class X : public T
  { };

The convincing argument here is that if we had only the declaration of the template (including the default argument), we would expect it to be usable in exactly the same way as the version with the definition. However, the special access needed is visible only when the definition is available. So the above should be an error, and information from the definition cannot affect the access of the default arguments.

Proposed Resolution (April 2003):

Add a new paragraph 16 to 14.1  temp.param after paragraph 15:

Since a default template-argument is encountered before any base-clause there is no special access to members used in a default template-argument. [Example:

  class B {};
  template <class T> class C {
  protected:
     typedef T TT;
  };

  template <class U, class V = typename U::TT>
  class D : public U {};

  D <C<B> > d;  // access error, C::TT is protected

--- end example]

Notes from October 2003 meeting:

We decided that template parameter default arguments should have their access checked in the context where they appear without special access for the entity declared (i.e., they are different than normal function default arguments). One reason: we don't know the instance of the template when we need the value. Second reason: compilers want to parse and throw away the form of the template parameter default argument, not save it and check it for each instantiation.

Class templates should be treated the same as function templates in this regard. The base class information is in the same category as friend declarations inside the class itself -- not available. If the body were used one would need to instantiate it in order to know whether one can name it.

Proposed resolution (October, 2004):

Add the following as a new paragraph following the last paragraph of 11  class.access (but before the new paragraph inserted by the resolution of issue 372, if adopted):

The names in a default template-argument (14.1  temp.param) have their access checked in the context in which they appear rather than at any points of use of the default template-argument. [Example:

    class B {};
    template <class T> class C {
    protected:
       typedef T TT;
    };

    template <class U, class V = typename U::TT>
    class D : public U {};

    D <C<B> >* d;  // access error, C::TT is protected

—end example]

354. Null as nontype template argument

[Voted into WP at October 2005 meeting.]

The standard does not permit a null value to be used as a nontype template argument for a nontype template parameter that is a pointer.

This code is accepted by EDG, Microsoft, Borland and Cfront, but rejected by g++ and Sun:

  template <int *p> struct A {};
  A<(int*)0> ai;

I'm not sure this was ever explicitly considered by the committee. Is there any reason to permit this kind of usage?

Jason Merrill: I suppose it might be useful for a program to be able to express a degenerate case using a null template argument. I think allowing it would be harmless.

Notes from October 2004 meeting:

CWG decided that it would be desirable to allow null pointers as nontype template arguments, even though they are not representable in some current ABIs. There was some discussion over whether to allow a bare 0 to be used with a pointer nontype template parameter. The following case was decisive:

    template<int i> void foo();
    template<int* i> void foo();
    ...
    foo<0>();

The current wording of 14.3  temp.arg paragraph 7 disambiguates the function call in favor of the int version. If the null pointer conversion were allowed for pointer nontype template parameters, this case would become ambiguous, so it was decided to require a cast.

Proposed resolution (April, 2005):

1. In 14.3.2  temp.arg.nontype paragraph 1, insert the following after the third bullet:

   • a constant expression that evaluates to a null pointer value (4.10  conv.ptr); or

   • a constant expression that evaluates to a null member pointer value (4.11  conv.mem); or

2. Add the indicated text to the note in the second bullet of 14.3.2  temp.arg.nontype paragraph 5:

   [Note: In particular, neither the null pointer conversion (4.10  conv.ptr) nor the derived-to-base conversion (4.10  conv.ptr) are applied. Although 0 is a valid template-argument for a non-type template-parameter of integral type, it is not a valid template-argument for a non-type template-parameter of pointer type. However, (int*)0 is a valid template-argument for a non-type template-parameter of type “pointer to int.” —end note]

3. Replace the normative wording of 14.4  temp.type paragraph 1 with the following:

   Two template-ids refer to the same class or function if

   • their template-names refer to the same template, and
   • their corresponding type template-arguments are the same type, and
   • their corresponding non-type template-arguments of integral or enumeration type have identical values, and
   • their corresponding non-type template-arguments of pointer type refer to the same external object or function or are both the null pointer value, and
   • their corresponding non-type template-arguments of pointer-to-member type refer to the same class member or are both the null member pointer value, and
   • their corresponding non-type template-argumentss for template parameters of reference type refer to the same external object or function, and
   • their corresponding template template-arguments refer to the same template.

382. Allow typename outside of templates

[Voted into WP at October 2005 meeting.]

P. J. Plauger, among others, has noted that typename is hard to use, because in a given context it's either required or forbidden, and it's often hard to tell which. It would make life easier for programmers if typename could be allowed in places where it is not required, e.g., outside of templates.

Notes from the April 2003 meeting:

There was unanimity on relaxing this requirement on typename. The question was how much to relax it. Everyone agreed on allowing it on all qualified names, which is an easy fix (no syntax change required). But should it be allowed other places? P.J. Plauger said he'd like to see it allowed anywhere a type name is allowed, and that it could actually be a decades-late assist for the infamous "the ice is thin here" typedef problem noted in K&R I.

Proposed resolution (April 2003):

Replace the text at the start of 14.6  temp.res paragraph 3:

A qualified-id that refers to a type and in which the nested-name-specifier depends on a template-parameter (14.6.2  temp.dep) shall be prefixed by the keyword typename to indicate that the qualified-id denotes a type, forming an elaborated-type-specifier (7.1.5.3  dcl.type.elab).

With:

The keyword typename can only be applied to a qualified-id. A qualified-id that refers to a type and in which the nested-name-specifier depends on a template-parameter (14.6.2  temp.dep) shall be prefixed by the keyword typename to indicate that the qualified-id denotes a type, forming an elaborated-type-specifier (7.1.5.3  dcl.type.elab). If a qualified-id which has been prefixed by the keyword typename does not denote a type the program is ill-formed. [ Note: The keyword is only required on a qualified-id within a template declaration or definition in which the nested-name-specifier depends on a template-parameter. ]

Remove 14.6  temp.res paragraph 5:

The keyword typename shall only be used in template declarations and definitions, including in the return type of a function template or member function template, in the return type for the definition of a member function of a class template or of a class nested within a class template, and in the type-specifier for the definition of a static member of a class template or of a class nested within a class template. The keyword typename shall be applied only to qualified names, but those names need not be dependent. The keyword typename shall be used only in contexts in which dependent names can be used. This includes template declarations and definitions but excludes explicit specialization declarations and explicit instantiation declarations. The keyword typename is not permitted in a base-specifier or in a mem-initializer; in these contexts a qualified-id that depends on a template-parameter (temp.dep) is implicitly assumed to be a type name.

Note: the claim here that a qualified name preceded by typename forms an elaborated type specifier conflicts with the changes made in issue 254 (see N1376=02-0034), which introduces typename-specifier.

Notes from October 2003 meeting:

We considered whether typename should be allowed in more places, and decided we only wanted to allow it in qualified names (for now at least).

Core issue 254 changed elaborated-type-specifier to typename-specifier. It also changed 14.6  temp.res paragraph 5, which this proposed resolution deletes.

See also issue 468.

Proposed resolution (October, 2004):

1. Change 14.6  temp.res paragraph 3 as follows:

   A When a qualified-id that refers to a type and in which the nested-name-specifier depends on a template-parameter (14.6.2  temp.dep) is intended to refer to a type, it shall be prefixed by the keyword typename to indicate that the qualified-id denotes a type, forming a typename-specifier. If the qualified-id in a typename-specifier does not denote a type, the program is ill-formed.

2. Change 14.6  temp.res paragraph 5 as follows:

   The keyword typename shall only be used in template declarations and definitions, including in the return type of a function template or member function template, in the return type for the definition of a member function of a class template or of a class nested within a class template, and in the type-specifier for the definition of a static member of a class template or of a class nested within a class template. The keyword typename shall be applied only to qualified names, but those names need not be dependent. The keyword typename shall be used only in contexts in which dependent names can be used. This includes template declarations and definitions but excludes explicit specialization declarations and explicit instantiation declarations. A qualified name used as the name in a mem-initializer-id, a base-specifier, or an elaborated-type-specifier is implicitly assumed to name a type, without the use of the typename keyword. [Note: the typename keyword is not permitted by the syntax of these constructs. —end note]

447. Is offsetof type-dependent?

[Voted into WP at October 2005 meeting.]

As far as I can tell, the standard doesn't say whether "offsetof(...)" is type-dependent. In the abstract, it shouldn't be -- an "offsetof" expression is always of type "size_t". But the standard doesn't say to what the definition of the macro is, so I don't think one can deduce that it will always be considered non-dependent by a conforming compiler.

John Spicer: (1) I agree that you can't know if offsetof is dependent because you don't know what it expands to. (2) In principle, offsetof should be like sizeof -- it is value-dependent if its argument is type-dependent.

Mark Mitchell: I think we should say that: (a) offsetof is not type-dependent, and (b) offsetof is value dependent iff the first argument is type-dependent

Everyone is using slightly different builtins to implement this functionality, and I don't think that there's any guarantee that they're all behaving the same here.

Notes from the March 2004 meeting:

Note that any such requirement would be in the library section, not core.

Proposed resolution (October, 2004):

1. At the end of 14.6.2.2  temp.dep.expr paragraph 4, add after the list that ends with throw assignment-expression:

   [Note: For the standard library macro offsetof, see 18.1  lib.support.types. —end note]

2. At the end of 14.6.2.3  temp.dep.constexpr paragraph 2, add after the list that ends with sizeof(type-id):

   [Note: For the standard library macro offsetof, see 18.1  lib.support.types. —end note]

3. In 18.1  lib.support.types paragraph 4, replace

   The macro offsetof accepts a restricted set of type arguments in this International Standard. If type is not a POD structure or a POD union the results are undefined. The result of applying the offsetof macro to a field that is a static data member or a function member is undefined.

   with

   The macro offsetof(type, member-designator) accepts a restricted set of type arguments in this International Standard. If type is not a POD structure or a POD union (clause 9  class), the results are undefined. The expression offsetof(type, member-designator) is never type-dependent (14.6.2.2  temp.dep.expr) and it is value-dependent (14.6.2.3  temp.dep.constexpr) if and only if type is dependent. The result of applying the offsetof macro to a field that is a static data member or a function member is undefined.

   [Note: the original wording shown here reflects the resolutions of library issues 306 and 449.]

197. Issues with two-stage lookup of dependent names

[Voted into WP at October 2005 meeting.]

The example in 14.6  temp.res paragraph 9 is incorrect, according to 14.6.4.2  temp.dep.candidate . The example reads,

    void f(char);

    template <class T> void g(T t)
    {
        f(1);        // f(char);
        f(T(1));     // dependent
        f(t);        // dependent
        dd++;        // not dependent
                     // error: declaration for dd not found
    }

    void f(int);

    double dd;
    void h()
    {
        g(2);        // will cause one call of f(char) followed
                     // by two calls of f(int)
        g('a');      // will cause three calls of f(char)
    }

Notes from 04/00 meeting:

The core working group agreed that the example as written is incorrect and should be reformulated to use a class type instead of a fundamental type. It was also decided to open a new issue dealing more generally with Koenig lookup and fundamental types.

(See also issues 213 and 225.)

Proposed resolution (April, 2005):

Change the example in 14.6  temp.res paragraph 9 as follows:

    void f(char);

    template <class T> void g(T t)
    {
        f(1);        // f(char);
        f(T(1));     // dependent
        f(t);        // dependent
        dd++;        // not dependent
                     // error: declaration for dd not found
    }

    enum E { e };
    void f(intE);

    double dd;
    void h()
    {
        g(2e);       // will cause one call of f(char) followed
                     // by two calls of f(intE)
        g('a');      // will cause three calls of f(char)
    }

237. Explicit instantiation and base class members

[Voted into WP at October 2005 meeting.]

In 14.7.2  temp.explicit paragraph 7 we read:

The explicit instantiation of a class template specialization implies the instantiation of all of its members not previously explicitly specialized in the translation unit containing the explicit instantiation.

Is "member" intended to mean "non-inherited member?" If yes, maybe it should be clarified since 10  class.derived paragraph 1 says,

Unless redefined in the derived class, members of a base class are also considered to be members of the derived class.

Proposed resolution (October, 2004):

Fixed by the resolution of issue 470.

470. Instantiation of members of an explicitly-instantiated class template

[Voted into WP at October 2005 meeting.]

14.7.2  temp.explicit paragraph 7 says,

The explicit instantiation of a class template specialization implies the instantiation of all of its members not previously explicitly specialized in the translation unit containing the explicit instantiation.

It's not clear whether this “implied” instantiation is implicit or explicit instantiation. It makes a difference in cases like the following:

    template <typename T> struct foo {
        struct bar { };
    };

    template struct foo<int>;         // #1

    template struct foo<int>::bar;    // #2

If the instantiation of foo<int>::bar implied by #1 is implicit, the explicit instantiation in #2 is well-formed. Otherwise, #2 violates the requirement in 14.7  temp.spec that

No program shall explicitly instantiate any template more than once ... for a given set of template-arguments.

It's also unclear whether the implied instantiation applies only to direct members of the class template or to inherited members, as well.

John Spicer: I have always interpreted this as meaning only the members declared in the class, not those inherited from other classes. This is what EDG does, and appears to be what g++, Microsoft and Sun do, too. I also think this is the correct thing for the Standard to require. If I were to derive a class from a class in the standard library, an explicit instantiation of my class should not cause the explicit instantiation of things in the standard library (because the library might provide such explicit instantiations, thus causing my program to run afoul of the "can't instantiate more than once" rule).

Proposed resolution (October, 2004):

Change 14.7.2  temp.explicit paragraph 7 as follows:

The explicit instantiation of a class template specialization implies the instantiation of all also explicitly instantiates each of its members not (not including members inherited from base classes) whose definition is visible at the point of instantiation and that has not been previously explicitly specialized in the translation unit containing the explicit instantiation.

Issues with "WP" Status

362. Order of initialization in instantiation units

[Voted into WP at March 2004 meeting.]

Should this program do what its author obviously expects? As far as I can tell, the standard says that the point of instantiation for Fib<n-1>::Value is the same as the point of instantiation as the enclosing specialization, i.e., Fib<n>::Value. What in the standard actually says that these things get initialized in the right order?

  template<int n>
  struct Fib { static int Value; };

  template <>
  int Fib<0>::Value = 0;

  template <>
  int Fib<1>::Value = 1;

  template<int n>
  int Fib<n>::Value = Fib<n-1>::Value + Fib<n-2>::Value;

  int f ()
  {
    return Fib<40>::Value;
  }

John Spicer: My opinion is that the standard does not specify the behavior of this program. I thought there was a core issue related to this, but I could not find it. The issue that I recall proposed tightening up the static initialization rules to make more cases well defined.

Your comment about point of instantiation is correct, but I don't think that really matters. What matters is the order of execution of the initialization code at execution time. Instantiations don't really live in "translation units" according to the standard. They live in "instantiation units", and the handling of instantiation units in initialization is unspecified (which should probably be another core issue). See 2.1  lex.phases paragraph 8.

Notes from October 2002 meeting:

We discussed this and agreed that we really do mean the the order is unspecified. John Spicer will propose wording on handling of instantiation units in initialization.

Proposed resolution (April 2003):

TC1 contains the following text in 3.6.2  basic.start.init paragraph 1:

Objects with static storage duration defined in namespace scope in the same translation unit and dynamically initialized shall be initialized in the order in which their definition appears in the translation unit.

This was revised by issue 270 to read:

Dynamic initialization of an object is either ordered or unordered. Explicit specializations and definitions of class template static data members have ordered initialization. Other class template static data member instances have unordered initialization. Other objects defined in namespace scope have ordered initialization. Objects defined within a single translation unit and with ordered initialization shall be initialized in the order of their definitions in the translation unit. The order of initialization is unspecified for objects with unordered initialization and for objects defined in different translation units.

This addresses this issue but while reviewing this issue some additional changes were suggested for the above wording:

Dynamic initialization of an object is either ordered or unordered. Definitions of explicitly specialized Explicit specializations and definitions of class template static data members have ordered initialization. Other class template static data members (i.e., implicitly or explicitly instantiated specializations) instances have unordered initialization. Other objects defined in namespace scope have ordered initialization. Objects defined within a single translation unit and with ordered initialization shall be initialized in the order of their definitions in the translation unit. The order of initialization is unspecified for objects with unordered initialization and for objects defined in different translation units.

261. When is a deallocation function "used?"

[Moved to DR at October 2002 meeting.]

3.2  basic.def.odr paragraph 2 says that a deallocation function is "used" by a new-expression or delete-expression appearing in a potentially-evaluated expression. 3.2  basic.def.odr paragraph 3 requires only that "used" functions be defined.

This wording runs afoul of the typical implementation technique for polymorphic delete-expressions in which the deallocation function is invoked from the virtual destructor of the most-derived class. The problem is that the destructor must be defined, because it's virtual, and if it contains an implicit reference to the deallocation function, the deallocation function must also be defined, even if there are no relevant new-expressions or delete-expressions in the program.

For example:

        struct B { virtual ~B() { } };

        struct D: B {
            void operator delete(void*);
            ~D() { }
        };

Is it required that D::operator delete(void*) be defined, even if no B or D objects are ever created or deleted?

Suggested resolution: Add the words "or if it is found by the lookup at the point of definition of a virtual destructor (12.4  class.dtor)" to the specification in 3.2  basic.def.odr paragraph 2.

Notes from 04/01 meeting:

The consensus was in favor of requiring that any declared non-placement operator delete member function be defined if the destructor for the class is defined (whether virtual or not), and similarly for a non-placement operator new if a constructor is defined.

Proposed resolution (10/01):

In 3.2  basic.def.odr paragraph 2, add the indicated text:

An allocation or deallocation function for a class is used by a new expression appearing in a potentially-evaluated expression as specified in 5.3.4  expr.new and 12.5  class.free. A deallocation function for a class is used by a delete expression appearing in a potentially-evaluated expression as specified in 5.3.5  expr.delete and 12.5  class.free. A non-placement allocation or deallocation function for a class is used by the definition of a constructor of that class. A non-placement deallocation function for a class is 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 (12.4  class.dtor). [Footnote: An implementation is not required to call allocation and deallocation functions from constructors or destructors; however, this is a permissible implementation technique.]

289. Incomplete list of contexts requiring a complete type

[Moved to DR at October 2002 meeting.]

3.2  basic.def.odr paragraph 4 has a note listing the contexts that require a class type to be complete. It does not list use as a base class as being one of those contexts.

Proposed resolution (10/01):

In 3.2  basic.def.odr paragraph 4 add a new bullet at the end of the note as the next-to-last bullet:

• a class with a base class of type T is defined (10  class.derived), or

433. Do elaborated type specifiers in templates inject into enclosing namespace scope?

[Voted into WP at March 2004 meeting.]

Consider the following translation unit:

  template<class T> struct S {
    void f(union U*);  // (1)
  };
  template<class T> void S<T>::f(union U*) {}  // (2)
  U *p;  // (3)

Does (1) introduce U as a visible name in the surrounding namespace scope?

If not, then (2) could presumably be an error since the "union U" in that definition does not find the same type as the declaration (1).

If yes, then (3) is OK too. However, we have gone through much trouble to allow template implementations that do not pre-parse the template definitions, but requiring (1) to be visible would change that.

A slightly different case is the following:

  template<typename> void f() { union U *p; }
  U *q;  // Should this be valid?

Notes from October 2003 meeting:

There was consensus that example 1 should be allowed. (Compilers already parse declarations in templates; even MSVC++ 6.0 accepts this case.) The vote was 7-2.

Example 2, on the other hand, is wrong; the union name goes into a block scope anyway.

Proposed resolution:

In 3.3.1  basic.scope.pdecl change the second bullet of paragraph 5 as follows:

for an elaborated-type-specifier of the form

   class-key identifier

if the elaborated-type-specifier is used in the decl-specifier-seq or parameter-declaration-clause of a function defined in namespace scope, the identifier is declared as a class-name in the namespace that contains the declaration; otherwise, except as a friend declaration, the identifier is declared in the smallest non-class, non-function-prototype scope that contains the declaration. [Note: These rules also apply within templates.] [Note: ...]

432. Is injected class name visible in base class specifier list?

[Voted into WP at March 2004 meeting.]

Consider the following example (inspired by a question from comp.lang.c++.moderated):

  template<typename> struct B {};
  template<typename T> struct D: B<D> {};

Most (all?) compilers reject this code because D is handled as a template name rather than as the injected class name.

9  class/2 says that the injected class name is "inserted into the scope of the class."

3.3.6  basic.scope.class/1 seems to be the text intended to describe what "scope of a class" means, but it assumes that every name in that scope was introduced using a "declarator". For an implicit declaration such as the injected-class name it is not clear what that means.

So my questions:

1. Should the injected class name be available in the base class specifiers?

   John Spicer: I do not believe the injected class name should be available in the base specifier. I think the semantics of injected class names should be as if a magic declaration were inserted after the opening "{" of the class definition. The injected class name is a member of the class and members don't exist at the point where the base specifiers are scanned.

2. Do you agree the wording should be clarified whatever the answer to the first question?

   John Spicer: I believe the 3.3.6  basic.scope.class wording should be updated to reflect the fact that not all names come from declarators.

Notes from October 2003 meeting:

We agree with John Spicer's suggested answers above.

Proposed Resolution (October 2003):

The answer to question 1 above is No and no change is required.

For question 1, change 3.3.6  basic.scope.class paragraph 1 rule 1 to:

1) The potential scope of a name declared in a class consists not only of the declarative region following the name's point of declaration declarator, but also of all function bodies, default arguments, and constructor ctor-initializers in that class (including such things in nested classes). The point of declaration of an injected-class-name (clause 9  class) is immediately following the opening brace of the class definition.

(Note that this change overlaps a change in issue 417.)

Also change 3.3.1  basic.scope.pdecl by adding a new paragraph 8 for the injected-class-name case:

The point of declaration for an injected-class-name (clause 9  class) is immediately following the opening brace of the class definition.

Alternatively this paragraph could be added after paragraph 5 and before the two note paragraphs (i.e. it would become paragraph 5a).

139. Error in friend lookup example

[Moved to DR at 10/01 meeting.]

The example in 3.4.1  basic.lookup.unqual paragraph 3 is incorrect:

    typedef int f;
    struct A {
        friend void f(A &);
        operator int();
        void g(A a) {
            f(a);
        }
    };

One possible repair of the example would be to make f a class with a constructor taking either A or int as its parameter.

(See also issues 95, 136, 138, 143, 165, and 166.)

Proposed resolution (04/01):

1. Change the example in 3.4.1  basic.lookup.unqual paragraph 3 to read:

       typedef int f;
       namespace N {
           struct A {
               friend int f(A &);
               operator int();
               void g(A a) {
                   int i = f(a);
                         // f is the typedef, not the friend function:
                         // equivalent to int(a)
               }
           };
       }
   

2. Delete the sentence immediately following the example:

   The expression f(a) is a cast-expression equivalent to int(a).

143. Friends and Koenig lookup

[Moved to DR at 4/02 meeting.]

Paragraphs 1 and 2 of 3.4.2  basic.lookup.argdep say, in part,

When an unqualified name is used as the postfix-expression in a function call (5.2.2  expr.call )... namespace-scope friend function declarations (11.4  class.friend ) not otherwise visible may be found... the set of declarations found by the lookup of the function name [includes] the set of declarations found in the... classes associated with the argument types.

Consider the following example:

    namespace A {
	class S;
    };
    namespace B {
	void f(A::S);
    };
    namespace A {
	class S {
	    int i;
	    friend void B::f(S);
	};
    }
    void g() {
	A::S s;
	f(s); // should find B::f(A::S)
    }

Another interpretation is that, instead of finding the friend declarations in associated classes, one only looks for namespace-scope functions, visible or invisible, in the namespaces of which the the associated classes are members; the only use of the friend declarations in the associated classes is to validate whether an invisible function declaration came from an associated class or not and thus whether it should be included in the overload set or not. By this interpretation, the call f(s) in the example will fail, because B::f(A::S) is not a member of namespace A and thus is not found by the lookup.

Notes from 10/99 meeting: The second interpretation is correct. The wording should be revised to make clear that Koenig lookup works by finding "invisible" declarations in namespace scope and not by finding friend declarations in associated classes.

Proposed resolution (04/01): The "associated classes" are handled adequately under this interpretation by 3.4.2  basic.lookup.argdep paragraph 3, which describes the lookup in the associated namespaces as including the friend declarations from the associated classes. Other mentions of the associated classes should be removed or qualified to avoid the impression that there is a lookup in those classes:

1. In 3.4.2  basic.lookup.argdep, change

   When an unqualified name is used as the postfix-expression in a function call (5.2.2  expr.call), other namespaces not considered during the usual unqualified lookup (3.4.1  basic.lookup.unqual) may be searched, and namespace-scope friend function declarations (11.4  class.friend) not otherwise visible may be found.

   to

   When an unqualified name is used as the postfix-expression in a function call (5.2.2  expr.call), other namespaces not considered during the usual unqualified lookup (3.4.1  basic.lookup.unqual) may be searched, and in those namespaces, namespace-scope friend function declarations (11.4  class.friend) not otherwise visible may be found.

2. In 3.4.2  basic.lookup.argdep paragraph 2, delete the words and classes in the following two sentences:

   If the ordinary unqualified lookup of the name finds the declaration of a class member function, the associated namespaces and classes are not considered. Otherwise the set of declarations found by the lookup of the function name is the union of the set of declarations found using ordinary unqualified lookup and the set of declarations found in the namespaces and classes associated with the argument types.

(See also issues 95, 136, 138, 139, 165, 166, and 218.)

403. Reference to a type as a template-id

[Voted into WP at March 2004 meeting.]

Spun off from issue 384.

3.4.2  basic.lookup.argdep says:

If T is a template-id, its associated namespaces and classes are the namespace in which the template is defined; for member templates, the member template's class; the namespaces and classes associated with the types of the template arguments provided for template type parameters (excluding template template parameters); the namespaces in which any template template arguments are defined; and the classes in which any member templates used as template template arguments are defined. [Note: non-type template arguments do not contribute to the set of associated namespaces. ]

Proposed Resolution (October 2003):

In 3.4.2  basic.lookup.argdep, paragraph 2, bullet 8, replace

If T is a template-id ...

If T is a class template specialization ...

298. T::x when T is cv-qualified

[Voted into WP at April 2003 meeting.]

Can a typedef T to a cv-qualified class type be used in a qualified name T::x?

    struct A { static int i; };
    typedef const A CA;
    int main () {
      CA::i = 0;  // Okay?
    }

Suggested answer: Yes. All the compilers I tried accept the test case.

Proposed resolution (10/01):

In 3.4.3.1  class.qual paragraph 1 add the indicated text:

If the nested-name-specifier of a qualified-id nominates a class, the name specified after the nested-name-specifier is looked up in the scope of the class (10.2  class.member.lookup), except for the cases listed below. The name shall represent one or more members of that class or of one of its base classes (clause 10  class.derived). If the class-or-namespace-name of the nested-name-specifier names a cv-qualified class type, it nominates the underlying class (the cv-qualifiers are ignored).

Notes from 4/02 meeting:

There is a problem in that class-or-namespace-name does not include typedef names for cv-qualified class types. See 7.1.3  dcl.typedef paragraph 4:

Argument and text removed from proposed resolution (October 2002):

7.1.3  dcl.typedef paragraph 5:

Here's a good question: in this example, should X be used as a name-for-linkage-purposes (FLP name)?

  typedef class { } const X;

Because a type-qualifier is parsed as a decl-specifier, it isn't possible to declare cv-qualified and cv-unqualified typedefs for a type in a single declaration. Also, of course, there's no way to declare a typedef for the cv-unqualified version of a type for which only a cv-qualified version has a name. So, in the above example, if X isn't used as the FLP name, then there can be no FLP name. Also note that a FLP name usually represents a parameter type, where top-level cv-qualifiers are usually irrelevant anyway.

Data points: for the above example, Microsoft uses X as the FLP name; GNU and EDG do not.

My recommendation: for consistency with the direction we're going on this issue, for simplicity of description (e.g., "the first class-name declared by the declaration"), and for (very slightly) increased utility, I think Microsoft has this right.

If the typedef declaration defines an unnamed class type (or enum type), the first typedef-name declared by the declaration to be have that class type (or enum type) or a cv-qualified version thereof is used to denote the class type (or enum type) for linkage purposes only (3.5  basic.link). [Example: ...

Proposed resolution (October 2002):

3.4.4  basic.lookup.elab paragraphs 2 and 3:

This sentence is deleted twice:

... If this name lookup finds a typedef-name, the elaborated-type-specifier is ill-formed. ...

Note that the above changes are included in N1376 as part of the resolution of issue 245.

5.1  expr.prim paragraph 7:

This is only a note, and it is at least incomplete (and quite possibly inaccurate), despite (or because of) its complexity. I propose to delete it.

... [Note: a typedef-name that names a class is a class-name (9.1  class.name). Except as the identifier in the declarator for a constructor or destructor definition outside of a class member-specification (12.1  class.ctor, 12.4  class.dtor), a typedef-name that names a class may be used in a qualified-id to refer to a constructor or destructor. ]

7.1.3  dcl.typedef paragraph 4:

My first choice would have been to make this the primary statement about the equivalence of typedef-name and class-name, since the equivalence comes about as a result of a typedef declaration. Unfortunately, references to class-name point to 9.1  class.name, so it would seem that the primary statement should be there instead. To avoid the possiblity of conflicts in the future, I propose to make this a note.

[Note: A typedef-name that names a class type, or a cv-qualified version thereof, is also a class-name (9.1  class.name). If a typedef-name is used following the class-key in an elaborated-type-specifier (7.1.5.3  dcl.type.elab), or in the class-head of a class declaration (9  class), or is used as the identifier in the declarator for a constructor or destructor declaration (12.1  class.ctor, 12.4  class.dtor), to identify the subject of an elaborated-type-specifier (7.1.5.3  dcl.type.elab), class declaration (clause 9  class), constructor declaration (12.1  class.ctor), or destructor declaration (12.4  class.dtor), the program is ill-formed. ] [Example: ...

7.1.5.3  dcl.type.elab paragraph 2:

This is the only remaining (normative) statement that a typedef-name can't be used in an elaborated-type-specifier. The reference to template type-parameter is deleted by the resolution of issue 283.

... If the identifier resolves to a typedef-name or a template type-parameter, the elaborated-type-specifier is ill-formed. [Note: ...

8  dcl.decl grammar rule declarator-id:

When I looked carefully into the statement of the rule prohibiting a typedef-name in a constructor declaration, it appeared to me that this grammar rule (inadvertently?) allows something that's always forbidden semantically.

declarator-id:
id-expression
::_opt nested-name-specifier_opt type-name class-name

9.1  class.name paragraph 5:

Unlike the prohibitions against appearing in an elaborated-type-specifier or constructor or destructor declarator, each of which was expressed more than once, the prohibition against a typedef-name appearing in a class-head was previously stated only in 7.1.3  dcl.typedef. It seems to me that that prohibition belongs here instead. Also, it seems to me important to clarify that a typedef-name that is a class-name is still a typedef-name. Otherwise, the various prohibitions can be argued around easily, if perversely ("But that isn't a typedef-name, it's a class-name; it says so right there in 9.1  class.name.")

A typedef-name (7.1.3  dcl.typedef) that names a class type or a cv-qualified version thereof is also a class-name, but shall not be used in an elaborated-type-specifier; see also 7.1.3  dcl.typedef. as the identifier in a class-head.

12.1  class.ctor paragraph 3:

The new nonterminal references are needed to really nail down what we're talking about here. Otherwise, I'm just eliminating redundancy. (A typedef-name that doesn't name a class type is no more valid here than one that does.)

A typedef-name that names a class is a class-name (7.1.3  dcl.typedef); however, a A typedef-name that names a class shall not be used as the identifier class-name in the declarator declarator-id for a constructor declaration.

12.4  class.dtor paragraph 1:

The same comments apply here as to 12.1  class.ctor.

... A typedef-name that names a class is a class-name (7.1.3); however, a A typedef-name that names a class shall not be used as the identifier class-name following the ~ in the declarator for a destructor declaration.

318. struct A::A should not name the constructor of A

[Voted into WP at April 2003 meeting.]

A use of an injected-class-name in an elaborated-type-specifier should not name the constructor of the class, but rather the class itself, because in that context we know that we're looking for a type. See issue 147.

Proposed Resolution (revised October 2002):

This clarifies the changes made in the TC for issue 147.

In 3.4.3.1  class.qual paragraph 1a replace:

If the nested-name-specifier nominates a class C, and the name specified after the nested-name-specifier, when looked up in C, is the injected class name of C (clause 9  class), the name is instead considered to name the constructor of class C.

with

In a lookup in which the constructor is an acceptable lookup result, if the nested-name-specifier nominates a class C and the name specified after the nested-name-specifier, when looked up in C, is the injected class name of C (clause 9  class), the name is instead considered to name the constructor of class C. [Note: For example, the constructor is not an acceptable lookup result in an elaborated type specifier so the constructor would not be used in place of the injected class name.]

Note that issue 263 updates a part of the same paragraph.

Append to the example:

  struct A::A a2;  // object of type A

400. Using-declarations and the "struct hack"

[Voted into WP at March 2004 meeting.]

Consider this code:

  struct A { int i; struct i {}; };
  struct B { int i; struct i {}; };
  struct D : public A, public B { using A::i; void f (); };
  void D::f () { struct i x; }

I can't find anything in the standard that says definitively what this means. 7.3.3  namespace.udecl says that a using-declaration shall name "a member of a base class" -- but here we have two members, the data member A::i and the class A::i.

Personally, I'd find it more attractive if this code did not work. I'd like "using A::i" to mean "lookup A::i in the usual way and bind B::i to that", which would mean that while "i = 3" would be valid in D::f, "struct i x" would not be. However, if there were no A::i data member, then "A::i" would find the struct and the code in D::f would be valid.

John Spicer: I agree with you, but unfortunately the standard committee did not.

I remembered that this was discussed by the committee and that a resolution was adopted that was different than what I hoped for, but I had a hard time finding definitive wording in the standard.

I went back though my records and found the paper that proposed a resolution and the associated committee motion that adopted the proposed resolution The paper is N0905, and "option 1" from that paper was adopted at the Stockholm meeting in July of 1996. The resolution is that "using A::i" brings in everything named i from A.

3.4.3.2  namespace.qual paragraph 2 was modified to implement this resolution, but interestingly that only covers the namespace case and not the class case. I think the class case was overlooked when the wording was drafted. A core issue should be opened to make sure the class case is handled properly.

Notes from April 2003 meeting:

This is related to issue 11. 7.3.3  namespace.udecl paragraph 10 has an example for namespaces.

Proposed resolution (October 2003):

Add a bullet to the end of 3.4.3.1  class.qual paragraph 1:

• the lookup for a name specified in a using-declaration (7.3.3  namespace.udecl) also finds class or enumeration names hidden within the same scope (3.3.7  basic.scope.hiding).

Change the beginning of 7.3.3  namespace.udecl paragraph 4 from

A using-declaration used as a member-declaration shall refer to a member of a base class of the class being defined, shall refer to a member of an anonymous union that is a member of a base class of the class being defined, or shall refer to an enumerator for an enumeration type that is a member of a base class of the class being defined.

to

In a using-declaration used as a member-declaration, the nested-name-specifier shall name a base class of the class being defined. Such a using-declaration introduces the set of declarations found by member name lookup (10.2  class.member.lookup, 3.4.3.1  class.qual).

245. Name lookup in elaborated-type-specifiers

[Voted into WP at April 2003 meeting.]

I have some concerns with the description of name lookup for elaborated type specifiers in 3.4.4  basic.lookup.elab:

1. Paragraph 2 has some parodoxical statements concerning looking up names that are simple identifers:

   If the elaborated-type-specifier refers to an enum-name and this lookup does not find a previously declared enum-name, the elaborated-type-specifier is ill-formed. If the elaborated-type-specifier refers to an [sic] class-name and this lookup does not find a previously declared class-name... the elaborated-type-specifier is a declaration that introduces the class-name as described in 3.3.1  basic.scope.pdecl."

   It is not clear how an elaborated-type-specifier can refer to an enum-name or class-name given that the lookup does not find such a name and that class-name and enum-name are not part of the syntax of an elaborated-type-specifier.

2. The second sentence quoted above seems to suggest that the name found will not be used if it is not a class name. typedef-name names are ill-formed due to the sentence preceding the quote. If lookup finds, for instance, an enum-name then a new declaration will be created. This differs from C, and from the enum case, and can have surprising effects:

       struct S {
          enum E {
              one = 1
          };
          class E* p;     // declares a global class E?
       };
   

   Was this really the intent? If this is the case then some more work is needed on 3.4.4  basic.lookup.elab. Note that the section does not make finding a type template formal ill-formed, as is done in 7.1.5.3  dcl.type.elab. I don't see anything that makes a type template formal name a class-name. So the example in 7.1.5.3  dcl.type.elab of friend class T; where T is a template type formal would no longer be ill-formed with this interpretation because it would declare a new class T.

(See also issue 254.)

Notes from the 4/02 meeting:

This will be consolidated with the changes for issue 254. See also issue 298.

Proposed resolution (October 2002):

As given in N1376=02-0034. Note that the inserts and strikeouts in that document do not display correctly in all browsers; <del> --> <strike> and <ins> --> <b>, and the similar changes for the closing delimiters, seem to do the trick.

254. Definitional problems with elaborated-type-specifiers

[Voted into WP at April 2003 meeting.]

1. The text in 3.4.4  basic.lookup.elab paragraph 2 twice refers to the possibility that an elaborated-type-specifier might have the form

           class-key identifier ;
   

   However, the grammar for elaborated-type-specifier does not include a semicolon.

2. In both 3.4.4  basic.lookup.elab and 7.1.5.3  dcl.type.elab, the text asserts that an elaborated-type-specifier that refers to a typedef-name is ill-formed. However, it is permissible for the form of elaborated-type-specifier that begins with typename to refer to a typedef-name.

   This problem is the result of adding the typename form to the elaborated-type-name grammar without changing the verbiage correspondingly. It could be fixed either by updating the verbiage or by moving the typename syntax into its own production and referring to both nonterminals when needed.

(See also issue 180. If this issue is resolved in favor of a separate nonterminal in the grammar for the typename forms, the wording in that issue's resolution must be changed accordingly.)

Notes from 04/01 meeting:

The consensus was in favor of moving the typename forms out of the elaborated-type-specifier grammar.

Notes from the 4/02 meeting:

This will be consolidated with the changes for issue 245.

Proposed resolution (October 2002):

As given in N1376=02-0034.

381. Incorrect example of base class member lookup

[Voted into WP at October 2004 meeting.]

The example in 3.4.5  basic.lookup.classref paragraph 4 is wrong (see 11.2  class.access.base paragraph 5; the cast to the naming class can't be done) and needs to be corrected. This was noted when the final version of the algorithm for issue 39 was checked against it.

Proposed Resolution (October 2003):

Remove the entire note at the end of 3.4.5  basic.lookup.classref paragraph 4, including the entire example.

216. Linkage of nameless class-scope enumeration types

[Moved to DR at 10/01 meeting.]

A name having namespace scope has external linkage if it is the name of

• [...]
• a named enumeration (7.2  dcl.enum), or an unnamed enumeration defined in a typedef declaration in which the enumeration has the typedef name for linkage purposes (7.1.3  dcl.typedef)

    typedef enum { e1 } *PE;
    void f(PE) {}  // Cannot declare a function (with linkage) using a 
		   // type with no linkage.

However, the same prohibition was not made for class scope types. Indeed, 3.5  basic.link paragraph 5 says:

In addition, a member function, static data member, class or enumeration of class scope has external linkage if the name of the class has external linkage.

That allows for:

    struct S {
       typedef enum { e1 } *MPE;
       void mf(MPE) {}
    };

My guess is that this is an unintentional consequence of 3.5  basic.link paragraph 5, but I would like confirmation on that.

Proposed resolution:

Change text in 3.5  basic.link paragraph 5 from:

In addition, a member function, static data member, class or enumeration of class scope has external linkage if the name of the class has external linkage.

In addition, a member function, a 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 (7.1.3  dcl.typedef), has external linkage if the name of the class has external linkage.

319. Use of names without linkage in declaring entities with linkage

[Voted into WP at October 2004 meeting.]

According to 3.5  basic.link paragraph 8, "A name with no linkage ... shall not be used to declare an entity with linkage." This would appear to rule out code such as:

  typedef struct {
    int i;
  } *PT;
  extern "C" void f(PT);

  static enum { a } e;

See issue 132, which dealt with a closely related issue.

Andrei Iltchenko submitted the same issue via comp.std.c++ on 17 Dec 2001:

Paragraph 8 of Section 3.5  basic.link contains the following sentences: "A name with no linkage shall not be used to declare an entity with linkage. If a declaration uses a typedef name, it is the linkage of the type name to which the typedef refers that is considered."

The problem with this wording is that it doesn't cover cases where the type to which a typedef-name refers has no name. As a result it's not clear whether, for example, the following program is well-formed:

#include <vector>

int  main()
{
   enum  {   sz = 6u   };
   typedef int  (* aptr_type)[sz];
   typedef struct  data  {
      int   i,  j;
   }  * elem_type;
   std::vector<aptr_type>   vec1;
   std::vector<elem_type>   vec2;
}

Suggested resolution:

My feeling is that the rules for whether or not a typedef-name used in a declaration shall be treated as having or not having linkage ought to be modelled after those for dependent types, which are explained in 14.6.2.1  temp.dep.type.

Add the following text at the end of Paragraph 8 of Section 3.5  basic.link and replace the following example:

In case of the type referred to by a typedef declaration not having a name, the newly declared typedef-name has linkage if and only if its referred type comprises no names of no linkage excluding local names that are eligible for appearance in an integral constant-expression (5.19  expr.const). [Note: if the referred type contains a typedef-name that does not denote an unnamed class, the linkage of that name is established by the recursive application of this rule for the purposes of using typedef names in declarations.] [Example:

  void f()
  {
     struct A { int x; };        // no linkage
     extern A a;                 // ill-formed
     typedef A Bl
     extern B b;                 // ill-formed

     enum  {   sz = 6u   };
     typedef int  (* C)[sz];     // C has linkage because sz can
                                 // appear in a constant expression
  }

--end example.]

Additional issue (13 Jan 2002, from Andrei Iltchenko):

Paragraph 2 of Section 14.3.1  temp.arg.type is inaccurate and unnecessarily prohibits a few important cases; it says "A local type, a type with no linkage, an unnamed type or a type compounded from any of these types shall not be used as a template-argument for a template-parameter." The inaccuracy stems from the fact that it is not a type but its name that can have a linkage.

For example based on the current wording of 14.3.1  temp.arg.type, the following example is ill-formed.

  #include <vector>
  struct  data  {
    int   i,  j;
  };
  int  main()
  {
    enum  {   sz = 6u   };
    std::vector<int(*)[sz]>   vec1; // The types 'int(*)[sz]' and 'data*'
    std::vector<data*>        vec2; // have no names and are thus illegal
                                    // as template type arguments.
  }

Suggested resolution:

Replace the whole second paragraph of Section 14.3.1  temp.arg.type with the following wording:

A type whose name does not have a linkage or a type compounded from any such type shall not be used as a template-argument for a template-parameter. In case of a type T used as a template type argument not having a name, T constitutes a valid template type argument if and only if the name of an invented typedef declaration referring to T would have linkage; see 3.5. [Example:

  template <class T> class X { /* ... */ };
  void f()
  {
    struct S { /* ... */ };
    enum  {   sz = 6u   };

    X<S> x3;                     // error: a type name with no linkage
                                 // used as template-argument
    X<S*> x4;                    // error: pointer to a type name with
                                 // no linkage used as template-argument
    X<int(*)[sz]> x5;            // OK: since the name of typedef int
                                 // (*pname)[sz] would have linkage
  }

--end example] [Note: a template type argument may be an incomplete type (3.9  basic.types).]

Proposed resolution:

This is resolved by the changes for issue 389. The present issue was moved back to Review status in February 2004 because 389 was moved back to Review.

389. Unnamed types in entities with linkage

[Voted into WP at October 2004 meeting.]

3.5  basic.link paragraph 8 says (among other things):

A name with no linkage (notably, the name of a class or enumeration declared in a local scope (3.3.2  basic.scope.local)) shall not be used to declare an entity with linkage. If a declaration uses a typedef name, it is the linkage of the type name to which the typedef refers that is considered.

I would expect this to catch situations such as the following:

  // File 1:
  typedef struct {} *UP;
  void f(UP) {}

  // File 2:
  typedef struct {} *UP; // Or: typedef struct {} U, *UP;
  void f(UP);

The problem here is that most implementations must generate the same mangled name for "f" in two translation units. The quote from the standard above isn't quite clear, unfortunately: There is no type name to which the typedef refers.

A related situation is the following:

  enum { no, yes } answer;

Finally, these problems are much less relevant when declaring names with internal linkage. For example, I would expect there to be few problems with:

  typedef struct {} *UP;
  static void g(UP);

I recently tried to interpret 3.5  basic.link paragraph 8 with the assumption that types with no names have no linkage. Surprisingly, this resulted in many diagnostics on variable declarations (mostly like "answer" above).

I'm pretty sure the standard needs clarifying words in this matter, but which way should it go?

See also issue 319.

Notes from April 2003 meeting:

There was agreement that this check is not needed for variables and functions with extern "C" linkage, and a change there is desirable to allow use of legacy C headers. The check is also not needed for entities with internal linkage, but there was no strong sentiment for changing that case.

We also considered relaxing this requirement for extern "C++" variables but decided that we did not want to change that case.

We noted that if extern "C" functions are allowed an additional check is needed when such functions are used as arguments in calls of function templates. Deduction will put the type of the extern "C" function into the type of the template instance, i.e., there would be a need to mangle the name of an unnamed type. To plug that hole we need an additional requirement on the template created in such a case.

Proposed resolution (April 2003, revised slightly October 2003 and March 2004):

In 3.5  basic.link paragraph 8, change

A name with no linkage (notably, the name of a class or enumeration declared in a local scope (3.3.2  basic.scope.local)) shall not be used to declare an entity with linkage. If a declaration uses a typedef name, it is the linkage of the type name to which the typedef refers that is considered.

to

A type is said to have linkage if and only if

• it is a class or enumeration type that is named (or has a name for linkage purposes (7.1.3  dcl.typedef)) and the name has linkage; or
• it is a specialization of a class template (14  temp) [Footnote: a class template always has external linkage, and the requirements of 14.3.1  temp.arg.type and 14.3.2  temp.arg.nontype ensure that the template arguments will also have appropriate linkage]; or
• it is a fundamental type (3.9.1  basic.fundamental); or
• it is a compound type (3.9.2  basic.compound) other than a class or enumeration, compounded exclusively from types that have linkage; or
• it is a cv-qualified (3.9.3  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 linkage, unless the variable or function has extern "C" linkage (7.5  dcl.link). [Note: in other words, a type without linkage contains a class or enumeration that cannot be named outside of 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 is thus not permitted. 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.]

Change 14.3.1  temp.arg.type paragraph 2 from (note: this is the wording as updated by issue 62)

The following types shall not be used as a template-argument for a template type-parameter:

• a type whose name has no linkage
• an unnamed class or enumeration type that has no name for linkage purposes (7.1.3  dcl.typedef)
• a cv-qualified version of one of the types in this list
• a type created by application of declarator operators to one of the types in this list
• a function type that uses one of the types in this list

to

A type without linkage (3.5  basic.link) shall not be used as a template-argument for a template type-parameter.

Once this issue is ready, issue 319 should be moved back to ready as well.

270. Order of initialization of static data members of class templates

[Moved to DR at 4/02 meeting.]

The Standard does not appear to address how the rules for order of initialization apply to static data members of class templates.

Suggested resolution: Add the following verbiage to either 3.6.2  basic.start.init or 9.4.2  class.static.data:

Initialization of static data members of class templates shall be performed during the initialization of static data members for the first translation unit to have static initialization performed for which the template member has been instantiated. This requirement shall apply to both the static and dynamic phases of initialization.

Notes from 04/01 meeting:

Enforcing an order of initialization on static data members of class templates will result in substantial overhead on access to such variables. The problem is that the initialization be required as the result of instantiation in a function used in the initialization of a variable in another translation unit. In current systems, the order of initialization of static data data members of class templates is not predictable. The proposed resolution is to state that the order of initialization is undefined.

Proposed resolution (04/01, updated slightly 10/01):

Replace the following sentence in 3.6.2  basic.start.init paragraph 1:

Objects with static storage duration defined in namespace scope in the same translation unit and dynamically initialized shall be initialized in the order in which their definition appears in the translation unit.

with

Dynamic initialization of an object is either ordered or unordered. Explicit specializations and definitions of class template static data members have ordered initialization. Other class template static data member instances have unordered initialization. Other objects defined in namespace scope have ordered initialization. Objects defined within a single translation unit and with ordered initialization shall be initialized in the order of their definitions in the translation unit. The order of initialization is unspecified for objects with unordered initialization and for objects defined in different translation units.

Note that this wording is further updated by issue 362.

Note (07/01):

Brian McNamara argues against the proposed resolution. The following excerpt captures the central point of a long message on comp.std.c++:

I have a class for representing linked lists which looks something like

    template <class T>
    class List {
       ...  static List<T>* sentinel; ...
    };
 
    template <class T>
    List<T>* List<T>::sentinel( new List<T> ); // static member definition

The sentinel list node is used to represent "nil" (the null pointer cannot be used with my implementation, for reasons which are immaterial to this discussion). All of the List's non-static member functions and constructors depend upon the value of the sentinel. Under the proposed resolution for issue #270, Lists cannot be safely instantiated before main() begins, as the sentinel's initialization is "unordered".

(Some readers may propose that I should use the "singleton pattern" in the List class. This is undesirable, for reasons I shall describe at the end of this post at the location marked "[*]". For the moment, indulge me by assuming that "singleton" is not an adequate solution.)

Though this is a particular example from my own experience, I believe it is representative of a general class of examples. It is common to use static data members of a class to represent the "distinguished values" which are important to instances of that class. It is imperative that these values be initialized before any instances of the class are created, as the instances depend on the values.

In a comp.std.c++ posting on 28 Jul 2001, Brian McNamara proposes the following alternative resolution:

Replace the following sentence in 3.6.2  basic.start.init paragraph 1:

Objects with static storage duration defined in namespace scope in the same translation unit and dynamically initialized shall be initialized in the order in which their definition appears in the translation unit.

Objects with static storage duration defined in namespace scope shall be initialized in the order described below.

Dynamic initialization is either ordered or quasi-ordered. Explicit specializations of class template static data members have ordered initialization. Other class template static data member instances have quasi-ordered initialization. All other objects defined in namespace scope have ordered initialization. The order of initialization is specified as follows:

• Objects that are defined within a single translation unit and that have ordered initialization shall be initialized in the order of their definitions in the translation unit.
• Objects that are defined only within a single translation unit and that have quasi-ordered initialization shall also be initialized in the order of their definitions in the translation unit -- that is, as though these objects had ordered initialization.
• Objects that are defined within multiple translation units (which, therefore, must have quasi-ordered initialization) shall be initialized as follows: in exactly one translation unit (which one is unspecified), the object shall be treated as though it has ordered initialization; in the other translation units which define the object, the object will be initialized before all other objects that have ordered initialization in those translation units.
• For any two objects, "X" and "Y", with static storage duration and defined in namespace scope, if the previous bullets do not imply a relationship for the initialization ordering between "X" and "Y", then the relative initialization order of these objects is unspecified.

[ Note: The intention is that translation units can each be compiled separately with no knowledge of what objects may be re-defined in other translation units. Each translation unit can contain a method which initializes all objects (both quasi-ordered and ordered) as though they were ordered. When these translation units are linked together to create an executable program, all of these objects can be initialized by simply calling the initialization methods (one from each translation unit) in any order. Quasi-ordered objects require some kind of guard to ensure that they are not initialized more than once (the first attempt to initialize such an object should succeed; any subsequent attempts should simply be ignored). ]

Erwin Unruh replies: There is a point which is not mentioned with this posting. It is the cost for implementing the scheme. It requires that each static template variable is instantiated in ALL translation units where it is used. There has to be a flag for each of these variables and this flag has to be checked in each TU where the instantiation took place.

I would reject this idea and stand with the proposed resolution of issue 270.

There just is no portable way to ensure the "right" ordering of construction.

Notes from 10/01 meeting:

The Core Working Group reaffirmed its previous decision.

441. Ordering of static reference initialization

[Voted into WP at April 2005 meeting.]

I have a couple of questions about 3.6.2  basic.start.init, "Initialization of non-local objects." I believe I recall some discussion of related topics, but I can't find anything relevant in the issues list.

The first question arose when I discovered that different implementations treat reference initialization differently. Consider, for example, the following (namespace-scope) code:

  int i;
  int& ir = i;
  int* ip = &i;

Zero-initialization and initialization with a constant expression are collectively called static initialization; all other initialization is dynamic initialization.

However, that does not mean that both ir and ip must be initialized at the same time:

Objects of POD types (3.9) with static storage duration initialized with constant expressions (5.19) shall be initialized before any dynamic initialization takes place.

Because "int&" is not a POD type, there is no requirement that it be initialized before dynamic initialization is performed, and implementations differ in this regard. Using a function called during dynamic initialization to print the values of "ip" and "&ir", I found that g++, Sun, HP, and Intel compilers initialize ir before dynamic initialization and the Microsoft compiler does not. All initialize ip before dynamic initialization. I believe this is conforming (albeit inconvenient :-) behavior.

So, my first question is whether it is intentional that a reference of static duration, initialized with a reference constant expression, need not be initialized before dynamic initialization takes place, and if so, why?

The second question is somewhat broader. As 3.6.2  basic.start.init is currently worded, it appears that there are no requirements on when ir is initialized. In fact, there is a whole category of objects -- non-POD objects initialized with a constant expression -- for which no ordering is specified. Because they are categorized as part of "static initialization," they are not subject to the requirement that they "shall be initialized in the order in which their definition appears in the translation unit." Because they are not POD types, they are not required to be initialized before dynamic initialization occurs. Am I reading this right?

My preference would be to change 3.6.2  basic.start.init paragraph 1 so that 1) references are treated like POD objects with respect to initialization, and 2) "static initialization" applies only to POD objects and references. Here's some sample wording to illustrate:

Suggested resolution:

Objects with static storage duration (3.7.1) shall be zero-initialized (8.5) before any other initialization takes place. Initializing a reference, or an object of POD type, of static storage duration with a constant expression (5.19) is called constant initialization. Together, zero-initialization and constant initialization are called static initialization; all other initialization is dynamic initialization. Static initialization shall be performed before any dynamic initialization takes place. [Remainder unchanged.]

Proposed Resolution:

Change 3.6.2  basic.start.init paragraph 1 as follows:

Objects with static storage duration (3.7.1) shall be zero-initialized (8.5) before any other initialization takes place. Initializing a reference, or an object of POD type, of static storage duration with a constant expression (5.19) is called constant initialization. Together, zero-initialization and constant initialization are Zero-initialization and initialization with a constant expression are collectively called static initialization; all other initialization is dynamic initialization. Static initialization shall be performed Objects of POD types (3.9) with static storage duration initialized with constant expressions (5.19) shall be initialized before any dynamic initialization takes place.

119. Object lifetime and aggregate initialization

[Moved to DR at 4/02 meeting.]

Jack Rouse: 3.8  basic.life paragraph 1 includes:

The lifetime of an object is a runtime property of the object. The lifetime of an object of type T begins when:

• storage with the proper alignment and size for type T is obtained, and
• if T is a class type with a non-trivial constructor (12.1  class.ctor ), the constructor call has completed.

    struct B {
        B( int = 0 );
        ~B();
    };

    struct S {
        B b1;
    };

    int main()
    {
        S s = { 1 };
        return 0;
    }

Mike Miller: I see a couple of ways of fixing the problem. One way would be to change "the constructor call has completed" to "the object's initialization is complete."

Another would be to add following "a class type with a non-trivial constructor" the phrase "that is not initialized with the brace notation (8.5.1  dcl.init.aggr )."

The first formulation treats aggregate initialization like a constructor call; even POD-type members of an aggregate could not be accessed before the aggregate initialization completed. The second is less restrictive; the POD-type members of the aggregate would be usable before the initialization, and the members with non-trivial constructors (the only way an aggregate can acquire a non-trivial constructor) would be protected by recursive application of the lifetime rule.

Proposed resolution (04/01):

In 3.8  basic.life paragraph 1, change

If T is a class type with a non-trivial constructor (12.1  class.ctor), the constructor call has completed.

to

If T is a class type with a non-trivial constructor (12.1  class.ctor), the initialization is complete. [Note: the initialization can be performed by a constructor call or, in the case of an aggregate with an implicitly-declared non-trivial default constructor, an aggregate initialization (8.5.1  dcl.init.aggr).]

274. Cv-qualification and char-alias access to out-of-lifetime objects

[Voted into WP at April 2003 meeting.]

The wording in 3.8  basic.life paragraph 6 allows an lvalue designating an out-of-lifetime object to be used as the operand of a static_cast only if the conversion is ultimately to "char&" or "unsigned char&". This description excludes the possibility of using a cv-qualified version of these types for no apparent reason.

Notes on 04/01 meeting:

The wording should be changed to allow cv-qualified char types.

Proposed resolution (04/01):

In 3.8  basic.life paragraph 6 change the third bullet:

• the lvalue is used as the operand of a static_cast (5.2.9  expr.static.cast) (except when the conversion is ultimately to char& or unsigned char&), or

• the lvalue is used as the operand of a static_cast (5.2.9  expr.static.cast) except when the conversion is ultimately to cv char& or cv unsigned char&, or

404. Unclear reference to construction with non-trivial constructor

[Voted into WP at March 2004 meeting.]

3.8  basic.life paragraph 1 second bullet says:

if T is a class type with a non-trivial constructor (12.1), the constructor call has completed.

This is confusing; what was intended is probably something like

if T is a class type and the constructor invoked to create the object is non-trivial (12.1), the constructor call has completed.

Proposed Resolution (October 2003):

As given above.

158. Aliasing and qualification conversions

[Moved to DR at 4/02 meeting.]

3.10  basic.lval paragraph 15 lists the types via which an lvalue can be used to access the stored value of an object; using an lvalue type that is not listed results in undefined behavior. It is permitted to add cv-qualification to the actual type of the object in this access, but only at the top level of the type ("a cv-qualified version of the dynamic type of the object").

However, 4.4  conv.qual paragraph 4 permits a "conversion [to] add cv-qualifiers at levels other than the first in multi-level pointers." The combination of these two rules allows creation of pointers that cannot be dereferenced without causing undefined behavior. For instance:

    int* jp;
    const int * const * p1 = &jp;
    *p1;    // undefined behavior!

The reason that *p1 results in undefined behavior is that the type of the lvalue is const int * const", which is not "a cv-qualified version of" int*.

Since the conversion is permitted, we must give it defined semantics, hence we need to fix the wording in 3.10  basic.lval to include all possible conversions of the type via 4.4  conv.qual.

Proposed resolution (04/01):

Add a new bullet to 3.10  basic.lval paragraph 15, following "a cv-qualified version of the dynamic type of the object:"

• A type similar (as defined in 4.4  conv.qual) to the dynamic type of the object,

351. Sequence point error: unspecified or undefined?

[Voted into WP at March 2004 meeting.]

I have found what looks like a bug in clause 5  expr, paragraph 4:

Between the previous and next sequence point a scalar object shall have its stored value modified at most once by the evaluation of an expression. Furthermore, the prior value shall be accessed only to determine the value to be stored. The requirements of this paragraph shall be met for each allowable ordering of the subexpressions of a full expression; otherwise the behavior is undefined. Example:

        i = v[i++];                     // the behavior is unspecified
        i = 7, i++, i++;                // i becomes 9

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

--end example]

So which is it, unspecified or undefined?

Notes from October 2002 meeting:

We should find out what C99 says and do the same thing.

Proposed resolution (April 2003):

Change the example in clause 5  expr, paragraph 4 from

[Example:

i = v[i++];                     //  the behavior is unspecified
i = 7, i++, i++;                //   i  becomes  9

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

--- end example]

to (changing "unspecified" to "undefined" twice)

[Example:

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

--- end example]

122. template-ids as unqualified-ids

[Moved to DR at 10/01 meeting.]

5.1  expr.prim paragraph 11 reads,

A template-id shall be used as an unqualified-id only as specified in 14.7.2  temp.explicit , 14.7  temp.spec , and 14.5.4  temp.class.spec .

What uses of template-ids as unqualified-ids is this supposed to prevent? And is the list of referenced sections correct/complete? For instance, what about 14.8.1  temp.arg.explicit, "Explicit template argument specification?" Does its absence from the list in 5.1  expr.prim paragraph 11 mean that "f<int>()" is ill-formed?

This is even more confusing when you recall that unqualified-ids are contained in qualified-ids:

qualified-id: ::_opt nested-name-specifier template_opt unqualified-id

Is the wording intending to say "used as an unqualified-id that is not part of a qualified-id?" Or something else?

Proposed resolution (10/00):

Remove the referenced sentence altogether.

125. Ambiguity in friend declaration syntax

[Voted into WP at March 2004 meeting.]

The example below is ambiguous.

    struct A{
      struct B{};
    };

    A::B C();

    namespace B{
      A C();
    }

    struct Test {
      friend A::B ::C();
    };

The ambiguity arises since both the simple-type-specifier (7.1.5.2  dcl.type.simple paragra 1) and an init-declararator (8  dcl.decl paragraph 1) contain an optional :: and an optional nested-name-specifier (5.1  expr.prim paragraph 1). Therefore, two different ways to analyse this code are possible:

simple-type-specifier = A::B
init-declarator = ::C()
simple-declaration = friend A::B ::C();

simple-type-specifier = A
init-declarator = ::B::C()
simple-declaration = friend A ::B::C();

Suggested Resolution: In the definition of nested-name-specifier, add a sentence saying that a :: token immediately following a nested-name-specifier is always considered as part of the nested-name-specifier. Under this interpretation, the example is ill-formed, and should be corrected as either

    friend A (::B::C)();   //or
    friend A::B (::C)();

An alternate suggestion — changing 7.1  dcl.spec to say that

The longest sequence of tokens that could possibly be a type name is taken as the decl-specifier-seq of a declaration.

— is undesirable because it would make the example well-formed rather than requiring the user to disambiguate the declaration explicitly.

Proposed resolution (04/01):

(See below for problem with this, from 10/01 meeting.)

In 5.1  expr.prim paragraph 7,

1. Before the grammar for qualified-id, start a new paragraph 7a with the text

   A qualified-id is an id-expression that contains the scope resolution operator ::.

2. Following the grammar fragment, insert the following:

   The longest sequence of tokens that could form a qualified-id constitutes a single qualified-id. [Example:

       // classes C, D; functions F, G, namespace N; non-class type T
       friend C ::D::F();   // ill-formed, means friend (C::D::F)();
       friend C (::D::F)(); // well-formed
       friend N::T ::G();   // ill-formed, means friend (N::T::G)();
       friend N::T (::G)(); // well-formed
   

   —end example]

3. Start a new paragraph 7b following the example.

(This resolution depends on that of issue 215.)

Notes from 10/01 meeting:

It was pointed out that the proposed resolution does not deal with cases like X::Y where X is a type but not a class type. The working group reaffirmed its decision that the disambiguation should be syntactic only, i.e., it should depend only on whether or not the name is a type.

At the Seattle meeting, I suggested that a solution might be to change the class-or-namespace-name in the nested-name-specifier rule to just be "identifier"; there was some resistance to this idea. FWIW, I've tried this in g++. I had to revise the idea so that only the second and subsequent names were open to being any identifier, but that seems to work just fine.

So, instead of

nested-name-specifier:
class-or-namespace-name :: nested-name-specifier_opt
class-or-namespace-name :: template nested-name-specifier

it would be

nested-name-specifier:
type-or-namespace-name :: (per issue 215)
nested-name-specifier identifier ::
nested-name-specifier template template-id ::

Or some equivalent but right-associative formulation, if people feel that's important, but it seems irrelevant to me.

Clark Nelson :

Personally, I prefer the left-associative rule. I think it makes it easier to understand. I was thinking about this production a lot at the meeting, considering also some issues related to 301. My formulation was getting kind of ugly, but with a left-associative rule, it gets a lot nicer.

Your proposal isn't complete, however, as it doesn't allow template arguments without an explicit template keyword. You probably want to add an alternative for:

nested-name-specifier type-or-namespace-name ::

There is admittedly overlap between this alternative and

nested-name-specifier identifier ::

but I think they're both necessary.

Notes from the 4/02 meeting:

The changes look good. Clark Nelson will merge the two proposals to produce a single proposed resolution.

Proposed resolution (April 2003):

nested-name-specifier is currently defined in 5.1  expr.prim paragraph 7 as:

nested-name-specifier:
class-or-namespace-name :: nested-name-specifier_opt
class-or-namespace-name :: template nested-name-specifier
class-or-namespace-name:
class-name
namespace-name

The proposed definition is instead:

nested-name-specifier:
type-name ::
namespace-name ::
nested-name-specifier identifier ::
nested-name-specifier template_opt template-id ::

Issue 215 is addressed by using type-name instead of class-name in the first alternative. Issue 125 (this issue) is addressed by using identifier instead of anything more specific in the third alternative. Using left association instead of right association helps eliminate the need for class-or-namespace-name (or type-or-namespace-name, as suggested for issue 215).

It should be noted that this formulation also rules out the possibility of A::template B::, i.e. using the template keyword without any template arguments. I think this is according to the purpose of the template keyword, and that the former rule allowed such a construct only because of the difficulty of formulation of a right-associative rule that would disallow it. But I wanted to be sure to point out this implication.

Notes from April 2003 meeting:

See also issue 96.

The proposed change resolves only part of issue 215.

113. Visibility of called function

[Moved to DR at 10/01 meeting.]

Christophe de Dinechin: In 5.2.2  expr.call , paragraph 2 reads:

If no declaration of the called function is visible from the scope of the call the program is ill-formed.

Mike Miller: "The called function" is unfortunate phraseology; it makes it sound as if it's referring to the function actually called, as opposed to the identifier in the postfix expression. It's wrong with respect to Koenig lookup, too (the declaration need not be visible if it can be found in a class or namespace associated with one or more of the arguments).

In fact, this paragraph should be a note. There's a general rule that says you have to find an unambiguous declaration of any name that is used (3.4  basic.lookup paragraph 1); the only reason this paragraph is here is to contrast with C's implicit declaration of called functions.

Proposed resolution:

If no declaration of the called function is visible from the scope of the call the program is ill-formed.

[Note: if a function or member function name is used, and name lookup (3.4  basic.lookup) does not find a declaration of that name, the program is ill-formed. No function is implicitly declared by such a call. ]

(See also issue 218.)

421. Is rvalue.field an rvalue?

[Voted into WP at March 2004 meeting.]

Consider

  typedef
    struct {
      int a;
    } A;

  A f(void)
  {
    A a;
    return a;
  }

  int main(void)
  {
    int* p = &f().a;   // #1
  }

Should #1 be rejected? The standard is currently silent.

Mike Miller: I don't believe the Standard is silent on this. I will agree that the wording of 5.2.5  expr.ref paragraph 4 bullet 2 is unfortunate, as it is subject to misinterpretation. It reads,

If E1 is an lvalue, then E1.E2 is an lvalue.

Notes from October 2003 meeting:

We agree the reference should be an rvalue, and a change along the lines of that recommended by Mike Miller is reasonable.

Proposed Resolution (October 2003):

Change the second bullet of 5.2.5  expr.ref paragraph 4 to read:

If E1 is an lvalue, then E1.E2 is an lvalue; otherwise, it is an rvalue.

54. Static_cast from private base to derived class

[Voted into WP at October 2004 meeting.]

Is it okay to use a static_cast to cast from a private base class to a derived class? That depends on what the words "valid standard conversion" in paragraph 8 mean — do they mean the conversion exists, or that it would not get an error if it were done? I think the former was intended — and therefore a static_cast from a private base to a derived class would be allowed.

Rationale (04/99): A static_cast from a private base to a derived class is not allowed outside a member from the derived class, because 4.10  conv.ptr paragraph 3 implies that the conversion is not valid. (Classic style casts work.)

Reopened September 2003:

Steve Adamczyk: It makes some sense to disallow the inverse conversion that is pointer-to-member of derived to pointer-to-member of private base. There's less justification for the pointer-to-private-base to pointer-to-derived case. EDG, g++ 3.2, and MSVC++ 7.1 allow the pointer case and disallow the pointer-to-member case. Sun disallows the pointer case as well.

  struct B {};
  struct D : private B {};
  int main() {
    B *p = 0;
    static_cast<D *>(p);  // Pointer case: should be allowed
    int D::*pm = 0;
    static_cast<int B::*>(pm);  // Pointer-to-member case: should get error
  }

There's a tricky case with old-style casts: because the static_cast interpretation is tried first, you want a case like the above to be considered a static_cast, but then issue an error, not be rejected as not a static cast; if you did the latter, you would then try the cast as a reinterpret_cast.

Ambiguity and casting to a virtual base should likewise be errors after the static_cast interpretation is selected.

Notes from the October 2003 meeting:

There was lots of sentiment for making things symmetrical: the pointer case should be the same as the pointer-to-member case. g++ 3.3 now issues errors on both cases.

We decided an error should be issued on both cases. The access part of the check should be done later; by some definition of the word the static_cast is valid, and then later an access error is issued. This is similar to the way standard conversions work.

Proposed Resolution (October 2003):

Replace paragraph 5.2.9  expr.static.cast/6:

The inverse of any standard conversion sequence (clause 4  conv), other than the lvalue-to-rvalue (4.1  conv.lval), array-to-pointer (4.2  conv.array), function-to-pointer (4.3  conv.func), and boolean (4.12  conv.bool) conversions, can be performed explicitly using static_cast. The lvalue-to-rvalue (4.1  conv.lval), array-to-pointer (4.2  conv.array), and function-to-pointer (4.3  conv.func) conversions are applied to the operand. Such a static_cast is subject to the restriction that the explicit conversion does not cast away constness (5.2.11  expr.const.cast), and the following additional rules for specific cases:

with two paragraphs:

The inverse of any standard conversion sequence (clause 4  conv), other than the lvalue-to-rvalue (4.1  conv.lval), array-to-pointer (4.2  conv.array), function-to-pointer (4.3  conv.func), and boolean (4.12  conv.bool) conversions, can be performed explicitly using static_cast. A program is ill-formed if it uses static_cast to perform the inverse of an ill-formed standard conversion sequence.[Example:

struct B {};
struct D : private B {};
void f() {
  static_cast<D*>((B*)0); // Error: B is a private base of D.
  static_cast<int B::*>((int D::*)0); // Error: B is a private base of D.
}

--- end example]

The lvalue-to-rvalue (4.1  conv.lval), array-to-pointer (4.2  conv.array), and function-to-pointer (4.3  conv.func) conversions are applied to the operand. Such a static_cast is subject to the restriction that the explicit conversion does not cast away constness (5.2.11  expr.const.cast), and the following additional rules for specific cases:

In addition, modify the second sentence of 5.4  expr.cast/5. The first two sentences of 5.4  expr.cast/5 presently read:

The conversions performed by

• a const_cast (5.2.11),
• a static_cast (5.2.9),
• a static_cast followed by a const_cast,
• a reinterpret_cast (5.2.10), or
• a reinterpret_cast followed by a const_cast,

can be performed using the cast notation of explicit type conversion. The same semantic restrictions and behaviors apply.

Change the second sentence to read:

The same semantic restrictions and behaviors apply, with the exception that in performing a static_cast in the following situations the conversion is valid even if the base class is inaccessible:

• a pointer to an object of derived class type or an lvalue of derived class type may be explicitly converted to a pointer or reference to an unambiguous base class type, respectively;
• a pointer to member of derived class type may be explicitly converted to a pointer to member of an unambiguous non-virtual base class type;
• a pointer to an object of an unambiguous non-virtual base class type, an lvalue of an unambiguous non-virtual base class type, or a pointer to member of an unambiguous non-virtual base class type may be explicitly converted to a pointer, a reference, or a pointer to member of a derived class type, respectively.

Remove paragraph 5.4  expr.cast/7, which presently reads:

In addition to those conversions, the following static_cast and reinterpret_cast operations (optionally followed by a const_cast operation) may be performed using the cast notation of explicit type conversion, even if the base class type is not accessible:

• a pointer to an object of derived class type or an lvalue of derived class type may be explicitly converted to a pointer or reference to an unambiguous base class type, respectively;
• a pointer to member of derived class type may be explicitly converted to a pointer to member of an unambiguous non-virtual base class type;
• a pointer to an object of non-virtual base class type, an lvalue of non-virtual base class type, or a pointer to member of non-virtual base class type may be explicitly converted to a pointer, a reference, or a pointer to member of a derived class type, respectively.

427. static_cast ambiguity: conversion versus cast to derived

[Voted into WP at October 2004 meeting.]

Consider this code:

  struct B {};

  struct D : public B { 
    D(const B&);
  };

  extern B& b;

  void f() {
    static_cast<const D&>(b);
  }

The rules for static_cast permit the conversion to "const D&" in two ways:

1. D is derived from B, and b is an lvalue, so a cast to D& is OK.
2. const D& t = b is valid, using the constructor for D. [Ed. note: actually, this should be parenthesized initialization.]

The first alternative is 5.2.9  expr.static.cast/5; the second is 5.2.9  expr.static.cast/2.

Presumably the first alternative is better -- it's the "simpler" conversion. The standard does not seem to make that clear.

Steve Adamczyk: I take the "Otherwise" at the beginning of 5.2.9  expr.static.cast/3 as meaning that the paragraph 2 interpretation is used if available, which means in your example above interpretation 2 would be used. However, that's not what EDG's compiler does, and I agree that it's not the "simpler" conversion.

Proposed Resolution (October 2003):

Move paragraph 5.2.9/5:

An lvalue of type ``cv1 B'', where B is a class type, can be cast to type ``reference to cv2 D'', where D is a class derived (clause 10  class.derived) from B, if a valid standard conversion from ``pointer to D'' to ``pointer to B'' exists (4.10  conv.ptr), cv2 is the same cv-qualification as, or greater cv-qualification than, cv1, and B is not a virtual base class of D. The result is an lvalue of type ``cv2 D.'' If the lvalue of type ``cv1 B'' is actually a sub-object of an object of type D, the lvalue refers to the enclosing object of type D. Otherwise, the result of the cast is undefined. [Example:

  struct B {};
  struct D : public B {};
  D d;
  B &br = d;

  static_cast<D&>(br);            //  produces lvalue to the original  d  object

--- end example]

before paragraph 5.2.9  expr.static.cast/2.

Insert Otherwise, before the text of paragraph 5.2.9  expr.static.cast/2 (which will become 5.2.9  expr.static.cast/3 after the above insertion), so that it reads:

Otherwise, an expression e can be explicitly converted to a type T using a static_cast of the form static_cast<T>(e) if the declaration "T t(e);" is well-formed, for some invented temporary variable t (8.5  dcl.init). The effect of such an explicit conversion is the same as performing the declaration and initialization and then using the temporary variable as the result of the conversion. The result is an lvalue if T is a reference type (8.3.2  dcl.ref), and an rvalue otherwise. The expression e is used as an lvalue if and only if the initialization uses it as an lvalue.

439. Guarantees on casting pointer back to cv-qualified version of original type

[Voted into WP at April 2005 meeting.]

Paragraph 5.2.9  expr.static.cast paragraph 10 says that:

A value of type pointer to object converted to "pointer to cv void" and back to the original pointer type will have its original value.

That guarantee should be stronger. In particular, given:

  T* p1 = new T;
  const T* p2 = static_cast<const T*>(static_cast<void *>(p1));
  if (p1 != p2)
    abort ();

A value of type pointer to object converted to "pointer to cv void" and back to the original pointer type (or a variant of the original pointer type that differs only in the cv-qualifiers applied to the object type) will have its original value. [Example:

T* p1 = new T;
const T* p2 = static_cast<const T*>(static_cast<void *>(p1));
bool b = p1 == p2; // b will have the value true.

---end example.]

Proposed resolution:

Change 5.2.9  expr.static.cast paragraph 10 as indicated:

A value of type pointer to object converted to "pointer to cv void" and back to the original pointer type, possibly with different cv-qualification, will have its original value. [Example:

  T* p1 = new T;
  const T* p2 = static_cast<const T*>(static_cast<void *>(p1));
  bool b = p1 == p2; // b will have the value true.

---end example]

Rationale: The wording "possibly with different cv-qualification" was chosen over the suggested wording to allow for changes in cv-qualification at different levels in a multi-level pointer, rather than only at the object type level.

195. Converting between function and object pointers

[Voted into WP at April 2005 meeting.]

It is currently not permitted to cast directly between a pointer to function type and a pointer to object type. This conversion is not listed in 5.2.9  expr.static.cast and 5.2.10  expr.reinterpret.cast and thus requires a diagnostic to be issued. However, if a sufficiently long integral type exists (as is the case in many implementations), it is permitted to cast between pointer to function types and pointer to object types using that integral type as an intermediary.

In C the cast results in undefined behavior and thus does not require a diagnostic, and Unix C compilers generally do not issue one. This fact is used in the definition of the standard Unix function dlsym, which is declared to return void* but in fact may return either a pointer to a function or a pointer to an object. The fact that C++ compilers are required to issue a diagnostic is viewed as a "competitive disadvantage" for the language.

Suggested resolution: Add wording to 5.2.10  expr.reinterpret.cast allowing conversions between pointer to function and pointer to object types, if the implementation has an integral data type that can be used as an intermediary.

Several points were raised in opposition to this suggestion:

1. Early C++ supported this conversion and it was deliberately removed during the standardization process.
2. The existence of an appropriate integral type is irrelevant to whether the conversion is "safe." The condition should be on whether information is lost in the conversion or not.
3. There are numerous ways to address the problem at an implementation level rather than changing the language. For example, the compiler could recognize the specific case of dlsym and omit the diagnostic, or the C++ binding to dlsym could be changed (using templates, for instance) to circumvent the violation.
4. The conversion is, in fact, not supported by C; the dlsym function is simply relying on non-mandated characteristics of C implementations, and we would be going beyond the requirements of C compatibility in requiring (some) implementations to support the conversion.
5. This issue is in fact not a defect (omitted or self-contradictory requirements) in the current Standard; the proposed change would actually be an extension and should not be considered until the full review of the IS.
6. dlsym appears not to be used very widely, and the declaration in the header file is not problematic, only calls to it. Since C code generally requires some porting to be valid C++ anyway, this should be considered one of those items that requires porting.

Martin O'Riordan suggested an alternative approach:

• Do not allow the conversions to be performed via old-style casts.
• Allow reinterpret_cast to perform the conversions with the C-style semantics (undefined behavior if not supported).
• Allow dynamic_cast to perform the conversions with well-defined behavior: the result would be a null pointer if the implementation could not support the requested conversion.

The advantage of this approach is that it would permit writing portable, well-defined programs involving such conversions. However, it breaks the current degree of compatibility between old and new casts, and it adds functionality to dynamic_cast which is not obviously related to its current meaning.

Notes from 04/00 meeting:

Andrew Koenig suggested yet another approach: specify that "no diagnostic is required" if the implementation supports the conversion.

Later note:

It was observed that conversion between function and data pointers is listed as a "common extension" in C99.

Notes on the 10/01 meeting:

It was decided that we want the conversion defined in such a way that it always exists but is always undefined (as opposed to existing only when the size relationship is appropriate, and being implementation-defined in that case). This would allow an implementation to issue an error at compile time if the conversion does not make sense.

Bill Gibbons notes that the definitions of the other similar casts are inconsistent in this regard. Perhaps they should be updated as well.

Proposed resolution (April 2003):

After 5.2.10  expr.reinterpret.cast paragraph 6, insert:

A pointer to a function can be explicitly converted to a pointer to a function of a different type. The effect of calling a function through a pointer to a function type (8.3.5  dcl.fct) that is not the same as the type used in the definition of the function is undefined. Except that converting an rvalue of type ``pointer to T1'' to the type ``pointer to T2'' (where T1 and T2 are function types) and back to its original type yields the original pointer value, the result of such a pointer conversion is unspecified. [Note: see also 4.10  conv.ptr for more details of pointer conversions. ] It is implementation defined whether a conversion from pointer to object to pointer to function and/or a conversion from pointer to function to pointer to object exist.

An lvalue expression of type T1 can be cast to the type ``reference to T2'' if T1 and T2 are object types and an expression of type ``pointer to T1'' can be explicitly converted to the type ``pointer to T2'' using a reinterpret_cast. That is, a reference cast reinterpret_cast< T& >(x) has the same effect as the conversion *reinterpret_cast< T* >(&x) with the built-in & and * operators. The result is an lvalue that refers to the same object as the source lvalue, but with a different type. No temporary is created, no copy is made, and constructors (12.1  class.ctor) or conversion functions (12.3  class.conv) are not called.

Drafting Note:

If either conversion exists, the implementation already has to define the behavior (paragraph 3).

Notes from April 2003 meeting:

The new consensus is that if the implementation allows this cast, pointer-to-function converted to pointer-to-object converted back to the original pointer-to-function should work; anything else is undefined behavior. If the implementation does not allow the cast, it should be ill-formed.

Tom Plum is investigating a new concept, that of a "conditionally-defined" feature, which may be applicable here.

Proposed Resolution (October, 2004):

(See paper J16/04-0067 = WG21 N1627 for background material and rationale for this resolution. The resolution proposed here differs only editorially from the one in the paper.)

1. Insert the following as 1.3.2 and renumber all following definitions accordingly:

   1.3.2  conditionally-supported behavior

   behavior evoked by a program construct that is not a mandatory requirement of this International Standard. If a given implementation supports the construct, the behavior shall be as described herein; otherwise, the implementation shall document that the construct is not supported and shall treat a program containing an occurrence of the construct as ill-formed (1.3.4  defns.ill.formed).

2. Add the indicated words to 1.4  intro.compliance paragraph 2, bullet 2:

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

3. Insert the following as a new paragraph following 5.2.10  expr.reinterpret.cast paragraph 7:

   Converting a pointer to a function to a pointer to an object type or vice versa evokes conditionally-supported behavior. In any such conversion supported by an implementation, converting from an rvalue of one type to the other and back (possibly with different cv-qualification) shall yield the original pointer value; mappings between pointers to functions and pointers to objects are otherwise implementation-defined.

4. Change 7.4  dcl.asm paragraph 1 as indicated:

   The meaning of an An asm declaration evokes conditionally-supported behavior. If supported, its meaning is implementation-defined.

5. Change 7.5  dcl.link paragraph 2 as indicated:

   The string-literal indicates the required language linkage. The meaning of the string-literal is implementation-defined. A linkage-specification with a string that is unknown to the implementation is ill-formed. This International Standard specifies the semantics of C and C++ language linkage. Other values of the string-literal evoke conditionally-supported behavior, with implementation-defined semantics. [Note: Therefore, a linkage-specification with a string-literal that is unknown to the implementation requires a diagnostic. When the string-literal in a linkage-specification names a programming language, the spelling of the programming language's name is implementation-defined. [Note: It is recommended that the spelling be taken from the document defining that language, for example Ada (not ADA) and Fortran or FORTRAN (depending on the vintage). The semantics of a language linkage other than C++ or C are implementation-defined. ]

6. Change 14  temp paragraph 4 as indicated:

   A template, a template explicit specialization (14.7.3  temp.expl.spec), or a class template partial specialization shall not have C linkage. If the linkage of one of these is something other than C or C++, the behavior is implementation-defined result is conditionally-supported behavior, with implementation-defined semantics.

324. Can "&" be applied to assignment to bit-field?

[Voted into WP at October 2003 meeting.]

An assignment returns an lvalue for its left operand. If that operand refers to a bit field, can the "&" operator be applied to the assignment? Can a reference be bound to it?

  struct S { int a:3; int b:3; int c:3; };

  void f()
  {
    struct S s;
    const int *p = &(s.b = 0);     // (a)
    const int &r = (s.b = 0);      // (b)
          int &r2 = (s.b = 0);     // (c)
  }

Notes from the 4/02 meeting:

The working group agreed that this should be an error.

Proposed resolution (October 2002):

In 5.3.2  expr.pre.incr paragraph 1 (prefix "++" and "--" operators), change

The value is the new value of the operand; it is an lvalue.

The result is the updated operand; it is an lvalue, and it is a bit-field if the operand is a bit-field.

In 5.16  expr.cond paragraph 4 ("?" operator), add the indicated text:

If the second and third operands are lvalues and have the same type, the result is of that type and is an lvalue and it is a bit-field if the second or the third operand is a bit-field, or if both are bit-fields.

In 5.17  expr.ass paragraph 1 (assignment operators) add the indicated text (the original text is as updated by issue 221, which is DR but not in TC1):

The assignment operator (=) and the compound assignment operators all group right-to-left. All require a modifiable lvalue as their left operand and return an lvalue with the type and value of the left operand after the assignment has taken place. The result in all cases is a bit-field if the left operand is a bit-field.

Note that issue 222 adds (non-conflicting) text at the end of this same paragraph (5.17  expr.ass paragraph 1).

In 5.18  expr.comma paragraph 1 (comma operator), change:

The type and value of the result are the type and value of the right operand; the result is an lvalue if its right operand is.

The type and value of the result are the type and value of the right operand; the result is an lvalue if the right operand is an lvalue, and is a bit-field if the right operand is an lvalue and a bit-field.

Relevant related text (no changes required):

5.3.1  expr.unary.op paragraph 4:

The operand of & shall not be a bit-field.

8.5.3  dcl.init.ref paragraph 5, bullet 1, sub-bullet 1 (regarding binding a reference to an lvalue):

... is an lvalue (but is not a bit-field) ...

429. Matching deallocation function chosen based on syntax or signature?

[Voted into WP at October 2004 meeting.]

What does this example do?

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

  struct A {
        void* operator new(size_t alloc_size, size_t dummy=0) {
                printf("A::operator new()\n");
                return malloc(alloc_size);
        };

        void operator delete(void* p, size_t s) {
                printf("A::delete %d\n", s);
        };


        A()  {printf("A constructing\n"); throw 2;};

  };

  int
  main() {
    try {
        A* ap = new A;
        delete ap;
    }
    catch(int) {printf("caught\n"); return 1;}
  }  

The fundamental issue here is whether the deletion-on-throw is driven by the syntax of the new (placement or non-placement) or by signature matching. If the former, the operator delete would be called with the second argument equal to the size of the class. If the latter, it would be called with the second argument 0.

Core issue 127 (in TC1) dealt with this topic. It removed some wording in 15.2  except.ctor paragraph 2 that implied a syntax-based interpretation, leaving wording in 5.3.4  expr.new paragraph 19 that is signature-based. But there is no accompanying rationale to confirm an explicit choice of the signature-based approach.

EDG and g++ get 0 for the second argument, matching the presumed core issue 127 resolution. But maybe this should be revisited.

Notes from October 2003 meeting:

There was widespread agreement that the compiler shouldn't just silently call the delete with either of the possible values. In the end, we decided it's smarter to issue an error on this case and force the programmer to say what he means.

Mike Miller's analysis of the status quo: 3.7.3.2  basic.stc.dynamic.deallocation paragraph 2 says that "operator delete(void*, std::size_t)" is a "usual (non-placement) deallocation function" if the class does not declare "operator delete(void*)." 3.7.3.1  basic.stc.dynamic.allocation does not use the same terminology for allocation functions, but the most reasonable way to understand the uses of the term "placement allocation function" in the Standard is as an allocation function that has more than one parameter and thus can (but need not) be called using the "new-placement" syntax described in 5.3.4  expr.new. (In considering issue 127, the core group discussed and endorsed the position that, "If a placement allocation function has default arguments for all its parameters except the first, it can be called using non-placement syntax.")

5.3.4  expr.new paragraph 19 says that any non-placement deallocation function matches a non-placement allocation function, and that a placement deallocation function matches a placement allocation function with the same parameter types after the first -- i.e., a non-placement deallocation function cannot match a placement allocation function. This makes sense, because non-placement ("usual") deallocation functions expect to free memory obtained from the system heap, which might not be the case for storage resulting from calling a placement allocation function.

According to this analysis, the example shows a placement allocation function and a non-placement deallocation function, so the deallocation function should not be invoked at all, and the memory will just leak.

Proposed Resolution (October 2003):

Add the following text at the end of 5.3.4  expr.new paragraph 19:

If the lookup finds the two-parameter form of a usual deallocation function (3.7.3.2  basic.stc.dynamic.deallocation), and that function, considered as a placement deallocation function, would have been selected as a match for the allocation function, the program is ill-formed. [Example:

struct S { 
  // Placement allocation function:
  static void* operator new(std::size_t, std::size_t); 

  // Usual (non-placement) deallocation function:
  static void operator delete(void*, std::size_t); 
}; 

S* p = new (0) S; // ill-formed: non-placement deallocation function matches 
                  // placement allocation function 

--- end example]

353. Is deallocation routine called if destructor throws exception in delete?

[Voted into WP at April 2003 meeting.]

In a couple of comp.std.c++ threads, people have asked whether the Standard guarantees that the deallocation function will be called in a delete-expression if the destructor throws an exception. Most/all people have expressed the opinion that the deallocation function must be called in this case, although no one has been able to cite wording in the Standard supporting that view.

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

static int flag = 0;

inline 
void operator delete(void* p) throw() 
{
   if (flag)
        printf("in deallocation function\n");
   free(p);
}

struct S {
    ~S() { throw 0; }
};

void f() {
    try {
        delete new S;
    }
    catch(...) { }
}

int main() {
       flag=1;
       f();
       flag=0;
       return 0;
}

Proposed resolution (October 2002):

Add to 5.3.5  expr.delete paragraph 7 the highlighted text:

The delete-expression will call a deallocation function (3.7.3.2  basic.stc.dynamic.deallocation) [Note: The deallocation function is called regardless of whether the destructor for the object or some element of the array throws an exception. ]

366. String literal allowed in integral constant expression?

[Voted into WP at October 2003 meeting.]

According to 16.1  cpp.cond paragraph 1, the if-group

#if "Hello, world"

is well-formed, since it is an integral constant expression. Since that may not be obvious, here is why:

5.19  expr.const paragraph 1 says that an integral constant expression may involve literals (2.13  lex.literal); "Hello, world" is a literal. It restricts operators to not use certain type conversions; this expression does not use type conversions. It further disallows functions, class objects, pointers, ... - this expression is none of those, since it is an array.

However, 16.1  cpp.cond paragraph 6 does not explain what to do with this if-group, since the expression evaluates neither to false(zero) nor true(non-zero).

Proposed resolution (October 2002):

Change the beginning of the second sentence of 5.19  expr.const paragraph 1 which currently reads

An integral constant-expression can involve only literals (2.13  lex.literal), ...

An integral constant-expression can involve only literals of arithmetic types (2.13  lex.literal, 3.9.1  basic.fundamental), ...

457. Wording nit on use of const variables in constant expressions

[Voted into WP at April 2005 meeting.]

I'm looking at 5.19  expr.const. I see:

An integral constant-expression can involve only ... const variables or static data members of integral or enumeration types initialized with constant expressions ...

Shouldn't that be "const non-volatile"?

It seems weird to say that:

  const volatile int i = 3;
  int j[i];

Steve Adamczyk: See issue 76, which made the similar change to 7.1.5.1  dcl.type.cv paragraph 2, and probably should have changed this one as well.

Proposed resolution (October, 2004):

Change the first sentence in the second part of 5.19  expr.const paragraph 1 as follows:

An integral constant-expression can involve only literals of arithmetic types (2.13  lex.literal, 3.9.1  basic.fundamental), enumerators, non-volatile const variables or static data members of integral or enumeration types initialized with constant expressions (8.5  dcl.init), non-type template parameters of integral or enumeration types, and sizeof expressions.

281. inline specifier in friend declarations

[Moved to DR at October 2002 meeting.]

There is currently no restriction on the use of the inline specifier in friend declarations. That would mean that the following usage is permitted:

    struct A {
        void f();
    };

    struct B {
        friend inline void A::f();
    };

    void A::f(){}

I believe this should be disallowed because a friend declaration in one class should not be able to change attributes of a member function of another class.

More generally, I think that the inline attribute should only be permitted in friend declarations that are definitions.

Notes from the 04/01 meeting:

The consensus agreed with the suggested resolution. This outcome would be similar to the resolution of issue 136.

Proposed resolution (10/01):

Add to the end of 7.1.2  dcl.fct.spec paragraph 3:

If the inline specifier is used in a friend declaration, that declaration shall be a definition or the function shall have previously been declared inline.

396. Misleading note regarding use of auto for disambiguation

[Voted into WP at March 2004 meeting.]

BTW, I noticed that the following note in 7.1.1  dcl.stc paragraph 2 doesn't seem to have made it onto the issues list or into the TR:

[Note: hence, the auto specifier is almost always redundant and not often used; one use of auto is to distinguish a declaration-statement from an expression-statement (stmt.ambig) explicitly. --- end note]

I thought that this was well known to be incorrect, because using auto does not disambiguate this. Writing:

  auto int f();

Proposed resolution: Replace that note with the following note:

[Note: hence, the auto specifier is always redundant and not often used. --- end note]

John Spicer: I support the proposed change, but I think the disambiguation case is not the one that you describe. An example of the supposed disambiguation is:

  int i;
  int j;
  int main()
  {
    int(i);  // declares i, not reference to ::i
    auto int(j);  // declares j, not reference to ::j
  }

cfront would take "int(i)" as a cast of ::i, so the auto would force what it would otherwise treat as a statement to be considered a declaration (cfront 3.0 warned that this would change in the future).

In a conforming compiler the auto is always redundant (as you say) because anything that could be considered a valid declaration should be treated as one.

Proposed resolution (April 2003):

Replace 7.1.1  dcl.stc paragraph 2

[Note: hence, the auto specifier is almost always redundant and not often used; one use of auto is to distinguish a declaration-statement from an expression-statement (6.8  stmt.ambig) explicitly. --- end note]

[Note: hence, the auto specifier is always redundant and not often used. One use of auto is to distinguish a declaration-statement from an expression-statement explicitly rather than relying on the disambiguation rules (6.8  stmt.ambig), which may aid readers. --- end note]

424. Wording problem with issue 56 resolution on redeclaring typedefs in class scope

[Voted into WP at March 2004 meeting.]

I wonder if perhaps the core issue 56 change in 7.1.3  dcl.typedef paragraph 2 wasn't quite careful enough. The intent was to remove the allowance for:

  struct S {
    typedef int I;
    typedef int I;
  };

but I think it also disallows the following:

  class B {
    typedef struct A {} A;
    void f(struct B::A*p);
  };

See also issue 407.

Proposed resolution (October 2003):

At the end of 7.1.3  dcl.typedef paragraph 2, add the following:

In a given class scope, a typedef specifier can be used to redefine any class-name declared in that scope that is not also a typedef-name to refer to the type to which it already refers. [Example:

  struct S {
    typedef struct A {} A;  // OK
    typedef struct B B;     // OK
    typedef A A;            // error
  };

]

283. Template type-parameters are not syntactically type-names

[Voted into WP at April 2003 meeting.]

Although 14.1  temp.param paragraph 3 contains an assertion that

A type-parameter defines its identifier to be a type-name (if declared with class or typename)

the grammar in 7.1.5.2  dcl.type.simple paragraph 1 says that a type-name is either a class-name, an enum-name, or a typedef-name. The identifier in a template type-parameter is none of those. One possibility might be to equate the identifier with a typedef-name instead of directly with a type-name, which would have the advantage of not requiring parallel treatment of the two in situations where they are treated the same (e.g., in elaborated-type-specifiers, see issue 245). See also issue 215.

Proposed resolution (Clark Nelson, March 2002):

In 14.1  temp.param paragraph 3, change "A type-parameter defines its identifier to be a type-name" to "A type-parameter defines its identifier to be a typedef-name"

In 7.1.5.3  dcl.type.elab paragraph 2, change "If the identifier resolves to a typedef-name or a template type-parameter" to "If the identifier resolves to a typedef-name".

This has been consolidated with the edits for some other issues. See N1376=02-0034.

172. Unsigned int as underlying type of enum

[Moved to DR at October 2002 meeting.]

According to 7.2  dcl.enum paragraph 5, the underlying type of an enum is an unspecified integral type, which could potentially be unsigned int. The promotion rules in 4.5  conv.prom paragraph 2 say that such an enumeration value used in an expression will be promoted to unsigned int. This means that a conforming implementation could give the value false for the following code:

    enum { zero };
    -1 < zero;       // might be false

On a related note, 7.2  dcl.enum paragraph 5 says,

the underlying type shall not be larger than int unless the value of an enumerator cannot fit in an int or unsigned int.

This specification does not allow for an enumeration like

    enum { a = -1, b = UINT_MAX };

Since each enumerator can fit in an int or unsigned int, the underlying type is required to be no larger than int, even though there is no such type that can represent all the enumerators.

Proposed resolution (04/01; obsolete, see below):

Change 7.2  dcl.enum paragraph 5 as follows:

It is implementation-defined which integral type is used as the underlying type for an enumeration except that the underlying type shall not be larger than int unless the value of an enumerator cannot fit in an int or unsigned int neither int nor unsigned int can represent all the enumerator values. Furthermore, the underlying type shall not be an unsigned type if the corresponding signed type can represent all the enumerator values.

See also issue 58.

Notes from 04/01 meeting:

It was noted that 4.5  conv.prom promotes unsigned types smaller than int to (signed) int. The signedness chosen by an implementation for small underlying types is therefore unobservable, so the last sentence of the proposed resolution above should apply only to int and larger types. This observation also prompted discussion of an alternative approach to resolving the issue, in which the b_min and b_max of the enumeration would determine the promoted type rather than the underlying type.

Proposed resolution (10/01):

Change 4.5  conv.prom paragraph 2 from

An rvalue of type wchar_t (3.9.1  basic.fundamental) or an enumeration type (7.2  dcl.enum) can be converted to an rvalue of the first of the following types that can represent all the values of its underlying type: int, unsigned int, long, or unsigned long.

An rvalue of type wchar_t (3.9.1  basic.fundamental) can be converted to an rvalue of the first of the following types that can represent all the values of its underlying type: int, unsigned int, long, or unsigned long. An rvalue of an enumeration type (7.2  dcl.enum) can be converted to an rvalue of the first of the following types that can represent all the values of the enumeration (i.e., the values in the range b_min to b_max as described in 7.2  dcl.enum): int, unsigned int, long, or unsigned long.

377. Enum whose enumerators will not fit in any integral type

[Voted into WP at April 2003 meeting.]

7.2  dcl.enum defines the underlying type of an enumeration as an integral type "that can represent all the enumerator values defined in the enumeration".

What does the standard say about this code:

  enum E { a = LONG_MIN, b = ULONG_MAX };

?

I think this should be ill-formed.

Proposed resolution:

In 7.2  dcl.enum paragraph 5 after

The underlying type of an enumeration is an integral type that can represent all the enumerator values defined in the enumeration.

If no integral type can represent all the enumerator values, the enumeration is ill-formed.

11. How do the keywords typename/template interact with using-declarations?

[Voted into WP at March 2004 meeting.]

Issue 1:

The working paper is not clear about how the typename/template keywords interact with using-declarations:

     template<class T> struct A {
         typedef int X;
     };
     
     template<class T> void f() {
         typename A<T>::X a;      // OK
         using typename A<T>::X;  // OK
         typename X b;  // ill-formed; X must be qualified
         X c;  // is this OK?
     }

1. Continue to allow typename in using-declarations, and template (for member templates) too. Attach the "is a type" or "is a template" attribute to the placeholder name which the using-declaration "declares"
2. Disallow typename and template in using-declarations (just as class-keys are disallowed now). Allow typename and template before unqualified names which refer to dependent qualified names through using-declarations.
3. Document that this is broken.

The core WG already resolved this issue according to (1), but the wording does not seem to have been added to the standard. New wording needs to be drafted.

Issue 2:

Either way, one more point needs clarification. If the first option is adopted:

     template<class T> struct A {
         struct X { };
     };
     
     template<class T> void g() {
         using typename A<T>::X;
         X c;    // if this is OK, then X by itself is a  type
         int X;  // is this OK?
     }

     template<class T> struct A {
         struct X { };
     };
     
     template<class T> void g() {
         using A<T>::X;
         int X;  // is this OK?
     }

From John Spicer:

The "not a new declaration" decision is more of a guiding principle than a hard and fast rule. For example, a name introduced in a using-declaration can have different access than the original declaration.

Like symbolic links, a using-declaration can be viewed as a declaration that declares an alias to another name, much like a typedef.

In my opinion, "X c;" is already well-formed. Why would we permit typename to be used in a using-declaration if not to permit this precise usage?

In my opinion, all that needs to be done is to clarify that the "typeness" or "templateness" attribute of the name referenced in the using-declaration is attached to the alias created by the using-declaration. This is solution #1.

The rules for multiple declarations with the same name in the same scope should treat a using-declaration which names a type as a typedef, just as a typedef of a class name is treated as a class declaration. This needs drafting work. Also see Core issue 36.

Rationale (04/99): Any semantics associated with the typename keyword in using-declarations should be considered an extension.

Notes from the April 2003 meeting:

This was reopened because we are now considering extensions again. We agreed that it is desirable for the typename to be "sticky" on a using-declaration, i.e., references to the name introduced by the using-declaration are known to be type names without the use of the typename keyword (which can't be specified on an unqualified name anyway, as of now). The related issue with the template keyword already has a separate issue 109.

Issue 2 deals with the "struct hack." There is an example in 7.3.3  namespace.udecl paragraph 10 that shows a use of using-declarations to import two names that coexist because of the "struct hack." After some deliberation, we decided that the template-dependent using-declaration case is different enough that we did not have to support the "struct hack" in that case. A name introduced in such a case is like a typedef, and no other hidden type can be accessed through an elaborated type specifier.

Proposed resolution (April 2003, revised October 2003):

Add a new paragraph to the bottom of 7.3.3  namespace.udecl:

If a using-declaration uses the keyword typename and specifies a dependent name (14.6.2  temp.dep), the name introduced by the using-declaration is treated as a typedef-name (7.1.3  dcl.typedef).

258. using-declarations and cv-qualifiers

[Voted into WP at April 2003 meeting.]

According to 7.3.3  namespace.udecl paragraph 12,

When a using-declaration brings names from a base class into a derived class scope, member functions in the derived class override and/or hide member functions with the same name and parameter types in a base class (rather than conflicting).

Note that this description says nothing about the cv-qualification of the hiding and hidden member functions. This means, for instance, that a non-const member function in the derived class hides a const member function with the same name and parameter types instead of overloading it in the derived class scope. For example,

    struct A {
      virtual int f() const;
      virtual int f();
    };
    struct B: A {
      B();
      int f();
      using A::f;
    };

    const B cb;
    int i = cb.f(); // ill-formed: A::f() const hidden in B

The same terminology is used in 10.3  class.virtual paragraph 2:

If a virtual member function vf is declared in a class Base and in a class Derived, derived directly or indirectly from Base, a member function vf with the same name and same parameter list as Base::vf is declared, then Derived::vf is also virtual (whether or not it is so declared) and it overrides Base::vf.

Notes on the 04/01 meeting:

The hiding and overriding should be on the basis of the function signature, which includes any cv-qualification on the function.

Proposed resolution (04/02):

In 7.3.3  namespace.udecl paragraph 12 change:

When a using-declaration brings names from a base class into a derived class scope, member functions in the derived class override and/or hide member functions with the same name and parameter types in a base class (rather than conflicting).

When a using-declaration brings names from a base class into a derived class scope, member functions and member function templates in the derived class override and/or hide member functions and member function templates with the same name, parameter-type-list (8.3.5  dcl.fct), and cv-qualification in a base class (rather than conflicting).

In 10.3  class.virtual paragraph 2 change:

If a virtual member function vf is declared in a class Base and in a class Derived, derived directly or indirectly from Base, a member function vf with the same name and same parameter list as Base::vf is declared, then Derived::vf is also virtual (whether or not it is so declared) and it overrides Base::vf.

If a virtual member function vf is declared in a class Base and in a class Derived, derived directly or indirectly from Base, a member function vf with the same name, parameter-type-list (8.3.5  dcl.fct), and cv-qualification as Base::vf is declared, then Derived::vf is also virtual (whether or not it is so declared) and it overrides Base::vf.

See issue 140 for the definition of parameter-type-list.

460. Can a using-declaration name a namespace?

[Voted into WP at April 2005 meeting.]

Can a using-declaration be used to import a namespace?

namespace my_namespace{
  namespace my_namespace2 {
    int function_of_my_name_space(){ return 2;}
  }
}

int main (){
  using  ::my_namespace::my_namespace2;
  return my_namespace2::function_of_my_name_space();
}

Several popular compilers give an error on this, but there doesn't seem to be anything in 7.3.3  namespace.udecl that prohibits it. It should be noted that the user can get the same effect by using a namespace alias:

  namespace my_namespace2 = ::my_namespace::my_namespace2;

Notes from the March 2004 meeting:

We agree that it should be an error.

Proposed resolution (October, 2004):

Add the following as a new paragraph after 7.3.3  namespace.udecl paragraph 5:

A using-declaration shall not name a namespace;

4. Does extern "C" affect the linkage of function names with internal linkage?

[Moved to DR at 4/01 meeting.]

7.5  dcl.link paragraph 6 says the following:

At most one of a set of overloaded functions with a particular name can have C linkage.

        extern "C" {
            static void f(int) {}
            static void f(float) {}
        };

Suggested resolution: The extern "C" linkage specification applies only to the type of functions with internal linkage, and therefore some of the rules that have to do with name overloading don't apply.

Proposed Resolution:

The intent is to distingush implicit linkage from explicit linkage for both name linkage and language (function type) linkage. (It might be more clear to use the terms name linkage and type linkage to distinguish these concepts. A function can have a name with one kind of linkage and a type with a different kind of linkage. The function itself has no linkage: it has no name, only the declaration has a name. This becomes more obvious when you consider function pointers.)

The tentatively agreed proposal is to apply implicit linkage to names declared in brace-enclosed linkage specifications and to non-top-level names declared in simple linkage specifications; and to apply explicit linkage to top-level names declared in simple linkage specifications.

The language linkage of any function type formed through a function declarator is that of the nearest enclosing linkage-specification. For purposes of determining whether the declaration of a namespace-scope name matches a previous declaration, the language linkage portion of the type of a function declaration (that is, the language linkage of the function itself, not its parameters, return type or exception specification) is ignored.

For a linkage-specification using braces, i.e.

extern string-literal { declaration-seq_opt }

For a linkage-specification without braces, i.e.

extern string-literal declaration

the linkage of the names declared in the top-level declarators of declaration is specified by string-literal; if this conflicts with the linkage of any matching previous declarations, the program is ill-formed. The language linkage of the type of any top-level function declarator is specified by string-literal; if this conflicts with the language linkage of the type of any matching previous function declarations, the program is ill-formed. The effect of the linkage-specification on other (non top-level) names declared in declaration is the same as that of the brace-enclosed form.

Bill Gibbons: In particular, these should be well-formed:

    extern "C" void f(void (*fp)());   // parameter type is pointer to
                                       // function with C language linkage
    extern "C++" void g(void (*fp)()); // parameter type is pointer to
                                       // function with C++ language linkage

    extern "C++" {                     // well-formed: the linkage of "f"
        void f(void(*fp)());           // and the function type used in the
    }                                  // parameter still "C"

    extern "C" {                       // well-formed: the linkage of "g"
        void g(void(*fp)());           // and the function type used in the
    }                                  // parameter still "C++"

but these should not:

    extern "C++" void f(void(*fp)());  // error - linkage of "f" does not
                                       // match previous declaration
                                       // (linkage of function type used in
                                       // parameter is still "C" and is not
                                       // by itself ill-formed)
    extern "C" void g(void(*fp)());    // error - linkage of "g" does not
                                       // match previous declaration
                                       // (linkage of function type used in
                                       // parameter is still "C++" and is not
                                       // by itself ill-formed)

That is, non-top-level declarators get their linkage from matching declarations, if any, else from the nearest enclosing linkage specification. (As already described, top-level declarators in a brace-enclosed linkage specification get the linkage from matching declarations, if any, else from the linkage specifcation; while top-level declarators in direct linkage specifications get their linkage from that specification.)

Mike Miller: This is a pretty significant change from the current specification, which treats the two forms of language linkage similarly for most purposes. I don't understand why it's desirable to expand the differences.

It seems very unintuitive to me that you could have a top-level declaration in an extern "C" block that would not receive "C" linkage.

In the current standard, the statement in 7.5  dcl.link paragraph 4 that

the specified language linkage applies to the function types of all function declarators, function names, and variable names introduced by the declaration(s)

applies to both forms. I would thus expect that in

    extern "C" void f(void(*)());
    extern "C++" {
        void f(void(*)());
    }
    extern "C++" f(void(*)());

both "C++" declarations would be well-formed, declaring an overloaded version of f that takes a pointer to a "C++" function as a parameter. I wouldn't expect that either declaration would be a redeclaration (valid or invalid) of the "C" version of f.

Bill Gibbons: The potential difficulty is the matching process and the handling of deliberate overloading based on language linkage. In the above examples, how are these two declarations matched:

    extern "C" void f(void (*fp1)());

    extern "C++" {
        void f(void(*fp2)());
    }

given that the linkage that is part of fp1 is "C" while the linkage (prior to the matching process) that is part of fp2 is "C++"?

The proposal is that the linkage which is part of the parameter type is not determined until after the match is attempted. This almost always correct because you can't overload "C" and "C++" functions; so if the function names match, it is likely that the declarations are supposed to be the same.

Mike Miller: This seems like more trouble than it's worth. This comparison of function types ignoring linkage specifications is, as far as I know, not found anywhere in the current standard. Why do we need to invent it?

Bill Gibbons: It is possible to construct pathological cases where this fails, e.g.

    extern "C" typedef void (*PFC)();  // pointer to "C" linkage function
    void f(PFC);         // parameter is pointer to "C" function
    void f(void (*)());  // matching declaration or overload based on
                         // difference in linkage type?

It is reasonable to require explicit typedefs in this case so that in the above example the second function declaration gets its parameter type function linkage from the first function declaration.

(In fact, I think you can't get into this situation without having already used typedefs to declare different language linkage for the top-level and parameter linkages.)

For example, if the intent is to overload based on linkage a typedef is needed:

    extern "C" typedef void (*PFC)();  // pointer to "C" linkage function
    void f(PFC);              // parameter is pointer to "C" function
    typedef void (*PFCPP)();  // pointer to "C++" linkage function
    void f(PFCPP);            // parameter is pointer to "C++" function

In this case the two function declarations refer to different functions.

Mike Miller: This seems pretty strange to me. I think it would be simpler to determine the type of the parameter based on the containing linkage specification (implicitly "C++") and require a typedef if the user wants to override the default behavior. For example:

    extern "C" {
        typedef void (*PFC)();    // pointer to "C" function
        void f(void(*)());        // takes pointer to "C" function
    }

    void f(void(*)());            // new overload of "f", taking
                                  // pointer to "C++" function

    void f(PFC);                  // redeclare extern "C" version

Notes from 04/00 meeting:

The following changes were tentatively approved, but because they do not completely implement the proposal above the issue is being kept for the moment in "drafting" status.

Notes from 10/00 meeting:

After further discussion, the core language working group determined that the more extensive proposal described above is not needed and that the following changes are sufficient.

Proposed resolution (04/01):

1. Change the first sentence of 7.5  dcl.link paragraph 1 from

   All function types, function names, and variable names have a language linkage.

   to

   All function types, function names with external linkage, and variable names with external linkage have a language linkage.

2. Change the following sentence of 7.5  dcl.link paragraph 4:

   In a linkage-specification, the specified language linkage applies to the function types of all function declarators, function names, and variable names introduced by the declaration(s).

   to

   In a linkage-specification, the specified language linkage applies to the function types of all function declarators, function names with external linkage, and variable names with external linkage declared within the linkage-specification.

3. Add at the end of the final example on 7.5  dcl.link paragraph 4:

       extern "C" {
         static void f4();    // the name of the function f4 has
                              // internal linkage (not C language
                              // linkage) and the function's type
                              // has C language linkage
       }
       extern "C" void f5() {
         extern void f4();    // Okay -- name linkage (internal)
                              // and function type linkage (C
                              // language linkage) gotten from
                              // previous declaration.
       }
       extern void f4();      // Okay -- name linkage (internal)
                              // and function type linkage (C
                              // language linkage) gotten from
                              // previous declaration.
       void f6() {
         extern void f4();    // Okay -- name linkage (internal)
                              // and function type linkage (C
                              // language linkage) gotten from
                              // previous declaration.
       }
   

4. Change 7.5  dcl.link paragraph 7 from

   Except for functions with internal linkage, a function first declared in a linkage-specification behaves as a function with external linkage. [Example:

       extern "C" double f();
       static double f();     // error
   

   is ill-formed (7.1.1  dcl.stc). ] The form of linkage-specification that contains a braced-enclosed declaration-seq does not affect whether the contained declarations are definitions or not (3.1  basic.def); the form of linkage-specification directly containing a single declaration is treated as an extern specifier (7.1.1  dcl.stc) for the purpose of determining whether the contained declaration is a definition. [Example:

       extern "C" int i;      // declaration
       extern "C" {
   	  int i;           // definition
       }
   

   —end example] A linkage-specification directly containing a single declaration shall not specify a storage class. [Example:

       extern "C" static void f(); // error
   

   —end example]

   to

   A declaration directly contained in a linkage-specification is treated as if it contains the extern specifier (7.1.1  dcl.stc) for the purpose of determining the linkage of the declared name and whether it is a definition. Such a declaration shall not specify a storage class. [Example:

       extern "C" double f();
       static double f();     // error
       extern "C" int i;      // declaration
       extern "C" {
   	    int i;         // definition
       }
       extern "C" static void g(); // error
   

   —end example]

29. Linkage of locally declared functions

[Moved to DR at October 2002 meeting. This was incorrectly marked as having DR status between 4/01 and 4/02. It was overlooked when issue 4 was moved to DR at the 4/01 meeting; this one should have been moved as well, because it's resolved by the changes there.]

Consider the following:

    extern "C" void foo()
    {
        extern void bar();
        bar();
    }

The ARM is explicit and says

A linkage-specification for a function also applies to functions and objects declared within it.

In a linkage-specification, the specified language linkage applies to the function types of all function declarators, function names, and variable names introduced by the declaration(s).

From Mike Miller:

Yes: from 7  dcl.dcl paragraph 1,

declaration:
function-definition

function-definition:
decl-specifier-seq_opt declarator ctor-initializer_opt function-body

From Dag Brück:

Consider the following where extern "C" has been moved to a separate declaration:

    extern "C" void foo();
    
    void foo() { extern void bar(); bar(); }

As a side note, I have always wanted to think that placing extern "C" on a function definition or a separate declaration would produce identical programs.

Proposed Resolution (04/01):

See the proposed resolution for Core issue 4, which covers this case.

The ODR should also be checked to see whether it addresses name and type linkage.

160. Missing std:: qualification

[Moved to DR at 10/01 meeting.]

8.2  dcl.ambig.res paragraph 3 shows an example that includes <cstddef> with no using declarations or directives and refers to size_t without the std:: qualification.

Many references to size_t throughout the document omit the std:: namespace qualification.

This is a typical case. The use of std:: is inconsistent throughout the document.

In addition, the use of exception specifications should be examined for consistency.

(See also issue 282.)

Proposed resolution:

In 1.9  intro.execution paragraph 9, replace all two instances of "sig_atomic_t" by "std::sig_atomic_t".

In 3.1  basic.def paragraph 4, replace all three instances of "string" by "std::string" in the example and insert "#include <string>" at the beginning of the example code.

In 3.6.1  basic.start.main paragraph 4, replace

Calling the function

void exit(int);

declared in <cstdlib>...

by

Calling the function std::exit(int) declared in <cstdlib>...

and also replace "exit" by "std::exit" in the last sentence of that paragraph.

In 3.6.1  basic.start.main first sentence of paragraph 5, replace "exit" by "std::exit".

In 3.6.2  basic.start.init paragraph 4, replace "terminate" by "std::terminate".

In 3.6.3  basic.start.term paragraph 1, replace "exit" by "std::exit" (see also issue 28).

In 3.6.3  basic.start.term paragraph 3, replace all three instances of "atexit" by "std::atexit" and both instances of "exit" by "std::exit" (see also issue 28).

In 3.6.3  basic.start.term paragraph 4, replace

Calling the function

void abort();

declared in <cstdlib>...

by

Calling the function std::abort() declared in <cstdlib>...

In 3.7.3.1  basic.stc.dynamic.allocation paragraph 1 third sentence, replace "size_t" by "std::size_t".

In 3.7.3.1  basic.stc.dynamic.allocation paragraph 3, replace "new_handler" by "std::new_handler". Furthermore, replace "set_new_handler" by "std::set_new_handler" in the note.

In 3.7.3.1  basic.stc.dynamic.allocation paragraph 4, replace "type_info" by "std::type_info" in the note.

In 3.7.3.2  basic.stc.dynamic.deallocation paragraph 3, replace all four instances of "size_t" by "std::size_t".

In 3.8  basic.life paragraph 5, replace "malloc" by "std::malloc" in the example code and insert "#include <cstdlib>" at the beginning of the example code.

In 3.9  basic.types paragraph 2, replace "memcpy" by "std::memcpy" (the only instance in the footnote and both instances in the example) and replace "memmove" by "std::memmove" in the footnote (see also issue 43).

In 3.9  basic.types paragraph 3, replace "memcpy" by "std::memcpy", once in the normative text and once in the example (see also issue 43).

In 3.9.1  basic.fundamental paragraph 8 last sentence, replace "numeric_limits" by "std::numeric_limits".

In 5.2.7  expr.dynamic.cast paragraph 9 second sentence, replace "bad_cast" by "std::bad_cast".

In 5.2.8  expr.typeid paragraph 2, replace "type_info" by "std::type_info" and "bad_typeid" by "std::bad_typeid".

In 5.2.8  expr.typeid paragraph 3, replace "type_info" by "std::type_info".

In 5.2.8  expr.typeid paragraph 4, replace both instances of "type_info" by "std::type_info".

In 5.3.3  expr.sizeof paragraph 6, replace both instances of "size_t" by "std::size_t".

In 5.3.4  expr.new paragraph 11 last sentence, replace "size_t" by "std::size_t".

In 5.7  expr.add paragraph 6, replace both instances of "ptrdiff_t" by "std::ptrdiff_t".

In 5.7  expr.add paragraph 8, replace "ptrdiff_t" by "std::ptrdiff_t".

In 6.6  stmt.jump paragraph 2, replace "exit" by "std::exit" and "abort" by "std::abort" in the note.

In 8.2  dcl.ambig.res paragraph 3, replace "size_t" by "std::size_t" in the example.

In 8.4  dcl.fct.def paragraph 5, replace "printf" by "std::printf" in the note.

In 12.4  class.dtor paragraph 13, replace "size_t" by "std::size_t" in the example.

In 12.5  class.free paragraph 2, replace all four instances of "size_t" by "std::size_t" in the example.

In 12.5  class.free paragraph 6, replace both instances of "size_t" by "std::size_t" in the example.

In 12.5  class.free paragraph 7, replace all four instances of "size_t" by "std::size_t" in the two examples.

In 12.7  class.cdtor paragraph 4, replace "type_info" by "std::type_info".

In 13.6  over.built paragraph 13, replace all five instances of "ptrdiff_t" by "std::ptrdiff_t".

In 13.6  over.built paragraph 14, replace "ptrdiff_t" by "std::ptrdiff_t".

In 13.6  over.built paragraph 21, replace both instances of "ptrdiff_t" by "std::ptrdiff_t".

In 14.2  temp.names paragraph 4, replace both instances of "size_t" by "std::size_t" in the example. (The example is quoted in issue 96.)

In 14.3  temp.arg paragraph 1, replace "complex" by "std::complex", once in the example code and once in the comment.

In 14.7.3  temp.expl.spec paragraph 8, issue 24 has already corrected the example.

In 15.1  except.throw paragraph 6, replace "uncaught_exception" by "std::uncaught_exception".

In 15.1  except.throw paragraph 7, replace "terminate" by "std::terminate" and both instances of "unexpected" by "std::unexpected".

In 15.1  except.throw paragraph 8, replace "terminate" by "std::terminate".

In 15.2  except.ctor paragraph 3, replace "terminate" by "std::terminate".

In 15.3  except.handle paragraph 9, replace "terminate" by "std::terminate".

In 15.4  except.spec paragraph 8, replace "unexpected" by "std::unexpected".

In 15.4  except.spec paragraph 9, replace "unexpected" by "std::unexpected" and "terminate" by "std::terminate".

In 15.5  except.special paragraph 1, replace "terminate" by "std::terminate" and "unexpected" by "std::unexpected".

In the heading of 15.5.1  except.terminate, replace "terminate" by "std::terminate".

In 15.5.1  except.terminate paragraph 1, footnote in the first bullet, replace "terminate" by "std::terminate". In the same paragraph, fifth bullet, replace "atexit" by "std::atexit". In the same paragraph, last bullet, replace "unexpected_handler" by "std::unexpected_handler".

In 15.5.1  except.terminate paragraph 2, replace

In such cases,

void terminate();

is called...

by

In such cases, std::terminate() is called...

and replace all three instances of "terminate" by "std::terminate".

In the heading of 15.5.2  except.unexpected, replace "unexpected" by "std::unexpected".

In 15.5.2  except.unexpected paragraph 1, replace

...the function

void unexpected();

is called...

by

...the function std::unexpected() is called...

In 15.5.2  except.unexpected paragraph 2, replace "unexpected" by "std::unexpected" and "terminate" by "std::terminate".

In 15.5.2  except.unexpected paragraph 3, replace "unexpected" by "std::unexpected".

In the heading of 15.5.3  except.uncaught, replace "uncaught_exception" by "std::uncaught_exception".

In 15.5.3  except.uncaught paragraph 1, replace

The function

bool uncaught_exception()

returns true...

by

The function std::uncaught_exception() returns true...

In the last sentence of the same paragraph, replace "uncaught_exception" by "std::uncaught_exception".

112. Array types and cv-qualifiers

[Moved to DR at 10/01 meeting.]

Steve Clamage: Section 8.3.4  dcl.array paragraph 1 reads in part as follows:

Any type of the form "cv-qualifier-seq array of N T" is adjusted to "array of N cv-qualifier-seq T," and similarly for "array of unknown bound of T." [Example:

    typedef int A[5], AA[2][3];
    typedef const A CA;     // type is "array of 5 const int"
    typedef const AA CAA;   // type is "array of 2 array of 3 const int"

—end example] [Note: an "array of N cv-qualifier-seq T" has cv-qualified type; such an array has internal linkage unless explicitly declared extern (7.1.5.1  dcl.type.cv ) and must be initialized as specified in 8.5  dcl.init . ]

Mike Miller: I disagree; all it says is that whether the qualification on the element type is direct ("const int x[5]") or indirect ("const A CA"), the array itself is qualified in the same way the elements are.

Steve Clamage: In addition, section 3.9.3  basic.type.qualifier paragraph 2 says:

A compound type (3.9.2  basic.compound ) is not cv-qualified by the cv-qualifiers (if any) of the types from which it is compounded. Any cv-qualifiers applied to an array type affect the array element type, not the array type (8.3.4  dcl.array )."

Mike Miller: Yes, but consider the last two sentences of 3.9.3  basic.type.qualifier paragraph 5:

Cv-qualifiers applied to an array type attach to the underlying element type, so the notation "cv T," where T is an array type, refers to an array whose elements are so-qualified. Such array types can be said to be more (or less) cv-qualified than other types based on the cv-qualification of the underlying element types.

Mike Ball: I find this a very far reach. The text in 8.3.4  dcl.array is essentially that which is in the C standard (and is a change from early versions of C++). I don't see any justification at all for the bidirectional equivalence. It seems to me that the note is left over from the earlier version of the language.

Steve Clamage: Finally, the Note seems to say that the declaration

    volatile char greet[6] = "Hello";

Have I missed something, or should that Note be entirely removed?

Mike Miller: At least the wording in the note should be repaired not to indicate that volatile-qualification gives an array internal linkage. Also, depending on how the discussion goes, either the wording in 3.9.3  basic.type.qualifier paragraph 2 or in paragraph 5 needs to be amended to be consistent regarding whether an array type is considered qualified by the qualification of its element type.

Steve Adamczyk pointed out that the current state of affairs resulted from the need to handle reference binding consistently. The wording is intended to define the question, "Is an array type cv-qualified?" as being equivalent to the question, "Is the element type of the array cv-qualified?"

Proposed resolution (10/00):

Replace the portion of the note in 8.3.4  dcl.array paragraph 1 reading

such an array has internal linkage unless explicitly declared extern (7.1.5.1  dcl.type.cv) and must be initialized as specified in 8.5  dcl.init.

with

see 3.9.3  basic.type.qualifier.

140. Agreement of parameter declarations

[Moved to DR at 10/01 meeting.]

8.3.5  dcl.fct paragraph 3 says,

All declarations for a function with a given parameter list shall agree exactly both in the type of the value returned and in the number and type of parameters.

    int f(const int);
    int f(int x) { ... }

Tom Plum: I think the FDIS quotation means that the pair of decls are valid. But it doesn't clearly answer whether x is const inside the function definition. As to intent, I know the intent was that if the function definition wants to specify that x is const, the const must appear specifically in the defining decl, not just on some decl elsewhere. But I can't prove that intent from the drafted words.

Mike Miller: I think the intent was something along the following lines:

Two function declarations denote the same entity if the names are the same and the function signatures are the same. (Two function declarations with C language linkage denote the same entity if the names are the same.) All declarations of a given function shall agree exactly both in the type of the value returned and in the number and type of parameters; the presence or absence of the ellipsis is considered part of the signature.

According to this paragraph, the type of a parameter is determined by considering its decl-specifier-seq and declarator and then applying the array-to-pointer and function-to-pointer adjustments. The cv-qualifier and storage class adjustments are performed for the function type but not for the parameter types.

If my interpretation of the intent of the second sentence of the paragraph is correct, the two declarations in the example violate that restriction — the parameter types are not the same, even though the function types are. Since there's no dispensation mentioned for "no diagnostic required," an implementation presumably must issue a diagnostic in this case. (I think "no diagnostic required" should be stated if the declarations occur in different translation units — unless there's a blanket statement to that effect that I have forgotten?)

(I'd also note in passing that, if my interpretation is correct,

    void f(int);
    void f(register int) { }

Proposed resolution (10/00):

1. In 1.3.10  defns.signature, change "the types of its parameters" to "its parameter-type-list (8.3.5  dcl.fct)".

2. In the third bullet of 3.5  basic.link paragraph 9 change "the function types are identical for the purposes of overloading" to "the parameter-type-lists of the functions (8.3.5  dcl.fct) are identical."

3. In the sub-bullets of the third bullet of 5.2.5  expr.ref paragraph 4, change all four occurrences of "function of (parameter type list)" to "function of parameter-type-list."

4. In 8.3.5  dcl.fct paragraph 3, change

   All declarations for a function with a given parameter list shall agree exactly both in the type of the value returned and in the number and type of parameters; the presence or absence of the ellipsis is considered part of the function type.

   to

   All declarations for a function shall agree exactly in both the return type and the parameter-type-list.

5. In 8.3.5  dcl.fct paragraph 3, change

   The resulting list of transformed parameter types is the function's parameter type list.

   to

   The resulting list of transformed parameter types and the presence or absence of the ellipsis is the function's parameter-type-list.

6. In 8.3.5  dcl.fct paragraph 4, change "the parameter type list" to "the parameter-type-list."

7. In the second bullet of 13.1  over.load paragraph 2, change all occurrences of "parameter types" to "parameter-type-list."

8. In 13.3  over.match paragraph 1, change "the types of the parameters" to "the parameter-type-list."

9. In the last sub-bullet of the third bullet of 13.3.1.2  over.match.oper paragraph 3, change "parameter type list" to "parameter-type-list."

Note, 7 Sep 2001:

Editorial changes while putting in issue 147 brought up the fact that injected-class-name is not a syntax term and therefore perhaps shouldn't be written with hyphens. The same can be said of parameter-type-list.

262. Default arguments and ellipsis

[Voted into WP at April 2003 meeting.]

The interaction of default arguments and ellipsis is not clearly spelled out in the current wording of the Standard. 8.3.6  dcl.fct.default paragraph 4 says,

In a given function declaration, all parameters subsequent to a parameter with a default argument shall have default arguments supplied in this or previous declarations.

Strictly speaking, ellipsis isn't a parameter, but this could be clearer. Also, in 8.3.5  dcl.fct paragraph 2,

If the parameter-declaration-clause terminates with an ellipsis, the number of arguments shall be equal to or greater than the number of parameters specified.

This could be interpreted to refer to the number of arguments after the addition of default arguments to the argument list given in the call expression, but again it could be clearer.

Notes from 04/01 meeting:

The consensus opinion was that an ellipsis is not a parameter and that default arguments should be permitted preceding an ellipsis.

Proposed Resolution (4/02):

Change the following sentence in 8.3.5  dcl.fct paragraph 2 from

If the parameter-declaration-clause terminates with an ellipsis, the number of arguments shall be equal to or greater than the number of parameters specified.

to

If the parameter-declaration-clause terminates with an ellipsis, the number of arguments shall be equal to or greater than the number of parameters that do not have a default argument.

As noted in the defect, section 8.3.6  dcl.fct.default is correct but could be clearer.

In 8.3.6  dcl.fct.default, add the following as the first line of the example in paragraph 4.

  void g(int = 0, ...);  // okay, ellipsis is not a parameter so it can follow 
                         // a parameter with a default argument

295. cv-qualifiers on function types

[Moved to DR at October 2002 meeting.]

This concerns the inconsistent treatment of cv qualifiers on reference types and function types. The problem originated with GCC bug report c++/2810. The bug report is available at http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=2810&database=gcc

8.3.2  dcl.ref describes references. Of interest is the statement (my emphasis)

Cv-qualified references are ill-formed except when the cv-qualifiers are introduced through the use of a typedef or of a template type argument, in which case the cv-qualifiers are ignored.

Though it is strange to ignore 'volatile' here, that is not the point of this defect report. 8.3.5  dcl.fct describes function types. Paragraph 4 states,

In fact, if at any time in the determination of a type a cv-qualified function type is formed, the program is ill-formed.

No allowance for typedefs or template type parameters is made here, which is inconsistent with the equivalent reference case.

The GCC bug report was template code which attempted to do,

    template <typename T> void foo (T const &);
    void baz ();
    ...
    foo (baz);

in the instantiation of foo, T is `void ()' and an attempt is made to const qualify that, which is ill-formed. This is a surprise.

Suggested resolution:

Replace the quoted sentence from paragraph 4 in 8.3.5  dcl.fct with

cv-qualified functions are ill-formed, except when the cv-qualifiers are introduced through the use of a typedef or of a template type argument, in which case the cv-qualifiers are ignored.

Adjust the example following to reflect this.

Proposed resolution (10/01):

In 8.3.5  dcl.fct paragraph 4, replace

The effect of a cv-qualifier-seq in a function declarator is not the same as adding cv-qualification on top of the function type, i.e., it does not create a cv-qualified function type. In fact, if at any time in the determination of a type a cv-qualified function type is formed, the program is ill-formed. [Example:

  typedef void F();
  struct S {
    const F f;          // ill-formed
  };

-- end example]

The effect of a cv-qualifier-seq in a function declarator is not the same as adding cv-qualification on top of the function type. In the latter case, the cv-qualifiers are ignored. [Example:

  typedef void F();
  struct S {
    const F f;          // ok; equivalent to void f();
  };

-- end example]

Strike the last bulleted item in 14.8.2  temp.deduct paragraph 2, which reads

Attempting to create a cv-qualified function type.

Nathan Sidwell comments (18 Dec 2001 ): The proposed resolution simply states attempts to add cv qualification on top of a function type are ignored. There is no mention of whether the function type was introduced via a typedef or template type parameter. This would appear to allow

  void (const *fptr) ();

  int &const ref;

Is this difference intentional? It seems needlessly confusing.

Notes from 4/02 meeting:

Yes, the difference is intentional. There is no way to add cv-qualifiers other than those cases.

Notes from April 2003 meeting:

Nathan Sidwell pointed out that some libraries use the inability to add const to a type T as a way of testing that T is a function type. He will get back to us if he has a proposal for a change.

136. Default arguments and friend declarations

[Moved to DR at 10/01 meeting.]

8.3.6  dcl.fct.default paragraph 4 says,

For non-template functions, default arguments can be added in later declarations of a function in the same scope. Declarations in different scopes have completely distinct sets of default arguments. That is, declarations in inner scopes do not acquire default arguments from declarations in outer scopes, and vice versa.

    void f(int, int, int=0);             // #1
    class C {
        friend void f(int, int=0, int);  // #2
    };
    void f(int=0, int, int);             // #3

There are several related questions involved with this issue:

1. Is the friend declaration in the scope of class C or in the surrounding namespace scope?

   Mike Miller: 8.3.6  dcl.fct.default paragraph 4 is speaking about the lexical location of the declaration... The friend declaration occurs in a different declarative region from the declaration at #1, so I would read [this paragraph] as saying that it starts out with a clean slate of default arguments.

   Bill Gibbons: Yes. It occurs in a different region, although it declares a name in the same region (i.e. a redeclaration). This is the same as with local externs and is intended to work the same way. We decided that local extern declarations cannot add (beyond the enclosing block) new default arguments, and the same should apply to friend declarations.

   John Spicer: The question is whether [this paragraph] does (or should) mean declarations that appear in the same lexical scope or declarations that declare names in the same scope. In my opinion, it really needs to be the latter. It seems somewhat paradoxical to say that a friend declaration declares a function in namespace scope yet the declaration in the class still has its own attributes. To make that work I think you'd have to make friends more like block externs that really do introduce a name into the scope in which the declaration is contained.

2. Should default arguments be permitted in friend function declarations, and what effect should they have?

   Bill Gibbons: In the absence of a declaration visible in class scope to which they could be attached, default arguments on friend declarations do not make sense. [They should be] ill-formed, to prevent surprises.

   John Spicer: It is important that the following case work correctly:

           class X {
                   friend void f(X x, int i = 1){}
           };
   
           int main()
           {
                   X x;
                   f(x);
           }
   

   In other words, a function first declared in a friend declaration must be permitted to have default arguments and those default arguments must be usable when the function is found by argument dependent lookup. The reason that this is important is that it is common practice to define functions in friend declarations in templates, and that definition is the only place where the default arguments can be specified.

3. What restrictions should be placed on default argument usage with friend declarations?

   John Spicer: We want to avoid instantiation side effects. IMO, the way to do this would be to prohibit a friend declaration from providing default arguments if a declaration of that function is already visible. Once a function has had a default specified in a friend declaration it should not be possible to add defaults in another declaration be it a friend or normal declaration.

   Mike Miller: The position that seems most reasonable to me is to allow default arguments in friend declarations to be used in Koenig lookup, but to say that they are completely unrelated to default arguments in declarations in the surrounding scope; and to forbid use of a default argument in a call if more than one declaration in the overload set has such a default, as in the proposed resolution for issue 1.

Notes from 10/99 meeting:

Four possible outcomes were identified:

1. If a friend declaration declares a default parameter, allow no other declarations of that function in the translation unit.
2. Same as preceding, but only allow the friend declaration if it is also a definition.
3. Disallow default arguments in friend declarations.
4. Treat the default arguments in each friend declaration as a distinct set, causing an error if the call would be ambiguous.

The core group eliminated the first and fourth options from consideration, but split fairly evenly between the remaining two.

A straw poll of the full committee yielded the following results (given as number favoring/could live with/"over my dead body"):

1. 0/14/5
2. 8/13/5
3. 11/7/14
4. 7/10/9

Additional discussion is recorded in the "Record of Discussion" for the meeting, J16/99-0036 = WG21 N1212. See also paper J16/00-0040 = WG21 N1263.

Proposed resolution (10/00):

In 8.3.6  dcl.fct.default, add following paragraph 4:

If a friend declaration specifies a default argument expression, that declaration must be a definition and shall be the only declaration of the function or function template in the translation unit.

5. CV-qualifiers and type conversions

[Moved to DR at 4/01 meeting.]

The description of copy-initialization in 8.5  dcl.init paragraph 14 says:

• If the destination type is a (possibly cv-qualified) class type:
   ...
• Otherwise (i.e. for the remaining copy-initialization cases), user-defined conversion sequences that can convert from the source type to the destination type or (when a conversion function is used) to a derived class thereof are enumerated ... if the function is a constructor, the call initializes a temporary of the destination type. ...

     struct A {
       A(A&);
     };
     struct B : A { };
     
     struct C {
       operator B&();
     };
     
     C c;
     const A a = c; // allowed?

The temporary created with the conversion function is an lvalue of type B. If the temporary must have the cv-qualifiers of the destination type (i.e. const) then the copy-constructor for A cannot be called to create the object of type A from the lvalue of type const B. If the temporary has the cv-qualifiers of the result type of the conversion function, then the copy-constructor for A can be called to create the object of type A from the lvalue of type const B. This last outcome seems more appropriate.

Steve Adamczyk:

Because of late changes to this area, the relevant text is now the third sub-bullet of the fourth bullet of 8.5  dcl.init paragraph 14:

Otherwise (i.e., for the remaining copy-initialization cases), user-defined conversion sequences that can convert from the source type to the destination type or (when a conversion function is used) to a derived class thereof are enumerated... The function selected is called with the initializer expression as its argument; if the function is a constructor, the call initializes a temporary of the destination type. The result of the call (which is the temporary for the constructor case) is then used to direct-initialize, according to the rules above, the object that is the destination of the copy-initialization.

The issue still remains whether the wording should refer to "the cv-unqualified version of the destination type." I think it should.

Notes from 10/00 meeting:

The original example does not illustrate the remaining problem. The following example does:

    struct C { };
    C c;
    struct A {
        A(const A&);
        A(const C&);
    };
    const volatile A a = c;    // Okay

Proposed Resolution (04/01):

In 8.5  dcl.init, paragraph 14, bullet 4, sub-bullet 3, change

if the function is a constructor, the call initializes a temporary of the destination type.

to

if the function is a constructor, the call initializes a temporary of the cv-unqualified version of the destination type.

177. Lvalues vs rvalues in copy-initialization

[Moved to DR at 4/02 meeting.]

Is the temporary created during copy-initialization of a class object treated as an lvalue or an rvalue? That is, is the following example well-formed or not?

    struct B { };
    struct A {
        A(A&);    // not const
        A(const B&);
    };
    B b;
    A a = b;

According to 8.5  dcl.init paragraph 14, the initialization of a is performed in two steps. First, a temporary of type A is created using A::A(const B&). Second, the resulting temporary is used to direct-initialize a using A::A(A&).

The second step requires binding a reference to non-const to the temporary resulting from the first step. However, 8.5.3  dcl.init.ref paragraph 5 requires that such a reference be bound only to lvalues.

It is not clear from 3.10  basic.lval whether the temporary created in the process of copy-initialization should be treated as an lvalue or an rvalue. If it is an lvalue, the example is well-formed, otherwise it is ill-formed.

Proposed resolution (04/01):

1. In 8.5  dcl.init paragraph 14, insert the following after "the call initializes a temporary of the destination type":

   The temporary is an rvalue.

2. In 15.1  except.throw paragraph 3, replace

   The temporary is used to initialize the variable...

   with

   The temporary is an lvalue and is used to initialize the variable...

(See also issue 84.)

277. Zero-initialization of pointers

[Moved to DR at 10/01 meeting.]

The intent of 8.5  dcl.init paragraph 5 is that pointers that are zero-initialized will contain a null pointer value. Unfortunately, the wording used,

...set to the value of 0 (zero) converted to T

does not match the requirements for creating a null pointer value given in 4.10  conv.ptr paragraph 1:

A null pointer constant is an integral constant expression (5.19  expr.const) rvalue of integer type that evaluates to zero. A null pointer constant can be converted to a pointer type; the result is the null pointer value of that type...

The problem is that the "value of 0" in the description of zero-initialization is not specified to be an integral constant expression. Nonconstant expressions can also have the value 0, and converting a nonconst 0 to pointer type need not result in a null pointer value.

Proposed resolution (04/01):

In 8.5  dcl.init paragraph 5, change

...set to the value 0 (zero) converted to T;

to

...set to the value 0 (zero), taken as an integral constant expression, converted to T; [footnote: as specified in 4.10  conv.ptr, converting an integral constant expression whose value is 0 to a pointer type results in a null pointer value.]

302. Value-initialization and generation of default constructor

[Moved to DR at October 2002 meeting.]

We've been looking at implementing value-initialization. At one point, some years back, I remember Bjarne saying that something like X() in an expression should produce an X object with the same value one would get if one created a static X object, i.e., the uninitialized members would be zero-initialized because the whole object is initialized at program startup, before the constructor is called.

The formulation for default-initialization that made it into TC1 (in 8.5  dcl.init) is written a little differently (see issue 178), but I had always assumed that it would still be a valid implementation to zero the whole object and then call the default constructor for the troublesome "non-POD but no user-written constructor" cases.

That almost works correctly, but I found a problem case:

    struct A {
      A();
      ~A();
    };
    struct B {
      // B is a non-POD with no user-written constructor.
      // It has a nontrivial generated constructor.
      const int i;
      A a;
    };
    int main () {
      // Value-initializing a "B" doesn't call the default constructor for
      // "B"; it value-initializes the members of B.  Therefore it shouldn't
      // cause an error on generation of the default constructor for the
      // following:
      new B();
    }

If the definition of the B::B() constructor is generated, an error is issued because the const member "i" is not initialized. But the definition of value-initialization doesn't require calling the constructor, and therefore it doesn't require generating it, and therefore the error shouldn't be detected.

So this is a case where zero-initializing and then calling the constructor is not equivalent to value-initializing, because one case generates an error and the other doesn't.

This is sort of unfortunate, because one doesn't want to generate all the required initializations at the point where the "()" initialization occurs. One would like those initializations to be packaged in a function, and the default constructor is pretty much the function one wants.

I see several implementation choices:

1. Zero the object, then call the default generated constructor. This is not valid unless the standard is changed to say that the default constructor might be generated for value-initialization cases like the above (that is, it's implementation-dependent whether the constructor definition is generated). The zeroing operation can of course be optimized, if necessary, to hit only the pieces of the object that would otherwise be left uninitialized. An alternative would be to require generation of the constructor for value-initialization cases, even if the implementation technique doesn't call the constructor at that point. It's pretty likely that the constructor is going to have to be generated at some point in the program anyway.
2. Make a new value-initialization "constructor," whose body looks a lot like the usual generated constructor, but which also zeroes other members. No errors would be generated while generating this modified constructor, because it generates code that does full initialization. (Actually, it wouldn't guarantee initialization of reference members, and that might be an argument for generating the constructor, in order to get that error.) This is standard-conforming, but it destroys object-code compatibility.
3. Variation on (1): Zero first, and generate the object code for the default constructor when it's needed for value-initialization cases, but don't issue any errors at that time. Issue the errors only if it turns out the constructor is "really" referenced. Aside from the essential shadiness of this approach, I fear that something in the generation of the constructor will cause a template instantiation which will be an abservable side effect.

Personally, I find option 1 the least objectionable.

Proposed resolution (10/01):

Add the indicated wording to the third-to-last sentence of 3.2  basic.def.odr pararaph 2:

A default constructor for a class is used by default initialization or value initialization as specified in 8.5  dcl.init.

Add a footnote to the indicated bullet in 8.5  dcl.init paragraph 5:

• if T is a non-union class type without a user-declared constructor, then every non-static data member and base-class component of T is value-initialized. [Footnote: Value-initialization for such a class object may be implemented by zero-initializing the object and then calling the default constructor.]

Add the indicated wording to the first sentence of 12.1  class.ctor paragraph 7:

An implicitly-declared default constructor for a class is implicitly defined when it is used (3.2  basic.def.odr) to create an object of its class type (1.8  intro.object).

175. Class name injection and base name access

[Moved to DR at 10/01 meeting.]

With class name injection, when a base class name is used in a derived class, the name found is the injected name in the base class, not the name of the class in the scope containing the base class. Consequently, if the base class name is not accessible (e.g., because is is in a private base class), the base class name cannot be used unless a qualified name is used to name the class in the class or namespace of which it is a member.

Without class name injection the following example is valid. With class name injection, A is inaccessible in class C.

    class A { };
    class B: private A { };
    class C: public B {
        A* p;    // error: A inaccessible
    };

At the least, the standard should be more explicit that this is, in fact, ill-formed.

(See paper J16/99-0010 = WG21 N1187.)

Proposed resolution (04/01):

Add to the end of 11.1  class.access.spec paragraph 3:

[Note: In a derived class, the lookup of a base class name will find the injected-class-name instead of the name of the base class in the scope in which it was declared. The injected-class-name might be less accessible than the name of the base class in the scope in which it was declared.] [Example:

    class A { };
    class B : private A { };
    class C : public B {
        A* p;    // error: injected-class-name A is inaccessible
        ::A* q;  // OK
    };

—end example]

273. POD classes and operator&()

[Moved to DR at October 2002 meeting.]

I think that the definition of a POD class in the current version of the Standard is overly permissive in that it allows for POD classes for which a user-defined operator function operator& may be defined. Given that the idea behind POD classes was to achieve compatibility with C structs and unions, this makes 'Plain old' structs and unions behave not quite as one would expect them to.

In the C language, if x and y are variables of struct or union type S that has a member m, the following expression are allowed: &x, x.m, x = y. While the C++ standard guarantees that if x and y are objects of a POD class type S, the expressions x.m, x = y will have the same effect as they would in C, it is still possible for the expression &x to be interpreted differently, subject to the programmer supplying an appropriate version of a user-defined operator function operator& either as a member function or as a non-member function.

This may result in surprising effects. Consider:

    // POD_bomb is a POD-struct. It has no non-static non-public data members,
    // no virtual functions, no base classes, no constructors, no user-defined
    // destructor, no user-defined copy assignment operator, no non-static data
    // members of type pointer to member, reference, non-POD-struct, or
    // non-POD-union.
    struct  POD_bomb  {
       int   m_value1;
       int   m_value2;
       int  operator&()
       {   return  m_value1++;   }
       int  operator&() const
       {   return  m_value1 + m_value2;   }
    };

3.9  basic.types paragraph 2 states:

For any complete POD object type T, whether or not the object holds a valid value of type T, the underlying bytes (1.7  intro.memory) making up the object can be copied into an array of char or unsigned char [footnote: By using, for example, the library functions (17.4.1.2  lib.headers) memcpy or memmove]. If the content of the array of char or unsigned char is copied back into the object, the object shall subsequently hold its original value. [Example:

    #define N sizeof(T)
    char buf[N];
    T obj;   // obj initialized to its original value
    memcpy(buf, &obj, N);
		// between these two calls to memcpy,
		// obj might be modified
    memcpy(&obj, buf, N);
		// at this point, each subobject of obj of scalar type
		// holds its original value

—end example]

Now, supposing that the complete POD object type T in the example above is POD_bomb, and we cannot any more count on the assertions made in the comments to the example. Given a standard conforming implementation, the code will not even compile. And I see no legal way of copying the contents of an object of a complete object type POD_bomb into an array of char or unsigned char with memcpy or memmove without making use of the unary & operator. Except, of course, by means of an ugly construct like:

    struct  POD_without_ampersand  {
       POD_bomb   a_bomb;
    }  obj;
    #define N sizeof(POD_bomb)
    char buf[N];
    memcpy(buf, &obj, N);
    memcpy(&obj, buf, N);

The fact that the definition of a POD class allows for POD classes for which a user-defined operator& is defined, may also present major obstacles to implementers of the offsetof macro from <cstddef>

18.1  lib.support.types paragraph 5 says:

The macro offsetof accepts a restricted set of type arguments in this International Standard. type shall be a POD structure or a POD union (clause 9  class). The result of applying the offsetof macro to a field that is a static data member or a function is undefined."

Consider a well-formed C++ program below:

    #include <cstddef>
    #include <iostream>


    struct  POD_bomb  {
       int   m_value1;
       int   m_value2;
       int  operator&()
       {   return  m_value1++;   }
       int  operator&() const
       {   return  m_value1 + m_value2;   }
    };


    // POD_struct is a yet another example of a POD-struct.
    struct  POD_struct  {
       POD_bomb   m_nonstatic_bomb1;
       POD_bomb   m_nonstatic_bomb2;
    };


    int  main()
    {

       std::cout << "offset of m_nonstatic_bomb2: " << offsetof(POD_struct,
           m_nonstatic_bomb2) << '\n';
       return  0;

    }

See Jens Maurer's paper 01-0038=N1324 for an analysis of this issue.

Notes from 10/01 meeting:

A consensus was forming around the idea of disallowing operator& in POD classes when it was noticed that it is permitted to declare global-scope operator& functions, which cause the same problems. After more discussion, it was decided that such functions should not be prohibited in POD classes, and implementors should simply be required to "get the right answer" in constructs such as offsetof and va_start that are conventionally implemented using macros that use the "&" operator. It was noted that one can cast the original operand to char & to de-type it, after which one can use the built-in "&" safely.

Proposed resolution:

• Add a footnote in 18.1  lib.support.types paragraph 5:

  [Footnote: Note that offsetof is required to work as specified even if unary operator& is overloaded for any of the types involved.]

• Add a footnote in 18.7  lib.support.runtime paragraph 3:

  [Footnote: Note that va_start is required to work as specified even if unary operator& is overloaded for the type of parmN.]

284. qualified-ids in class declarations

[Moved to DR at October 2002 meeting.]

Although 8.3  dcl.meaning requires that a declaration of a qualified-id refer to a member of the specified namespace or class and that the member not have been introduced by a using-declaration, it applies only to names declared in a declarator. It is not clear whether there is existing wording enforcing the same restriction for qualified-ids in class-specifiers and elaborated-type-specifiers or whether additional wording is required. Once such wording is found/created, the proposed resolution of issue 275 must be modified accordingly.

Notes from 10/01 meeting:

The sentiment was that this should be required on class definitions, but not on elaborated type specifiers in general (which are references, not declarations). We should also make sure we consider explicit instantiations, explicit specializations, and friend declarations.

Proposed resolution (10/01):

Add after the end of 9.1  class.name paragraph 3:

When a nested-name-specifier is specified in a class-head or in an elaborated-type-specifier, the resulting qualified name shall refer to a previously declared member of the class or namespace to which the nested-name-specifier refers, and the member shall not have been introduced by a using-declaration in the scope of the class or namespace nominated by the nested-name-specifier.

379. Change "class declaration" to "class definition"

[Voted into WP at March 2004 meeting.]

The ARM used the term "class declaration" to refer to what would usually be termed the definition of the class. The standard now often uses "class definition", but there are some surviving uses of "class declaration" with the old meaning. They should be found and changed.

Proposed resolution (April 2003):

Replace in 3.1  basic.def paragraph 2

A declaration is a definition unless it declares a function without specifying the function's body (8.4  dcl.fct.def), it contains the extern specifier (7.1.1  dcl.stc) or a linkage-specification [Footnote: Appearing inside the braced-enclosed declaration-seq in a linkage-specification does not affect whether a declaration is a definition. --- end footnote] (7.5  dcl.link) and neither an initializer nor a function-body, it declares a static data member in a class declaration definition (9.4  class.static), it is a class name declaration (9.1  class.name), or it is a typedef declaration (7.1.3  dcl.typedef), a using-declaration (7.3.3  namespace.udecl), or a using-directive (7.3.4  namespace.udir).

Replace in 7.1.2  dcl.fct.spec paragraphs 5 and 6

The virtual specifier shall only be used in declarations of nonstatic class member functions that appear within a member-specification of a class declaration definition; see 10.3  class.virtual.

The explicit specifier shall be used only in declarations of constructors within a class declaration definition; see 12.3.1  class.conv.ctor.

Replace in 7.1.3  dcl.typedef paragraph 4

A typedef-name that names a class is a class-name (9.1  class.name). If a typedef-name is used following the class-key in an elaborated-type-specifier (7.1.5.3  dcl.type.elab) or in the class-head of a class declaration definition (9  class), or is used as the identifier in the declarator for a constructor or destructor declaration (12.1  class.ctor, 12.4  class.dtor), the program is ill-formed.

Replace in 7.3.1.2  namespace.memdef paragraph 3

The name of the friend is not found by simple name lookup until a matching declaration is provided in that namespace scope (either before or after the class declaration definition granting friendship).

Replace in 8.3.2  dcl.ref paragraph 4

There shall be no references to references, no arrays of references, and no pointers to references. The declaration of a reference shall contain an initializer (8.5.3  dcl.init.ref) except when the declaration contains an explicit extern specifier (7.1.1  dcl.stc), is a class member (9.2  class.mem) declaration within a class declaration definition, or is the declaration of a parameter or a return type (8.3.5  dcl.fct); see 3.1  basic.def.

Replace in 8.5.3  dcl.init.ref paragraph 3

The initializer can be omitted for a reference only in a parameter declaration (8.3.5  dcl.fct), in the declaration of a function return type, in the declaration of a class member within its class declaration definition (9.2  class.mem), and where the extern specifier is explicitly used.

Replace in 9.1  class.name paragraph 2

A class definition declaration introduces the class name into the scope where it is defined declared and hides any class, object, function, or other declaration of that name in an enclosing scope (3.3  basic.scope). If a class name is declared in a scope where an object, function, or enumerator of the same name is also declared, then when both declarations are in scope, the class can be referred to only using an elaborated-type-specifier (3.4.4  basic.lookup.elab).

Replace in 9.4  class.static paragraph 4

Static members obey the usual class member access rules (clause 11  class.access). When used in the declaration of a class member, the static specifier shall only be used in the member declarations that appear within the member-specification of the class declaration definition.

Replace in 9.7  class.nest paragraph 1

A class can be defined declared within another class. A class defined declared within another is called a nested class. The name of a nested class is local to its enclosing class. The nested class is in the scope of its enclosing class. Except by using explicit pointers, references, and object names, declarations in a nested class can use only type names, static members, and enumerators from the enclosing class.

Replace in 9.8  class.local paragraph 1

A class can be defined declared within a function definition; such a class is called a local class. The name of a local class is local to its enclosing scope. The local class is in the scope of the enclosing scope, and has the same access to names outside the function as does the enclosing function. Declarations in a local class can use only type names, static variables, extern variables and functions, and enumerators from the enclosing scope.

Replace in 10  class.derived paragraph 1

... The class-name in a base-specifier shall not be an incompletely defined class (clause 9  class); this class is called a direct base class for the class being declared defined. During the lookup for a base class name, non-type names are ignored (3.3.7  basic.scope.hiding). If the name found is not a class-name, the program is ill-formed. A class B is a base class of a class D if it is a direct base class of D or a direct base class of one of D's base classes. A class is an indirect base class of another if it is a base class but not a direct base class. A class is said to be (directly or indirectly) derived from its (direct or indirect) base classes. [Note: See clause 11  class.access for the meaning of access-specifier.] Unless redefined redeclared in the derived class, members of a base class are also considered to be members of the derived class. The base class members are said to be inherited by the derived class. Inherited members can be referred to in expressions in the same manner as other members of the derived class, unless their names are hidden or ambiguous (10.2  class.member.lookup). [Note: the scope resolution operator :: (5.1  expr.prim) can be used to refer to a direct or indirect base member explicitly. This allows access to a name that has been redefined redeclared in the derived class. A derived class can itself serve as a base class subject to access control; see 11.2  class.access.base. A pointer to a derived class can be implicitly converted to a pointer to an accessible unambiguous base class (4.10  conv.ptr). An lvalue of a derived class type can be bound to a reference to an accessible unambiguous base class (8.5.3  dcl.init.ref).]

Replace in 10.1  class.mi paragraph 5

For another example,

class V { /* ... */ };
class A : virtual public V { /* ... */ };
class B : virtual public V { /* ... */ };
class C : public A, public B { /* ... */ };

for an object c of class type C, a single subobject of type V is shared by every base subobject of c that is declared to have has a virtual base class of type V.

Replace in the example in 10.2  class.member.lookup paragraph 6 (the whole paragraph was turned into a note by the resolution of core issue 39)

The names defined declared in V and the left hand instance of W are hidden by those in B, but the names defined declared in the right hand instance of W are not hidden at all.

Replace in 10.4  class.abstract paragraph 2

... A virtual function is specified pure by using a pure-specifier (9.2  class.mem) in the function declaration in the class declaration definition. ...

Replace in the footnote at the end of 11.2  class.access.base paragraph 1

[Footnote: As specified previously in clause 11  class.access, private members of a base class remain inaccessible even to derived classes unless friend declarations within the base class declaration definition are used to grant access explicitly.]

Replace in 11.3  class.access.dcl paragraph 1

The access of a member of a base class can be changed in the derived class by mentioning its qualified-id in the derived class declaration definition. Such mention is called an access declaration. ...

Replace in the example in 13.4  over.over paragraph 5

The initialization of pfe is ill-formed because no f() with type int(...) has been defined declared, and not because of any ambiguity.

Replace in C.1.5  diff.dcl paragraph 1

Rationale: Storage class specifiers don't have any meaning when associated with a type. In C++, class members can be defined declared with the static storage class specifier. Allowing storage class specifiers on type declarations could render the code confusing for users.

Replace in C.1.7  diff.class paragraph 3

In C++, a typedef name may not be redefined redeclared in a class declaration definition after being used in the declaration that definition

The resolution of core issue 45 (DR) deletes 11.8  class.access.nest paragraph 2.

The following occurrences of "class declaration" are not changed:

• 5.3.4  expr.new paragraph 4
• 7.1.3  dcl.typedef paragraph 1
• 9.1  class.name paragraphs 3 and 4
• 11.4  class.friend paragraph 9
• C.1.7  diff.class paragraph 1

383. Is a class with a declared but not defined destructor a POD?

[Voted into WP at March 2004 meeting.]

The standard (9  class par. 4) says that "A POD-struct is an aggregate class that has no non-static data members of type non-POD-struct, non-POD-union (or array of such types) or reference, and has no user-defined copy assignment operator and no user-defined destructor."

Note that it says 'user-defined', not 'user-declared'. Is it the intent that if e.g. a copy assignment operator is declared but not defined, this does not (per se) prevent the class to be a POD-struct?

Proposed resolution (April 2003):

Replace in 9  class paragraph 4

A POD-struct is an aggregate class that has no non-static data members of type non-POD-struct, non-POD-union (or array of such types) or reference, and has no user-defined declared copy assignment operator and no user-defined declared destructor. Similarly, a POD-union is an aggregate union that has no non-static data members of type non-POD-struct, non-POD-union (or array of such types) or reference, and has no user-defined declared copy assignment operator and no user-defined declared destructor.

Drafting note: The changes are shown relative to TC1, incorporating the changes from the resolution of core issue 148.

417. Using derived-class qualified name in out-of-class nested class definition

[Voted into WP at October 2004 meeting.]

We had a user complain that our compiler was allowing the following code:

  struct B {
    struct S;
  };

  struct D : B { };

  struct D::S {
  };

We took one look at the code and made the reasonable (I would claim) assumption that this was indeed a bug in our compiler. Especially as we had just fixed a very similar issue with the definition of static data members.

Imagine our surprise when code like this showed up in Boost and that every other compiler we tested accepts this code. So is this indeed legal (it seems like it must be) and if so is there any justification for this beyond 3.4.3.1  class.qual?

John Spicer: The equivalent case for a member function is covered by the declarator rules in 8.3  dcl.meaning paragraph 1. The committee has previously run into cases where a restriction should apply to both classes and non-classes, but fails to do so because there is no equivalent of 8.3  dcl.meaning paragraph 1 for classes.

Given that, by the letter of the standard, I would say that this case is allowed.

Notes from October 2003 meeting:

We feel this case should get an error.

Proposed Resolution (October 2003):

Note that the change here interacts with issue 432.

Add the following as a new paragraph immediately following 3.3.1  basic.scope.pdecl paragraph 2:

The point of declaration for a class first declared by a class-specifier is immediately after the identifier or template-id (if any) in its class-head (Clause 9  class). The point of declaration for an enumeration is immediately after the identifier (if any) in its enum-specifier (7.2  dcl.enum).

Change point 1 of 3.3.6  basic.scope.class paragraph 1 to read:

The potential scope of a name declared in a class consists not only of the declarative region following the name's declarator point of declaration, but also of all function bodies, default arguments, and constructor ctor-initializers in that class (including such things in nested classes).

[Note that the preceding change duplicates one of the changes in the proposed resolution of issue 432.]

Change 14.7.2  temp.explicit paragraph 2 to read:

If the explicit instantiation is for a member function, a member class or a static data member of a class template specialization, the name of the class template specialization in the qualified-id for the member declarator name shall be a template-id.

Add the following as paragraph 5 of Clause 9  class:

If a class-head contains a nested-name-specifier, the class-specifier shall refer to a class that was previously declared directly in the class or namespace to which the nested-name-specifier refers (i.e., neither inherited nor introduced by a using-declaration), and the class-specifier shall appear in a namespace enclosing the previous declaration.

Delete 9.1  class.name paragraph 4 (this was added by issue 284):

When a nested-name-specifier is specified in a class-head or in an elaborated-type-specifier, the resulting qualified name shall refer to a previously declared member of the class or namespace to which the nested-name-specifier refers, and the member shall not have been introduced by a using-declaration in the scope of the class or namespace nominated by the nested-name-specifier.

328. Missing requirement that class member types be complete

[Voted into WP at March 2004 meeting.]

Is it legal to use an incomplete type (3.9  basic.types paragraph 6) as a class member, if no object of such class is ever created ?

And as a class template member, even if the template is instantiated, but no object of the instantiated class is created?

The consensus seems to be NO, but no wording was found in the standard which explicitly disallows it.

The problem seems to be that most of the restrictions on incomplete types are on their use in objects, but class members are not objects.

A possible resolution, if this is considered a defect, is to add to 3.2  basic.def.odr paragraph 4, (situations when T must be complete), the use of T as a member of a class or instantiated class template.

The thread on comp.std.c++ which brought up the issue was "Compiler differences: which is correct?", started 2001 11 30. <3c07c8fb$0$8507$ed9e5944@reading.news.pipex.net>

Proposed Resolution (April 2002, revised April 2003):

Change the first bullet of the note in 3.2  basic.def.odr paragraph 4 and add two new bullets following it, as follows:

• an object of type T is defined (3.1  basic.def, 5.3.4  expr.new), or
• a non-static class data member of type T is declared (9.2  class.mem), or
• T is used as the object type or array element type in a new-expression (5.3.4  expr.new), or

Replace 9.2  class.mem paragraph 8 by:

Non-static (9.4  class.static) data members shall not have incomplete types. In particular, a class C shall not contain a non-static member of class C, but it can contain a pointer or reference to an object of class C.

See also 3.9  basic.types paragraph 6, which is relevant but not changed by the Proposed Resolution.

437. Is type of class allowed in member function exception specification?

[Voted into WP at April 2005 meeting.]

I've encountered a C++ program in which a member function wants to declare that it may throw an object of its own class, e.g.:

  class Foo {
  private:
     int val;
  public:
     Foo( int &initval ) { val = initval; };
     void throwit() throw(Foo) { throw (*this); };
  };

The compiler is complaining that Foo is an incomplete type, and can't be used in the exception specification.

My reading of the standard [basic.types] is inconclusive. Although it does state that the class declaration is considered complete when the closing brace is read, I believe it also intends that the member function declarations should not be semantically validated until the class has been completely declared.

If this isn't allowed, I don't know how else a member function could be declared to throw an object of its own class.

John Spicer: The type is considered complete within function bodies, but not in their declaration (see 9.2  class.mem paragraph 2).

Proposed Resolution:

Change 9.2  class.mem paragraph 2 as follows:

Within the class member-specification, the class is regarded as complete within function bodies, default arguments, exception-specifications, and constructor ctor-initializers (including such things in nested classes).

Rationale: Taken with 8.3.5  dcl.fct paragraph 6, the exception-specification is the only part of a function declaration/definition in which the class name cannot be used because of its putative incompleteness. There is no justification for singling out exception specifications this way; both in the function body and in a catch clause, the class type will be complete, so there is no harm in allowing the class name to be used in the exception-specification.

406. Static data member in class with name for linkage purposes

[Voted into WP at March 2004 meeting.]

The following test program is claimed to be a negative C++ test case for "Unnamed classes shall not contain static data members", c.f. ISO/IEC 14882 section 9.4.2  class.static.data paragraph 5.

 
  struct B {
         typedef struct {
                 static int i;          // Is this legal C++ ?
         } A;
  };
 
  int B::A::i = 47;      // Is this legal C++ ?

We are not quite sure about what an "unnamed class" is. There is no exact definition in ISO/IEC 14882; the closest we can come to a hint is the wording of section 7.1.3  dcl.typedef paragraph 5, where it seems to be understood that a class-specifier with no identifier between "class" and "{" is unnamed. The identifier provided after "}" ( "A" in the test case above) is there for "linkage purposes" only.

To us, class B::A in the test program above seems "named" enough, and there is certainly a mechanism to provide the definition for B::A::i (in contrast to the note in section 9.4.2  class.static.data paragraph 5).

Our position is therefore that the above test program is indeed legal C++. Can you confirm or reject this claim?

Herb Sutter replied to the submitter as follows: Here are my notes based on a grep for "unnamed class" in the standard:

• 3.5  basic.link paragraph 4, bullet 3, makes a note that does not directly speak to your question but which draws the same distinction:

  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 (7.1.3  dcl.typedef);

  Likewise in your example, you have an unnamed class defined in a typedef declaration.
• 9.4.2  class.static.data paragraph 5 does indeed appear to me to make your example not supported by ISO C++ (although implementations could allow it as an extension, and many implementation do happen to allow it).
• 7.1.3  dcl.typedef paragraph 5 does indeed likewise confirm the interpretation you give of an unnamed class.

So yes, an unnamed class is one where there is no identifier (class name) between the class-key and the {. This is also in harmony with the production for class-name in 9  class paragraph 1:

class-name:
identifier
template-id

Notes from the October 2003 meeting:

We agree that the example is not valid; this is an unnamed class. We will add wording to define an unnamed class. The note in 9.4.2  class.static.data paragraph 5 should be corrected or deleted.

Proposed Resolution (October 2003):

At the end of clause 9  class, paragraph 1, add the following:

A class-specifier where the class-head omits the optional identifier defines an unnamed class.

Delete the following from 9.4.2  class.static.data paragraph 5:

[ Note: this is because there is no mechanism to provide the definitions for such static data members. ]

436. Problem in example in 9.6 paragraph 4

[Voted into WP at October 2004 meeting.]

It looks like the example on 9.6  class.bit paragraph 4 has both the enum and function contributing the identifier "f" for the same scope.

  enum BOOL { f=0, t=1 };
  struct A {
    BOOL b:1;
  };
  A a;
  void f() {
    a.b = t;
    if (a.b == t) // shall yield true
    { /* ... */ }
  }

Proposed resolution:

Change the example at the end of 9.6  class.bit/4 from:

  enum BOOL { f=0, t=1 };
  struct A {
    BOOL b:1;
  };
  A a;
  void f() {
    a.b = t;
    if (a.b == t) // shall yield true
    { /* ... */ }
  }

To:

  enum BOOL { FALSE=0, TRUE=1 };
  struct A {
    BOOL b:1;
  };
  A a;
  void f() {
    a.b = TRUE;
    if (a.b == TRUE) // shall yield true
    { /* ... */ }
  }

198. Definition of "use" in local and nested classes

[Voted into WP at April 2003 meeting.]

9.8  class.local paragraph 1 says,

Declarations in a local class can use only type names, static variables, extern variables and functions, and enumerators from the enclosing scope.

This definition of "use" would presumably allow code like

    void foo() {
        int i;
        struct S {
            int a[sizeof(i)];
        };
    };

However, the restrictions on nested classes in 9.7  class.nest paragraph 1 are very similar to those for local classes, and the example there explicitly states that a reference in a sizeof expression is a forbidden use (abbreviated for exposition):

    class enclose {
    public:
        int x;
        class inner {
            void f(int i)
            {
                int a = sizeof(x);  // error: refers to enclose::x
            }
        };
    };

[As a personal note, I have seen real-world code that was exactly like this; it was hard to persuade the author that the required writearound, sizeof(((enclose*) 0)->x), was an improvement over sizeof(x). —wmm]

Similarly, 9.2  class.mem paragraph 9 would appear to prohibit examples like the following:

    struct B {
        char x[10];
    };
    struct D: B {
        char y[sizeof(x)];
    };

Suggested resolution: Add cross-references to 3.2  basic.def.odr following the word "use" in both 9.7  class.nest and 9.8  class.local , and change the example in 9.7  class.nest to indicate that a reference in a sizeof expression is permitted. In 9.2  class.mem paragraph 9, "referred to" should be changed to "used" with a cross_reference to 3.2  basic.def.odr.

Notes from 10/01 meeting:

It was noted that the suggested resolution did not make the sizeof() example in 9.7  class.nest valid. Although the reference to the argument of sizeof() is not regarded as a use, the right syntax must be used nonetheless to reference a non-static member from the enclosing class. The use of the member name by itself is not valid. The consensus within the core working group was that nothing should be done about this case. It was later discovered that 9.4  class.static paragraph 3 states that

The definition of a static member shall not use directly the names of the nonstatic members of its class or of a base class of its class (including as operands of the sizeof operator). The definition of a static member may only refer to these members to form pointer to members (5.3.1  expr.unary.op) or with the class member access syntax (5.2.5  expr.ref).

This seems to reinforce the decision of the working group.

The use of "use" should still be cross-referenced. The statements in 9.7  class.nest and 9.8  class.local should also be rewritten to state the requirement positively rather than negatively as the list of "can't"s is already missing some cases such as template parameters.

Notes from the 4/02 meeting:

We backed away from "use" in the technical sense, because the requirements on the form of reference are the same whether or not the reference occurs inside a sizeof.

Proposed Resolution (revised October 2002):

In 9.2  class.mem paragraph 9, replace

Except when used to form a pointer to member (5.3.1  expr.unary.op), when used in the body of a nonstatic member function of its class or of a class derived from its class (9.3.1  class.mfct.nonstatic), or when used in a mem-initializer for a constructor for its class or for a class derived from its class (12.6.2  class.base.init), a nonstatic data or function member of a class shall only be referred to with the class member access syntax (5.2.5  expr.ref).

with the following paragraph

Each occurrence in an expression of the name of a nonstatic data member or nonstatic member function of a class shall be expressed as a class member access (5.2.5  expr.ref), except when it appears in the formation of a pointer to member (5.3.1  expr.unary.op), when it appears in the body of a nonstatic member function of its class or of a class derived from its class (9.3.1  class.mfct.nonstatic), or when it appears in a mem-initializer for a constructor for its class or for a class derived from its class (12.6.2  class.base.init).

In 9.7  class.nest paragraph 1, replace the last sentence,

Except by using explicit pointers, references, and object names, declarations in a nested class can use only type names, static members, and enumerators from the enclosing class.

with the following

[Note: In accordance with 9.2  class.mem, except by using explicit pointers, references, and object names, declarations in a nested class shall not use nonstatic data members or nonstatic member functions from the enclosing class. This restriction applies in all constructs including the operands of the sizeof operator.]

In the example following 9.7  class.nest paragraph 1, change the comment on the first statement of function f to emphasize that sizeof(x) is an error. The example reads in full:

  int x;
  int y;
  class enclose {
  public:
    int x;
    static int s;
    class inner {
      void f(int i)
      {
        int a = sizeof(x);  // error: direct use of enclose::x even in sizeof
        x = i;              // error: assign to enclose::x
        s = i;              // OK: assign to enclose::s
        ::x = i;            // OK: assign to global x
        y = i;              // OK: assign to global y
      }
      void g(enclose* p, int i)
      {
        p->x = i;        // OK: assign to enclose::x
      }
    };
  };
   
  inner* p = 0;             // error: inner not in scope

39. Conflicting ambiguity rules

[Voted into WP at April 2005 meeting.]

The ambiguity text in 10.2  class.member.lookup may not say what we intended. It makes the following example ill-formed:

    struct A {
        int x(int);
    };
    struct B: A {
        using A::x;
        float x(float);
    };
    
    int f(B* b) {
        b->x(3);  // ambiguous
    }

... Each of these declarations that was introduced by a using-declaration is considered to be from each sub-object of C that is of the type containing the declaration designated by the using-declaration. If the resulting set of declarations are not all from sub-objects of the same type, or the set has a nonstatic member and includes members from distinct sub-objects, there is an ambiguity and the program is ill-formed.

Proposed Resolution (10/00):

1. Replace the two cited sentences from 10.2  class.member.lookup paragraph 2 with the following:

   The resulting set of declarations shall all be from sub-objects of the same type, or there shall be a set of declarations from sub-objects of a single type that contains using-declarations for the declarations found in all other sub-object types. Furthermore, for nonstatic members, the resulting set of declarations shall all be from a single sub-object, or there shall be a set of declarations from a single sub-object that contains using-declarations for the declarations found in all other sub-objects. Otherwise, there is an ambiguity and the program is ill-formed.

2. Replace the examples in 10.2  class.member.lookup paragraph 3 with the following:

       struct A {
           int x(int);
           static int y(int);
       };
       struct V {
           int z(int);
       };
       struct B: A, virtual V {
           using A::x;
           float x(float);
           using A::y;
           static float y(float);
           using V::z;
           float z(float);
       };
       struct C: B, A, virtual V {
       };
   
       void f(C* c) {
           c->x(3);    // ambiguous -- more than one sub-object A
           c->y(3);    // not ambiguous
           c->z(3);    // not ambiguous
       }
   

Notes from 04/01 meeting:

The following example should be accepted but is rejected by the wording above:

    struct A { static void f(); };

    struct B1: virtual A {
        using A::f;
    };

    struct B2: virtual A {
        using A::f;
    };

    struct C: B1, B2 { };

    void g() {
        C::f();        // OK, calls A::f()
    }

Notes from 10/01 meeting (Jason Merrill):

The example in the issues list:

    struct A {
        int x(int);
    };
    struct B: A {
        using A::x;
        float x(float);
    };
    
    int f(B* b) {
        b->x(3);  // ambiguous
    }

... Each of these declarations that was introduced by a using-declaration is considered to be from each sub-object of C that is of the type containing the declaration designated by the using-declaration. If the resulting set of declarations are not all from sub-objects of the same type, or the set has a nonstatic member and includes members from distinct sub-objects, there is an ambiguity and the program is ill-formed.

The first proposed solution:

The resulting set of declarations shall all be from sub-objects of the same type, or there shall be a set of declarations from sub-objects of a single type that contains using-declarations for the declarations found in all other sub-object types. Furthermore, for nonstatic members, the resulting set of declarations shall all be from a single sub-object, or there shall be a set of declarations from a single sub-object that contains using-declarations for the declarations found in all other sub-objects. Otherwise, there is an ambiguity and the program is ill-formed.

    struct A { static void f(); };

    struct B1: virtual A {
        using A::f;
    };

    struct B2: virtual A {
        using A::f;
    };

    struct C: B1, B2 { };

    void g() {
        C::f();        // OK, calls A::f()
    }

The solution is to separate the notions of lookup and definition context. I have taken an algorithmic approach to describing the strategy.

Incidentally, the earlier proposal allows one base to have a superset of the declarations in another base; that was an extension, and my proposal does not do that. One algorithmic benefit of this limitation is to simplify the case of a virtual base being hidden along one arm and not another ("domination"); if we allowed supersets, we would need to remember which subobjects had which declarations, while under the following resolution we need only keep two lists, of subobjects and declarations.

Proposed resolution (October 2002):

Replace 10.2  class.member.lookup paragraph 2 with:

The following steps define the result of name lookup for a member name f in a class scope C.

The lookup set for f in C, called S(f,C), consists of two component sets: the declaration set, a set of members named f; and the subobject set, a set of subobjects where declarations of these members (possibly including using-declarations) were found. In the declaration set, using-declarations are replaced by the members they designate, and type declarations (including injected-class-names) are replaced by the types they designate. S(f,C) is calculated as follows.

If C contains a declaration of the name f, the declaration set contains every declaration of f in C (excluding bases), the subobject set contains C itself, and calculation is complete.

Otherwise, S(f,C) is initially empty. If C has base classes, calculate the lookup set for f in each direct base class subjobject B_i, and merge each such lookup set S(f,B_i) in turn into S(f,C).

The following steps define the result of merging lookup set S(f,B_i) into the intermediate S(f,C):

• If each of the subobject members of S(f,B_i) is a base class subobject of at least one of the subobject members of S(f,C), S(f,C) is unchanged and the merge is complete. Conversely, if each of the subobject members of S(f,C) is a base class subobject of at least one of the subobject members of S(f,B_i), the new S(f,C) is a copy of S(f,B_i).
• Otherwise, if the declaration sets of S(f,B_i) and S(f,C) differ, the merge is ambiguous: the new S(f,C) is a lookup set with an invalid declaration set and the union of the subobject sets. In subsequent merges, an invalid declaration set is considered different from any other.
• Otherwise, consider each declaration d in the set, where d is a member of class A. If d is a nonstatic member, compare the A base class subobjects of the subobject members of S(f,B_i) and S(f,C). If they do not match, the merge is ambiguous, as in the previous step. [Note: It is not necessary to remember which A subobject each member comes from, since using-declarations don't disambiguate. ]
• Otherwise, the new S(f,C) is a lookup set with the shared set of declarations and the union of the subobject sets.

The result of name lookup for f in C is the declaration set of S(f,C). If it is an invalid set, the program is ill-formed.

[Example:

    struct A { int x; };                    // S(x,A) = {{ A::x }, { A }}
    struct B { float x; };                  // S(x,B) = {{ B::x }, { B }}
    struct C: public A, public B { };       // S(x,C) = { invalid, { A in C, B in C }}
    struct D: public virtual C { };         // S(x,D) = S(x,C)
    struct E: public virtual C { char x; }; // S(x,E) = {{ E::x }, { E }}
    struct F: public D, public E { };       // S(x,F) = S(x,E)

    int main() {
      F f;
      f.x = 0;   // OK, lookup finds { E::x }
    }

S(x,F) is unambiguous because the A and B base subobjects of D are also base subobjects of E, so S(x,D) is discarded in the first merge step. --end example]

Turn 10.2  class.member.lookup paragraphs 5 and 6 into notes.

Notes from October 2003 meeting:

Mike Miller raised some new issues in N1543, and we adjusted the proposed resolution as indicated in that paper.

Further information from Mike Miller (January 2004):

Unfortunately, I've become aware of a minor glitch in the proposed resolution for issue 39 in N1543, so I'd like to suggest a change that we can discuss in Sydney.

A brief review and background of the problem: the major change we agreed on in Kona was to remove detection of multiple-subobject ambiguity from class lookup (10.2  class.member.lookup) and instead handle it as part of the class member access expression. It was pointed out in Kona that 11.2  class.access.base/5 has this effect:

If a class member access operator, including an implicit "this->," is used to access a nonstatic data member or nonstatic member function, the reference is ill-formed if the left operand (considered as a pointer in the "." operator case) cannot be implicitly converted to a pointer to the naming class of the right operand.

After the meeting, however, I realized that this requirement is not sufficient to handle all the cases. Consider, for instance,

    struct B {
        int i;
    };

    struct I1: B { };
    struct I2: B { };

    struct D: I1, I2 {
        void f() {
            i = 0;    // not ill-formed per 11.2p5
        }
    };

Here, both the object expression ("this") and the naming class are "D", so the reference to "i" satisfies the requirement in 11.2  class.access.base/5, even though it involves a multiple-subobject ambiguity.

In order to address this problem, I proposed in N1543 to add a paragraph following 5.2.5  expr.ref/4:

If E2 is a non-static data member or a non-static member function, the program is ill-formed if the class of E1 cannot be unambiguously converted (10.2) to the class of which E2 is directly a member.

That's not quite right. It does diagnose the case above as written; however, it breaks the case where qualification is used to circumvent the ambiguity:

    struct D2: I1, I2 {
        void f() {
            I2::i = 0;    // ill-formed per proposal
        }
    };

In my proposed wording, the class of "this" can't be converted to "B" (the qualifier is ignored), so the access is ill-formed. Oops.

I think the following is a correct formulation, so the proposed resolution we discuss in Sydney should contain the following paragraph instead of the one in N1543:

If E2 is a nonstatic data member or a non-static member function, the program is ill-formed if the naming class (11.2) of E2 cannot be unambiguously converted (10.2) to the class of which E2 is directly a member.

This reformulation also has the advantage of pointing readers to 11.2  class.access.base, where the the convertibility requirement from the class of E1 to the naming class is located and which might otherwise be overlooked.

Notes from the March 2004 meeting:

We discussed this further and agreed with these latest recommendations. Mike Miller has produced a paper N1626 that gives just the final collected set of changes.

(This resolution also resolves isssue 306.)

306. Ambiguity by class name injection

[Voted into WP at April 2005 meeting.]

Is the following well-formed?

    struct A {
        struct B { };
    };
    struct C : public A, public A::B {
        B *p;
    };

What if a struct is found along one path and a typedef to that struct is found along another path? That should probably be valid, but does the standard say so?

This is resolved by issue 39

February 2004: Moved back to "Review" status because issue 39 was moved back to "Review".

390. Pure virtual must be defined when implicitly called

[Voted into WP at March 2004 meeting.]

In clause 10.4  class.abstract paragraph 2, it reads:

A pure virtual function need be defined only if explicitly called with the qualified-id syntax (5.1  expr.prim).

This is IMHO incomplete. A dtor is a function (well, a "special member function", but this also makes it a function, right?) but it is called implicitly and thus without a qualified-id syntax. Another alternative is that the pure virtual function is called directly or indirectly from the ctor. Thus the above sentence which specifies when a pure virtual function need be defined ("...only if...") needs to be extended:

A pure virtual function need be defined only if explicitly called with the qualified-id syntax (5.1  expr.prim) or if implicitly called (12.4  class.dtor or 12.7  class.cdtor).

Proposed resolution:

Change 10.4  class.abstract paragraph 2 from

A pure virtual function need be defined only if explicitly called with the qualified-id syntax (5.1  expr.prim).

to

A pure virtual function need be defined only if explicitly called with, or as if with (12.4  class.dtor), the qualified-id syntax (5.1  expr.prim).

Note: 12.4  class.dtor paragraph 6 defines the "as if" cited.

8. Access to template arguments used in a function return type and in the nested name specifier

[Moved to DR at 4/01 meeting.]

Consider the following example:

    class A {
       class A1{};
       static void func(A1, int);
       static void func(float, int);
       static const int garbconst = 3;
     public:
       template < class T, int i, void (*f)(T, int) > class int_temp {};
       template<> class int_temp<A1, 5, func> { void func1() };
       friend int_temp<A1, 5, func>::func1();
       int_temp<A1, 5, func>* func2();
   };
   A::int_temp<A::A1, A::garbconst + 2, &A::func>* A::func2() {...}

In 11  class.access paragraph 5 we have:

All access controls in clause 11 affect the ability to access a class member name from a particular scope... In particular, access controls apply as usual to member names accessed as part of a function return type, even though it is not possible to determine the access privileges of that use without first parsing the rest of the function declarator.

      A::int_temp
      A::A1
      A::garbconst (part of an expression)
      A::func (after overloading is done)

Rationale:

Not a defect. This behavior is as intended.

ISSUE 2:

Now consider void A::int_temp<A::A1, A::garbconst + 2, &A::func>::func1() {...} By my reading of 11.8  class.access.nest , the references to A::A1, A::garbconst and A::func are now illegal, and there is no way to define this function outside of the class. Is there any need to do anything about either of these Issues?

Proposed resolution (04/01):

The resolution for this issue is contained in the resolution for issue 45.

9. Clarification of access to base class members

[Moved to DR at 4/01 meeting.]

11.2  class.access.base paragraph 4 says:

A base class is said to be accessible if an invented public member of the base class is accessible. If a base class is accessible, one can implicitly convert a pointer to a derived class to a pointer to that base class.

    class D;
     
    class B
    {
     protected:
       int b1;
 
       friend void foo( D* pd );
    };
     
    class D : protected B { };
     
    void foo( D* pd )
    {
       if ( pd->b1 > 0 ); // Is 'b1' accessible?
    }

1st interpretation:

A public member of B is accessible within foo (since foo is a friend), therefore foo can refer to b1 and convert a D* to a B*.

2nd interpretation:

B is a protected base class of D, and a public member of B is a protected member of D and can only be accessed within members of D and friends of D. Therefore foo cannot refer to b1 and cannot convert a D* to a B*.

(See J16/99-0042 = WG21 N1218.)

Proposed Resolution (04/01):

1. Add preceding 11.2  class.access.base paragraph 4:

   A base class B of N is accessible at R, if

   • an invented public member of B would be a public member of N, or
   • R occurs in a member or friend of class N, and an invented public member of B would be a private or protected member of N, or
   • R occurs in a member or friend of a class P derived from N, and an invented public member of B would be a private or protected member of P, or
   • there exists a class S such that B is a base class of S accessible at R and S is a base class of N accessible at R. [Example:

         class B {
         public:
             int m;
         };
     
         class S: private B {
             friend class N;
         };
     
         class N: private S {
             void f() {
     	    B* p = this;  // OK because class S satisfies the
     			// fourth condition above: B is a base
     			// class of N accessible in f() because
     			// B is an accessible base class of S
     			// and S is an accessible base class of N.
             }
         };
     

     —end example]

2. Delete the first sentence of 11.2  class.access.base paragraph 4:

   A base class is said to be accessible if an invented public member of the base class is accessible.

3. Replace the last sentence ("A member m is accessible...") by the following:

   A member m is accessible at the point R when named in class N if

   • m as a member of N is public, or
   • m as a member of N is private, and R occurs in a member or friend of class N, or
   • m as a member of N is protected, and R occurs in a member or friend of class N, or in a member or friend of a class P derived from N, where m as a member of P is private or protected, or
   • there exists a base class B of N that is accessible at R, and m is accessible at R when named in class B. [Example:...

The resolution for issue 207 modifies this wording slightly.

16. Access to members of indirect private base classes

[Moved to DR at 4/01 meeting.]

The text in 11.2  class.access.base paragraph 4 does not seem to handle the following cases:

    class D;
     
    class B {
    private:
        int i;
        friend class D;
    };
     
    class C : private B { };
     
    class D : private C {
        void f() {
            B::i; //1: well-formed?
            i;    //2: well-formed?
        }
    };

Proposed Resolution (04/01): The resolution for this issue is contained in the resolution for issue 9..

207. using-declarations and protected access

[Moved to DR at 10/01 meeting.]

Consider the following example:

  class A {
  protected:
    static void f() {};
  };

  class B : A {
  public:
    using A::f;
    void g() {
      A::f();
    }
  };

The standard says in 11.2  class.access.base paragraph 4 that the call to A::f is ill-formed:

A member m is accessible when named in class N if

• m as a member of N is public, or
• m as a member of N is private, and the reference occurs in a member or friend of class N, or
• m as a member of N is protected, and the reference occurs in a member or friend of class N, or in a member or friend of a class P derived from N, where m as a member of P is private or protected, or
• there exists a base class B of N that is accessible at the point of reference, and m is accessible when named in class B.

Here, m is A::f and N is A.

• f as a member of A is public? No.
• f as a member of A is private? No.
• f as a member of A is protected? Yes.
• reference in a member or friend of A? No.
• reference in a member or friend of a class derived from A? Yes, B.
• f as a member of B private or protected? No, public.
• base of A accessible at point of reference? No.

It seems clear to me that the third bullet should say "public, private or protected".

Steve Adamczyk:The words were written before using-declarations existed, and therefore didn't anticipate this case.

Proposed resolution (04/01):

Modify the third bullet of the third change ("A member m is accessible...") in the resolution of issue 9 to read "public, private, or protected" instead of "private or protected."

77. The definition of friend does not allow nested classes to be friends

[Moved to DR at 4/02 meeting.]

The definition of "friend" in 11.4  class.friend says:

A friend of a class is a function or class that is not a member of the class but is permitted to use the private and protected member names from the class. ...

    class OUTER {
        class INNER;
        friend class INNER;
        class INNER {};
    };

Proposed resolution (04/01):

Change the first sentence of 11.4  class.friend as follows:

A friend of a class is a function or class that is not a member of the class but is allowed given permission to use the private and protected member names from the class. The name of a friend is not in the scope of the class, and the friend is not called with the member access operators (5.2.5  expr.ref) unless it is a member of another class. A class specifies its friends, if any, by way of friend declarations. Such declarations give special access rights to the friends, but they do not make the nominated friends members of the befriending class.

385. How does protected member check of 11.5 interact with using-declarations?

[Voted into WP at October 2004 meeting.]

We consider it not unreasonable to do the following

  class A { 
    protected: 
    void g();
  }; 
  class B : public A { 
    public: 
      using A::g; // B::g is a public synonym for A::g 
  }; 

  class C: public A {
    void foo();
  };

  void C::foo() { 
    B b; 
    b.g(); 
  } 

However the EDG front-end does not like and gives the error

  #410-D: protected function "A::g" is not accessible through a "B" pointer or  object 
    b.g();
      ^

Steve Adamczyk: The error in this case is due to 11.5  class.protected of the standard, which is an additional check on top of the other access checking. When that section says "a protected nonstatic member function ... of a base class" it doesn't indicate whether the fact that there is a using-declaration is relevant. I'd say the current wording taken at face value would suggest that the error is correct -- the function is protected, even if the using-declaration for it makes it accessible as a public function. But I'm quite sure the wording in 11.5  class.protected was written before using-declarations were invented and has not been reviewed since for consistency with that addition.

Notes from April 2003 meeting:

We agreed that the example should be allowed.

Proposed resolution (April 2003, revised October 2003):

Change 11.5  class.protected paragraph 1 from

When a friend or a member function of a derived class references a protected nonstatic member function or protected nonstatic data member of a base class, an access check applies in addition to those described earlier in clause 11  class.access. [Footnote: This additional check does not apply to other members, e.g. static data members or enumerator member constants.] Except when forming a pointer to member (5.3.1  expr.unary.op), the access must be through a pointer to, reference to, or object of the derived class itself (or any class derived from that class (5.2.5  expr.ref). If the access is to form a pointer to member, the nested-name-specifier shall name the derived class (or any class derived from that class).

to

An additional access check beyond those described earlier in clause 11  class.access is applied when a nonstatic data member or nonstatic member function is a protected member of its naming class (11.2  class.access.base). [Footnote: This additional check does not apply to other members, e.g., static data members or enumerator member constants.] As described earlier, access to a protected member is granted because the reference occurs in a friend or member of some class C. If the access is to form a pointer to member (5.3.1  expr.unary.op), the nested-name-specifier shall name C or a class derived from C. All other accesses involve a (possibly implicit) object expression (5.2.5  expr.ref). In this case, the class of the object expression shall be C or a class derived from C.

Additional discussion (September, 2004):

Steve Adamczyk: I wonder if this wording is incorrect. Consider:

    class A {
    public:
      int p;
    };
    class B : protected A {
      // p is a protected member of B
    };
    class C : public B {
      friend void fr();
    };
    void fr() {
      B *pb = new B;
      pb->p = 1;  // Access okay?  Naming class is B, p is a protected member of B,
                  // the "C" of the issue 385 wording is C, but access is not via
                  // an object of type C or a derived class thereof.
    }

I think the formulation that the member is a protected member of its naming class is not what we want. I think we intended that the member is protected in the declaration that is found, where the declaration found might be a using-declaration.

Mike Miller: I think the proposed wording makes the access pb->p ill-formed, and I think that's the right thing to do.

First, protected inheritance of A by B means that B intends the public and protected members of A to be part of B's implementation, available to B's descendants only. (That's why there's a restriction on converting from B* to A*, to enforce B's intention on the use of members of A.) Consequently, I see no difference in access policy between your example and

    class B {
    protected:
        int p;
    };

Second, the reason we have this rule is that C's use of inherited protected members might be different from their use in a sibling class, say D. Thus members and friends of C can only use B::p in a manner consistent with C's usage, i.e., in C or derived-from-C objects. If we rewrote your example slightly,

    class D: public B { };

    void fr(B* pb) {
        pb->p = 1;
    }

    void g() {
        fr(new D);
    }

it's clear that the intent of this rule is broken — fr would be accessing B::p assuming C's policies when the object in question actually required D's policies.

(See also issues 471 and 472.)

10. Can a nested class access its own class name as a qualified name if it is a private member of the enclosing class?

[Moved to DR at 4/01 meeting.]

Paragraph 1 says: "The members of a nested class have no special access to members of an enclosing class..."

This prevents a member of a nested class from being defined outside of its class definition. i.e. Should the following be well-formed?

    class D {
        class E {
            static E* m;
        };
    };
     
    D::E* D::E::m = 1; // ill-formed

Here is another example:

    class C {
        class B
        {
            C::B *t; //2 error, C::B is inaccessible
        };
    };

    class C {
        class B
        {
            B& operator= (const B&);
        };
    };
     
    C::B& C::B::operator= (const B&) { } //3

Proposed resolution (04/01):

The resolution for this issue is incorporated into the resolution for issue 45.

45. Access to nested classes

[Moved to DR at 4/01 meeting.]

Example:

    #include <iostream.h>
    
    class C {  // entire body is private
        struct Parent {
            Parent() { cout << "C::Parent::Parent()\n"; }
        };
    
        struct Derived : Parent {
            Derived() { cout << "C::Derived::Derived()\n"; }
        };
    
        Derived d;
    };
    
    
    int main() {
        C c;      //  Prints message from both nested classes
        return 0;
    }

11  class.access paragraph 1:

A member of a class can be

• private; that is, its name can be used only by members and friends of the class in which it is declared. [...]

The members of a nested class have no special access to members of an enclosing class, nor to classes or functions that have granted friendship to an enclosing class; the usual access rules (clause 11  class.access ) shall be obeyed. [...]

From Mike Miller:

I think it is completely legal, by the reasoning given in the (non-normative) 11.8  class.access.nest paragraph 2. The use of a private nested class as a base of another nested class is explicitly declared to be acceptable there. I think the rationale in the comments in the example ("// OK because of injection of name A in A") presupposes that public members of the base class will be public members in a (publicly-derived) derived class, regardless of the access of the base class, so the constructor invocation should be okay as well.

I can't find anything normative that explicitly says that, though.

(See also papers J16/99-0009 = WG21 N1186, J16/00-0031 = WG21 N1254, and J16/00-0045 = WG21 N1268.)

Proposed Resolution (04/01):

1. Insert the following as a new paragraph following 11  class.access paragraph 1:

   A member of a class can also access all names as the class of which it is a member. A local class of a member function may access the same names that the member function itself may access. [Footnote: Access permissions are thus transitive and cumulative to nested and local classes.]

2. Delete 11  class.access paragraph 6.

3. In 11.8  class.access.nest paragraph 1, change

   The members of a nested class have no special access to members of an enclosing class, nor to classes or functions that have granted friendship to an enclosing class; the usual access rules (clause 11  class.access) shall be obeyed.

   to

   A nested class is a member and as such has the same access rights as any other member.

   Change

       B b;       // error: E::B is private
   

   to

       B b;       // Okay, E::I can access E::B
   

   Change

       p->x = i;      // error: E::x is private
   

   to

       p->x = i;      // Okay, E::I can access E::x
   

4. Delete 11.8  class.access.nest paragraph 2.

(This resolution also resolves issues 8 and 10.

263. Can a constructor be declared a friend?

[Voted into WP at April 2003 meeting.]

According to 12.1  class.ctor paragraph 1, a declaration of a constructor has a special limited syntax, in which only function-specifiers are allowed. A friend specifier is not a function-specifier, so one interpretation is that a constructor cannot be declared in a friend declaration.

(It should also be noted, however, that neither friend nor function-specifier is part of the declarator syntax, so it's not clear that anything conclusive can be derived from the wording of 12.1  class.ctor.)

Notes from 04/01 meeting:

The consensus of the core language working group was that it should be permitted to declare constructors as friends.

Proposed Resolution (revised October 2002):

Change paragraph 1a in 3.4.3.1  class.qual (added by the resolution of issue 147) as follows:

If the nested-name-specifier nominates a class C, and the name specified after the nested-name-specifier, when looked up in C, is the injected-class-name of C (clause 9  class), the name is instead considered to name the constructor of class C. Such a constructor name shall be used only in the declarator-id of a constructor definition declaration that appears outside of the class definition names a constructor....

Note: the above does not allow qualified names to be used for in-class declarations; see 8.3  dcl.meaning paragraph 1. Also note that issue 318 updates the same paragraph.

Change the example in 11.4  class.friend, paragraph 4 as follows:

class Y {
  friend char* X::foo(int);
  friend X::X(char);   // constructors can be friends
  friend X::~X();      // destructors can be friends
  //...
};

326. Wording for definition of trivial constructor

[Voted into WP at October 2003 meeting.]

In 12.1  class.ctor paragraph 5, the standard says "A constructor is trivial if [...]", and goes on to define a trivial default constructor. Taken literally, this would mean that a copy constructor can't be trivial (contrary to 12.8  class.copy paragraph 6). I suggest changing this to "A default constructor is trivial if [...]". (I think the change is purely editorial.)

Proposed Resolution (revised October 2002):

Change 12.1  class.ctor paragraph 5-6 as follows:

A default constructor for a class X is a constructor of class X that can be called without an argument. If there is no user-declared user-declared constructor for class X, a default constructor is implicitly declared. An implicitly-declared implicitly-declared default constructor is an inline public member of its class. A default constructor is trivial if it is an implicitly-declared default constructor and if:

• its class has no virtual functions (10.3  class.virtual) and no virtual base classes (10.1  class.mi), and
• all the direct base classes of its class have trivial default constructors, and
• for all the nonstatic data members of its class that are of class type (or array thereof), each such class has a trivial default constructor.

Otherwise, the default constructor is non-trivial.

Change 12.4  class.dtor paragraphs 3-4 as follows (the main changes are removing italics):

If a class has no user-declared user-declared destructor, a destructor is declared implicitly. An implicitly-declared implicitly-declared destructor is an inline public member of its class. A destructor is trivial if it is an implicitly-declared destructor and if:

• all of the direct base classes of its class have trivial destructors and
• for all of the non-static data members of its class that are of class type (or array thereof), each such class has a trivial destructor.

Otherwise, the destructor is non-trivial non-trivial.

In 9.5  class.union paragraph 1, change "trivial constructor" to "trivial default constructor".

In 12.2  class.temporary paragraph 3, add to the reference to 12.1  class.ctor a second reference, to 12.8  class.copy.

331. Allowed copy constructor signatures

[Voted into WP at October 2003 meeting.]

12.1  class.ctor paragraph 10 states

A copy constructor for a class X is a constructor with a first parameter of type X & or of type const X &. [Note: see 12.8  class.copy for more information on copy constructors.]

No mention is made of constructors with first parameters of types volatile X & or const volatile X &. This statement seems to be in contradiction with 12.8  class.copy paragraph 2 which states

A non-template constructor for class X is a copy constructor if its first parameter is of type X &, const X &, volatile X & or const volatile X &, ...

12.8  class.copy paragraph 5 also mentions the volatile versions of the copy constructor, and the comparable paragraphs for copy assignment (12.8  class.copy paragraphs 9 and 10) all allow volatile versions, so it seems that 12.1  class.ctor is at fault.

Proposed resolution (October 2002):

Change 12.1  class.ctor paragraph 10 from

A copy constructor for a class X is a constructor with a first parameter of type X& or of type const X&. [Note: see 12.8  class.copy for more information on copy constructors. ]

A copy constructor (12.8  class.copy) is used to copy objects of class type.

124. Lifetime of temporaries in default initialization of class arrays

[Moved to DR at 4/01 meeting.]

Jack Rouse: 12.2  class.temporary states that temporary objects will normally be destroyed at the end of the full expression in which they are created. This can create some unique code generation requirements when initializing a class array with a default constructor that uses a default argument. Consider the code:

    struct T {
       int i;
       T( int );
       ~T();
    };

    struct S {
       S( int = T(0).i );
       ~S();
    };

    S* f( int n )
    {
       return new S[n];
    }

I believe that many existing implementations will destroy the temporaries needed by the default constructor after each array element is initialized. But I can't find anything in the standard that allows the temporaries to be destroyed early in this case.

I think the standard should allow the early destruction of temporaries used in the default initialization of class array elements. I believe early destruction is the status quo, and I don't think the users of existing C++ compilers have been adversely impacted by it.

Proposed resolution (04/01):

The proposed resolution is contained in the proposal for issue 201.

201. Order of destruction of temporaries in initializers

[Moved to DR at 4/01 meeting.]

According to 12.2  class.temporary paragraph 4, an expression appearing as the initializer in an object definition constitutes a context "in which temporaries are destroyed at a different point than the end of the full-expression." It goes on to say that the temporary containing the value of the expression persists until after the initialization is complete (see also issue 117). This seems to presume that the end of the full-expression is a point earlier than the completion of the initialization.

However, according to 1.9  intro.execution paragraphs 12-13, the full-expression in such cases is, in fact, the entire initialization. If this is the case, the behavior described for temporaries in an initializer expression is simply the normal behavior of temporaries in any expression, and treating it as an exception to the general rule is both incorrect and confusing.

Proposed resolution (04/01):

[Note: this proposal also addresses issue 124.]

1. Add to the end of 1.9  intro.execution paragraph 12:

   If the initializer for an object or sub-object is a full-expression, the initialization of the object or sub-object (e.g., by calling a constructor or copying an expression value) is considered to be part of the full-expression.

2. Replace 12.2  class.temporary paragraph 4 with:

   There are two contexts in which temporaries are destroyed at a different point than the end of the full-expression. The first context is when a default constructor is called to initialize an element of an array. If the constructor has one or more default arguments, any temporaries created in the default argument expressions are destroyed immediately after return from the constructor.

320. Question on copy constructor elision example

[Voted into WP at April 2005 meeting.]

Section 12.2  class.temporary paragraph 2, abridged:

  X f(X);
  void g()
  {
	X a;
	a = f(a);
  }

a=f(a) requires a temporary for either the argument a or the result of f(a) to avoid undesired aliasing of a.

The note seems to imply that an implementation is allowed to omit copying "a" to f's formal argument, or to omit using a temporary for the return value of f. I don't find that license in normative text.

Function f returns an X by value, and in the expression the value is assigned (not copy-constructed) to "a". I don't see how that temporary can be omitted. (See also 12.8  class.copy p 15)

Since "a" is an lvalue and not a temporary, I don't see how copying "a" to f's formal parameter can be avoided.

Am I missing something, or is 12.2  class.temporary p 2 misleading?

Proposed resolution (October, 2004):

In 12.2  class.temporary paragraph 2, change the last sentence as indicated:

On the other hand, the expression a=f(a) requires a temporary for either the argument a or the result of f(a) to avoid undesired aliasing of a the result of f(a), which is then assigned to a.

392. Use of full expression lvalue before temporary destruction

[Voted into WP at March 2004 meeting.]

class C {
public:
    C();
    ~C();
    int& get() { return p; } // reference return
private:
    int p;
};

int main ()
{
    if ( C().get() ) // OK?
}

Section 12.2  class.temporary paragraph 3 says a temp is destroyed as the last step in evaluating the full expression. But the expression C().get() has a reference type. Does 12.2  class.temporary paragraph 3 require that the dereference to get a boolean result occur before the destructor runs, making the code valid? Or does the code have undefined behavior?

Bill Gibbons: It has undefined behavior, though clearly this wasn't intended. The lvalue-to-rvalue conversion that occurs in the "if" statement is not currently part of the full-expression.

From section 12.2  class.temporary paragraph 3:

Temporary objects are destroyed as the last step in evaluating the full-expression (1.9  intro.execution) that (lexically) contains the point where they were created.

From section 1.9  intro.execution paragraph 12:

A full-expression is an expression that is not a subexpression of another expression. 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.

The note in section 1.9  intro.execution paragraph 12 goes on to explain that this covers expressions used as initializers, but it does not discuss lvalues within temporaries.

It is a small point but it is probably worth correcting 1.9  intro.execution paragraph 12. Instead of the "implicit call of a function" wording, it might be better to just say that a full-expression includes any implicit use of the expression value in the enclosing language construct, and include a note giving implicit calls and lvalue-to-rvalue conversions as examples.

Offhand the places where this matters include: initialization (including member initializers), selection statements, iteration statements, return, throw

Proposed resolution (April 2003):

Change 1.9  intro.execution paragraph 12-13 to read:

A full-expression is an expression that is not a subexpression of another expression. 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. 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.

[Note: certain contexts in C++ cause the evaluation of a full-expression that results from a syntactic construct other than expression (5.18  expr.comma). For example, in 8.5  dcl.init one syntax for initializer is

( expression-list )

but the resulting construct is a function call upon a constructor function with expression-list as an argument list; such a function call is a full-expression. For example, in 8.5  dcl.init, another syntax for initializer is

= initializer-clause

but again the resulting construct might be a function call upon a constructor function with one assignment-expression as an argument; again, the function call is a 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]

443. Wording nit in description of lifetime of temporaries

[Voted into WP at April 2005 meeting.]

There seems to be a typo in 12.2  class.temporary/5, which says "The temporary to which the reference is bound or the temporary that is the complete object TO a subobject OF which the TEMPORARY is bound persists for the lifetime of the reference except as specified below."

I think this should be "The temporary to which the reference is bound or the temporary that is the complete object OF a subobject TO which the REFERENCE is bound persists for the lifetime of the reference except as specified below."

I used upper-case letters for the parts I think need to be changed.

Proposed resolution (October, 2004):

Change 12.2  class.temporary paragraph 5 as indicated:

The temporary to which the reference is bound or the temporary that is the complete object to of a subobject of to which the temporary reference is bound persists for the lifetime of the reference except as specified below.

296. Can conversion functions be static?

[Moved to DR at October 2002 meeting.]

May user-defined conversion functions be static? That is, should this compile?

    class Widget {
    public:
      static operator bool() { return true; }
    };

All my compilers hate it. I hate it, too. However, I don't see anything in 12.3.2  class.conv.fct that makes it illegal. Is this a prohibition that arises from the grammar, i.e., the grammar doesn't allow "static" to be followed by a conversion-function-id in a member function declaration? Or am I just overlooking something obvious that forbids static conversion functions?

Proposed Resolution (4/02):

Add to 12.3.2  class.conv.fct as a new paragraph 7:

Conversion functions cannot be declared static.

244. Destructor lookup

[Moved to DR at October 2002 meeting.]

12.4  class.dtor contains this example:

    struct B {
        virtual ~B() { }
    };
    struct D : B {
        ~D() { }
    };

    D D_object;
    typedef B B_alias;
    B* B_ptr = &D_object;

    void f() {
        D_object.B::~B();               // calls B's destructor
        B_ptr->~B();                    // calls D's destructor
        B_ptr->~B_alias();              // calls D's destructor
        B_ptr->B_alias::~B();           // calls B's destructor
        B_ptr->B_alias::~B_alias();     // error, no B_alias in class B
    }

On the other hand, 3.4.3  basic.lookup.qual contains this example:

    struct C {
        typedef int I;
    };
    typedef int I1, I2;
    extern int* p;
    extern int* q;
    p->C::I::~I();       // I is looked up in the scope of C
    q->I1::~I2();        // I2 is looked up in the scope of
                         // the postfix-expression
    struct A {
        ~A();
    };
    typedef A AB;
    int main()
    {
        AB *p;
        p->AB::~AB();    // explicitly calls the destructor for A
    }

Note that

     B_ptr->B_alias::~B_alias();

is claimed to be an error, while the equivalent

     p->AB::~AB();

is claimed to be well-formed.

I believe that clause 3 is correct and that clause 12 is in error. We worked hard to get the destructor lookup rules in clause 3 to be right, and I think we failed to notice that a change was also needed in clause 12.

Mike Miller:

Unfortunately, I don't believe 3.4.3  basic.lookup.qual covers the case of p->AB::~AB(). It's clearly intended to do so, as evidenced by 3.4.3.1  class.qual paragraph 1 ("a destructor name is looked up as specified in 3.4.3  basic.lookup.qual"), but I don't think the language there does so.

The relevant paragraph is 3.4.3  basic.lookup.qual paragraph 5. (None of the other paragraphs in that section deal with this topic at all.) It has two parts. The first is

If a pseudo-destructor-name (5.2.4  expr.pseudo) contains a nested-name-specifier, the type-names are looked up as types in the scope designated by the nested-name-specifier.

This sentence doesn't apply, because ~AB isn't a pseudo-destructor-name. 5.2.4  expr.pseudo makes clear that this syntactic production (5.2  expr.post paragraph 1) only applies to cases where the type-name is not a class-name. p->AB::~AB is covered by the production using id-expression.

The second part of 3.4.3  basic.lookup.qual paragraph 5 says

In a qualified-id of the form:

::_opt nested-name-specifier ~ class-name

where the nested-name-specifier designates a namespace name, and in a qualified-id of the form:

::_opt nested-name-specifier class-name :: ~ class-name

the class-names are looked up as types in the scope designated by the nested-name-specifier.

This wording doesn't apply, either. The first one doesn't because the nested-name-specifier is a class-name, not a namespace name. The second doesn't because there's only one layer of qualification.

As far as I can tell, there's no normative text that specifies how the ~AB is looked up in p->AB::~AB(). 3.4.3.1  class.qual, where all the other class member qualified lookups are handled, defers to 3.4.3  basic.lookup.qual, and 3.4.3  basic.lookup.qual doesn't cover the case.

See also issue 305.

Jason Merrill: My thoughts on the subject were that the name we use in a destructor call is really meaningless; as soon as we see the ~ we know what the user means, all we're doing from that point is testing their ability to name the destructor in a conformant way. I think that everyone will agree that

  anything::B::~B()

Anyone writing two different names here is either deliberately writing obfuscated code, trying to call the destructor of a nested class, or fighting an ornery compiler (i.e. one that still wants to see B_alias::~B()). I think we can ignore the first case. The third would be handled by reverting to the old rule (look up the name after ~ in the normal way) with the lexical matching exception described above -- or we could decide to break such code, do no lookup at all, and only accept a matching name. In a good implementation, the second should probably get an error message telling them to write Outer::Inner::~Inner instead.

We discussed this at the meetings, but I don't remember if we came to any sort of consensus on a direction. I see three options:

1. Stick with the status quo, i.e. the special lookup rule such that if the name before ::~ is a class name, the name after ::~ is looked up in the same scope as the previous one. If we choose this option, we just need better wording that actually expresses this, as suggested in the issue list. This option breaks old B_alias::~B code where B_alias is declared in a different scope from B.
2. Revert to the old rules, whereby the name after ::~ is looked up just like a name after ::, with the exception that if it matches the name before ::~ then it is considered to name the same class. This option supports old code and code that writes B_alias::~B_alias. It does not support the q->I1::~I2 usage of 3.4.3  basic.lookup.qual, but that seems like deliberate obfuscation. This option is simpler to implement than #1.
3. Do no lookup for a name after ::~; it must match the name before. This breaks old code as #1, but supports the most important case where the names match. This option may be slightly simpler to implement than #2. It is certainly easier to teach.

My order of preference is 2, 3, 1.

Incidentally, it seems to me oddly inconsistent to allow Namespace::~Class, but not Outer::~Inner. Prohibiting the latter makes sense from the standpoint of avoiding ambiguity, but what was the rationale for allowing the former?

John Spicer: I agree that allowing Namespace::~Class is odd. I'm not sure where this came from. If we eliminated that special case, then I believe the #1 rule would just be that in A::B1::~B2 you look up B1 and B2 in the same place in all cases.

I don't like #2. I don't think the "old" rules represent a deliberate design choice, just an error in the way the lookup was described. The usage that rule permits p->X::~Y (where Y is a typedef to X defined in X), but I doubt people really do that. In other words, I think that #1 a more useful special case than #2 does, not that I think either special case is very important.

One problem with the name matching rule is handling cases like:

  A<int> *aip;

  aip->A<int>::~A<int>();  // should work
  aip->A<int>::~A<char>(); // should not

Proposed resolution (10/01):

Replace the normative text of 3.4.3  basic.lookup.qual paragraph 5 after the first sentence with:

Similarly, in a qualified-id of the form:

    ::opt nested-name-specifieropt class-name :: ~ class-name

the second class-name is looked up in the same scope as the first.

In 12.4  class.dtor paragraph 12, change the example to

D D_object;
typedef B B_alias;
B* B_ptr = &D_object;

void f() {
  D_object.B::~B();                //  calls  B's destructor
  B_ptr->~B();                    //  calls  D's destructor
  B_ptr->~B_alias();              //  calls  D's destructor
  B_ptr->B_alias::~B();           //  calls  B's destructor
  B_ptr->B_alias::~B_alias();     //  calls  B's destructor
}

April 2003: See issue 399.

252. Looking up deallocation functions in virtual destructors

[Moved to DR at 10/01 meeting.]

There is a mismatch between 12.4  class.dtor paragraph 11 and 12.5  class.free paragraph 4 regarding the lookup of deallocation functions in virtual destructors. 12.4  class.dtor says,

At the point of definition of a virtual destructor (including an implicit definition (12.8  class.copy)), non-placement operator delete shall be looked up in the scope of the destructor's class (3.4.1  basic.lookup.unqual) and if found shall be accessible and unambiguous. [Note: this assures that an operator delete corresponding to the dynamic type of an object is available for the delete-expression (12.5  class.free). ]

The salient features to note from this description are:

1. The lookup is "in the scope of the destructor's class," which implies that only members are found (cf 12.2  class.temporary). (The cross-reference would indicate otherwise, however, since it refers to the description of looking up unqualified names; this kind of lookup "spills over" into the surrounding scope.)
2. Only non-placement operator delete is looked up. Presumably this means that a placement operator delete is ignored in the lookup.

On the other hand, 12.5  class.free says,

If a delete-expression begins with a unary :: operator, the deallocation function's name is looked up in global scope. Otherwise, if the delete-expression is used to deallocate a class object whose static type has a virtual destructor, the deallocation function is the one found by the lookup in the definition of the dynamic type's virtual destructor (12.4  class.dtor). Otherwise, if the delete-expression is used to deallocate an object of class T or array thereof, the static and dynamic types of the object shall be identical and the deallocation function's name is looked up in the scope of T. If this lookup fails to find the name, the name is looked up in the global scope. If the result of the lookup is ambiguous or inaccessible, or if the lookup selects a placement deallocation function, the program is ill-formed.

Points of interest in this description include:

1. For a class type with a virtual destructor, the lookup is described as being "in the definition of the dynamic type's virtual destructor," rather than "in the scope of the dynamic type." That is, the lookup is assumed to be an unqualified lookup, presumably terminating in the global scope.
2. The assumption is made that the lookup in the virtual destructor was successful ("...the one found...", not "...the one found..., if any"). This will not be the case if the deallocation function was not declared as a member somewhere in the inheritance hierarchy.
3. The lookup in the non-virtual-destructor case does find placement deallocation functions and can fail as a result.

Suggested resolution: Change the description of the lookup in 12.4  class.dtor paragraph 11 to match the one in 12.5  class.free paragraph 4.

Proposed resolution (10/00):

1. Replace 12.4  class.dtor paragraph 11 with the following:

   At the point of definition of a virtual destructor (including an implicit definition), the non-array deallocation function is looked up in the scope of the destructor's class (10.2  class.member.lookup), and, if no declaration is found, the function is looked up in the global scope. If the result of this lookup is ambiguous or inaccessible, or if the lookup selects a placement deallocation function, the program is ill-formed. [Note: this assures that a deallocation function corresponding to the dynamic type of an object is available for the delete-expression (12.5  class.free).]

2. In 12.5  class.free paragraph 4, change

   ...the deallocation function is the one found by the lookup in the definition of the dynamic type's virtual destructor (12.4  class.dtor).

   to

   ...the deallocation function is the one selected at the point of definition of the dynamic type's virtual destructor (12.4  class.dtor).

272. Explicit destructor invocation and qualified-ids

[Moved to DR at 10/01 meeting.]

an explicit destructor call must always be written using a member access operator (5.2.5  expr.ref); in particular, the unary-expression ~X() in a member function is not an explicit destructor call (5.3.1  expr.unary.op).

This note is incorrect, as an explicit destructor call can be written as a qualified-id, e.g., X::~X(), which does not use a member access operator.

Proposed resolution (04/01):

Change 12.4  class.dtor paragraph 12 as follows:

[Note: an explicit destructor call must always be written using a member access operator (5.2.5  expr.ref) or a qualified-id (5.1  expr.prim); in particular, the unary-expression ~X() in a member function is not an explicit destructor call (5.3.1  expr.unary.op).]

162. (&C::f)() with nonstatic members

[Moved to DR at October 2002 meeting.]

13.3.1.1  over.match.call paragraph 3 says that when a call of the form

   (&C::f)()

This is clear, it's implementable, and it doesn't directly contradict anything else in the standard. However, I'm not sure it's consistent with some similar cases.

In 13.4  over.over paragraph 5, second example, it is made amply clear that when &C::f is used as the address of a function, e.g.,

   int (*pf)(int) = &C::f;

Similarly, 13.3.1.1.1  over.call.func paragraph 3 makes it clear that

   C::f();

Those paragraphs seem to suggest the general rule that you do overload resolution first and then you interpret the construct you have according to the function selected. The fact that there are static and nonstatic functions in the overload set is irrelevant; it's only necessary that the chosen function be static or nonstatic to match the context.

Given that, I think it would be more consistent if the (&C::f)() case would also do overload resolution first. If a nonstatic member is chosen, the program would be ill-formed.

Proposed resolution (04/01):

1. Change the indicated text in 13.3.1.1  over.match.call paragraph 3:

   The fourth case arises from a postfix-expression of the form &F, where F names a set of overloaded functions. In the context of a function call, the set of functions named by F shall contain only non-member functions and static member functions. [Footnote: If F names a non-static member function, &F is a pointer-to-member, which cannot be used with the function call syntax.] And in this context using &F behaves the same as using &F is treated the same as the name F by itself. Thus, (&F)(expression-list_opt) is simply (F)(expression-list_opt), which is discussed in 13.3.1.1.1  over.call.func. If the function selected by overload resolution according to 13.3.1.1.1  over.call.func is a nonstatic member function, the program is ill-formed. [Footnote: When F is a nonstatic member function, a reference of the form &A::F is a pointer-to-member, which cannot be used with the function-call syntax, and a reference of the form &F is an invalid use of the "&" operator on a nonstatic member function.] (The resolution of &F in other contexts is described in 13.4  over.over.)

239. Footnote 116 and Koenig lookup

[Moved to DR at 4/01 meeting.]

Because of the usual name hiding rules, these will be introduced by declarations or by using-directives all found in the same block or all found at namespace scope.

At least in terms of the current state of the Standard, this is not correct: a block extern declaration does not prevent Koenig lookup from occurring. For example,

    enum E { zero };
    void f(E);
    void g() {
        void f(int);
        f(zero);
    }

In this example, the overload set will include declarations from both namespace and block scope.

(See also issue 12.)

Proposed resolution (04/01):

1. In 3.4.2  basic.lookup.argdep paragraph 2, change

   If the ordinary unqualified lookup of the name finds the declaration of a class member function, the associated namespaces and classes are not considered.

   to

   If the ordinary unqualified lookup of the name finds the declaration of a class member function, or a block-scope function declaration that is not a using-declaration, the associated namespaces and classes are not considered.

   and change the example to:

       namespace NS {
           class T { };
           void f(T);
           void g(T, int);
       }
       NS::T parm;
       void g(NS::T, float);
       int main() {
           f(parm);            // OK: calls NS::f
           extern void g(NS::T, float);
           g(parm, 1);         // OK: calls g(NS::T, float)
       }
   

2. In 13.3.1.1.1  over.call.func paragraph 3 from:

   If the name resolves to a non-member function declaration, that function and its overloaded declarations constitute the set of candidate functions.

   to

   If the name resolves to a set of non-member function declarations, that set of functions constitutes the set of candidate functions.

   Note that this text is also edited by issue 364. Also, remove the associated footnote 116.

364. Calling overloaded function with static in set, with no object

[Voted into WP at October 2003 meeting.]

Consider this program:

   struct S {
     static void f (int);
     void f (char);
   };

   void g () {
     S::f ('a');
   }

G++ 3.1 rejects it, saying:

   test.C:7: cannot call member function `void S::f(char)' without object

Mark Mitchell: It looks to me like G++ is correct, given 13.3.1.1.1  over.call.func. This case is the "unqualified function call" case described in paragraph 3 of that section. ("Unqualified" here means that there is no "x->" or "x." in front of the call, not that the name is unqualified.)

That paragraph says that you first do name lookup. It then asks you to look at what declaration is returned. (That's a bit confusing; you presumably get a set of declarations. Or maybe not; the name lookup section says that if name lookup finds a non-static member in a context like this the program is in error. But surely this program is not erroneous. Hmm.)

Anyhow, you have -- at least -- "S::f(char)" as the result of the lookup.

The keyword "this" is not in scope, so "all overloaded declarations of the function name in T become candidate functions and a contrived object of type T becomes the implied object argument." That means we get both versions of "f" at this point. Then, "the call is ill-formed, however, if overload resolution selects one of the non-static members of T in this case." Since, in this case, "S::f(char)" is the winner, the program is ill-formed.

Steve Adamczyk: This result is surprising, because we've selected a function that we cannot call, when there is another function that can be called. This should either be ambiguous, or it should select the static member function. See also 13.3.1  over.match.funcs paragraph 2: "Similarly, when appropriate, the context can construct an argument list that contains an implied object argument..."

Notes from October 2002 meeting:

We agreed that g++ has it right, but the standard needs to be clearer.

Proposed resolution (October 2002, revised April 2003):

Change 13.3.1.1.1  over.call.func paragraphs 2 and 3 as follows:

In qualified function calls, the name to be resolved is an id-expression and is preceded by an -> or . operator. Since the construct A->B is generally equivalent to (*A).B, the rest of clause 13  over assumes, without loss of generality, that all member function calls have been normalized to the form that uses an object and the . operator. Furthermore, clause 13  over assumes that the postfix-expression that is the left operand of the . operator has type ``cv T'' where T denotes a class. [Footnote: Note that cv-qualifiers on the type of objects are significant in overload resolution for both lvalue and class rvalue objects. --- end footnote] Under this assumption, the id-expression in the call is looked up as a member function of T following the rules for looking up names in classes (10.2  class.member.lookup). If a member function is found, that function and its overloaded declarations The function declarations found by that lookup constitute the set of candidate functions. The argument list is the expression-list in the call augmented by the addition of the left operand of the . operator in the normalized member function call as the implied object argument (13.3.1  over.match.funcs).

In unqualified function calls, the name is not qualified by an -> or . operator and has the more general form of a primary-expression. The name is looked up in the context of the function call following the normal rules for name lookup in function calls (3.4.2  basic.lookup.argdep 3.4  basic.lookup). If the name resolves to a non-member function declaration, that function and its overloaded declarations The function declarations found by that lookup constitute the set of candidate functions. [Footnote: Because of the usual name hiding rules, these will be introduced by declarations or by using-directives all found in the same block or all found at namespace scope. --- end footnote] Because of the rules for name lookup, the set of candidate functions consists (1) entirely of non-member functions or (2) entirely of member functions of some class T. In case (1), tThe argument list is the same as the expression-list in the call. If the name resolves to a nonstatic member function, then the function call is actually a member function call. In case (2), the argument list is the expression-list in the call augmented by the addition of an implied object argument as in a qualified function call. If the keyword this (9.3.2  class.this) is in scope and refers to the class T of that member function, or a derived class thereof of T, then the function call is transformed into a normalized qualified function call using implied object argument is(*this) as the postfix-expression to the left of the . operator. The candidate functions and argument list are as described for qualified function calls above. If the keyword this is not in scope or refers to another class, then name resolution found a static member of some class T. In this case, all overloaded declarations of the function name in T become candidate functions and a contrived object of type T becomes the implied object argument. [Footnote: An implied object argument must be contrived to correspond to the implicit object parameter attributed to member functions during overload resolution. It is not used in the call to the selected function. Since the member functions all have the same implicit object parameter, the contrived object will not be the cause to select or reject a function. --- end footnote] If the argument list is augmented by a contrived object and The call is ill-formed, however, if overload resolution selects one of the non-static member functions of T, the call is ill-formed in this case.

Note that issue 239 also edits paragraph 3.

280. Access and surrogate call functions

[Voted into WP at October 2003 meeting.]

According to 13.3.1.1.2  over.call.object paragraph 2, when the primary-expression E in the function call syntax evaluates to a class object of type "cv T", a surrogate call function corresponding to an appropriate conversion function declared in a direct or indirect base class B of T is included or not included in the set of candidate functions based on class B being accessible.

For instance in the following code sample, as per the paragraph in question, the expression c(3) calls f2, instead of the construct being ill-formed due to the conversion function A::operator fp1 being inaccessible and its corresponding surrogate call function providing a better match than the surrogate call function corresponding to C::operator fp2:

    void  f1(int)  {   }
    void  f2(float)  {   }
    typedef void  (* fp1)(int);
    typedef void  (* fp2)(float);

    struct  A  {
       operator fp1()
       {   return  f1;   }
    };

    struct  B :  private A  {   };

    struct  C :  B  {
       operator  fp2()
       {   return  f2;   }
    };

    int  main()
    {
       C   c;
       c(3);  // f2 is called, instead of the construct being ill-formed.
       return  0;
    }

The fact that the accessibility of a base class influences the overload resolution process contradicts the fundamental language rule (3.4  basic.lookup paragraph 1, and 13.3  over.match paragraph 2) that access checks are applied only once name lookup and function overload resolution (if applicable) have succeeded.

Notes from 4/02 meeting:

There was some concern about whether 10.2  class.member.lookup (or anything else, for that matter) actually defines "ambiguous base class". See issue 39. See also issue 156.

Notes from October 2002 meeting:

It was suggested that the ambiguity check is done as part of the call of the conversion function.

Proposed resolution (revised October 2002):

In 13.3.1.1.2  over.call.object paragraph 2, replace the last sentence

Similarly, surrogate call functions are added to the set of candidate functions for each conversion function declared in an accessible base class provided the function is not hidden within T by another intervening declaration.

with

Similarly, surrogate call functions are added to the set of candidate functions for each conversion function declared in a base class of T provided the function is not hidden within T by another intervening declaration.

Replace 13.3.1.1.2  over.call.object paragraph 3

If such a surrogate call function is selected by overload resolution, its body, as defined above, will be executed to convert E to the appropriate function and then to invoke that function with the arguments of the call.

If such a surrogate call function is selected by overload resolution, the corresponding conversion function will be called to convert E to the appropriate function pointer or reference, and the function will then be invoked with the arguments of the call. If the conversion function cannot be called (e.g., because of an ambiguity), the program is ill-formed.

416. Class must be complete to allow operator lookup?

[Voted into WP at October 2004 meeting.]

Normally reference semantics allow incomplete types in certain contexts, but isn't this:

  class A;

  A& operator<<(A& a, const char* msg);
  void foo(A& a)
  {
    a << "Hello";
  }

required to be diagnosed because of the op<<? The reason being that the class may actually have an op<<(const char *) in it.

What is it? un- or ill-something? Diagnosable? No problem at all?

Steve Adamczyk: I don't know of any requirement in the standard that the class be complete. There is a rule that will instantiate a class template in order to be able to see whether it has any operators. But I wouldn't think one wants to outlaw the above example merely because the user might have an operator<< in the class; if he doesn't, he would not be pleased that the above is considered invalid.

Mike Miller: Hmm, interesting question. My initial reaction is that it just uses ::operator<<; any A::operator<< simply won't be considered in overload resolution. I can't find anything in the Standard that would say any different.

The closest analogy to this situation, I'd guess, would be deleting a pointer to an incomplete class; 5.3.5  expr.delete paragraph 5 says that that's undefined behavior if the complete type has a non-trivial destructor or an operator delete. However, I tend to think that that's because it deals with storage and resource management, not just because it might have called a different function. Generally, overload resolution that goes one way when it might have gone another with more declarations in scope is considered to be not an error, cf 7.3.3  namespace.udecl paragraph 9, 14.6.3  temp.nondep paragraph 1, etc.

So my bottom line take on it would be that it's okay, it's up to the programmer to ensure that all necessary declarations are in scope for overload resolution. Worst case, it would be like the operator delete in an incomplete class -- undefined behavior, and thus not required to be diagnosed.

13.3.1.2  over.match.oper paragraph 3, bullet 1, says, "If T1 is a class type, the set of member candidates is the result of the qualified lookup of T1::operator@ (13.3.1.1.1  over.call.func)." Obviously, that lookup is not possible if T1 is incomplete. Should 13.3.1.2  over.match.oper paragraph 3, bullet 1, say "complete class type"? Or does the inability to perform the lookup mean that the program is ill-formed? 3.2  basic.def.odr paragraph 4 doesn't apply, I don't think, because you don't know whether you'll be applying a class member access operator until you know whether the operator involved is a member or not.

Notes from October 2003 meeting:

We noticed that the title of this issue did not match the body. We checked the original source and then corrected the title (so it no longer mentions templates).

We decided that this is similar to other cases like deleting a pointer to an incomplete class, and it should not be necessary to have a complete class. There is no undefined behavior.

Proposed Resolution (October 2003):

Change the first bullet of 13.3.1.2  over.match.oper paragraph 3 to read:

If T1 is a complete class type, the set of member candidates is the result of the qualified lookup of T1::operator@ (13.3.1.1.1  over.call.func); otherwise, the set of member candidates is empty.

60. Reference binding and valid conversion sequences

[Moved to DR at October 2002 meeting.]

Does dropping a cv-qualifier on a reference binding prevent the binding as far as overload resolution is concerned? Paragraph 4 says "Other restrictions on binding a reference to a particular argument do not affect the formation of a conversion sequence." This was intended to refer to things like access checking, but some readers have taken that to mean that any aspects of reference binding not mentioned in this section do not preclude the binding.

Proposed resolution (10/01):

In 13.3.3.1.4  over.ics.ref paragraph 4 add the indicated text:

Other restrictions on binding a reference to a particular argument that are not based on the types of the reference and the argument do not affect the formation of a standard conversion sequence, however.

115. Address of template-id

[Voted into WP at October 2003 meeting.]

    template <class T> void f(T);
    template <class T> void g(T);
    template <class T> void g(T,T);

    int main()
    {
        (&f<int>);
        (&g<int>);
    }

13.4  over.over paragraph 1 says that a function template name is considered to name a set of overloaded functions. I believe it should be expanded to say that a function template name with an explicit template argument list is also considered to name a set of overloaded functions.

In the general case, you need to have a destination type in order to identify a unique function. While it is possible to permit this, I don't think it is a good idea because such code depends on there only being one template of that name that is visible.

The EDG front end issues an error on this use of "f". egcs 1.1.1 allows it, but the most current snapshot of egcs that I have also issues an error on it.

It has been pointed out that when dealing with nontemplates, the rules for taking the address of a single function differ from the rules for an overload set, but this asymmetry is needed for C compatibility. This need does not exist for the template case.

My feeling is that a general rule is better than a general rule plus an exception. The general rule is that you need a destination type to be sure that the operation will succeed. The exception is when there is only one template in the set and only then when you provide values for all of the template arguments.

It is true that in some cases you can provide a shorthand, but only if you encourage a fragile coding style (that will cause programs to break when additional templates are added).

I think the standard needs to specify one way or the other how this case should be handled. My recommendation would be that it is ill-formed.

Nico Josuttis: Consider the following example:

    template <int VAL>
    int add (int elem)
    {
	return elem + VAL;
    }

    std::transform(coll.begin(), coll.end(),
		   coll.begin(),
		   add<10>);

If John's recommendation is adopted, this code will become ill-formed. I bet there will be a lot of explanation for users necessary why this fails and that they have to change add<10> to something like (int (*)(int))add<10>.

This example code is probably common practice because this use of the STL is typical and is accepted in many current implementations. I strongly urge that this issue be resolved in favor of keeping this code valid.

Bill Gibbons: I find this rather surprising. Shouldn't a template-id which specifies all of the template arguments be treated like a declaration-only explicit instantiation, producing a set of ordinary function declarations? And when that set happens to contain only one function, shouldn't the example code work?

(See also issue 250.)

Notes from 04/01 meeting:

The consensus of the group was that the add example should not be an error.

Proposed resolution (October 2002):

In 13.4 add to the end of paragraph 2:

[Note: As described in 14.8.1  temp.arg.explicit, if deduction fails and the function template name is followed by an explicit template argument list, the template-id is then examined to see whether it identifies a single function template specialization. If it does, the template-id is considered to be an lvalue for that function template specialization. The target type is not used in that determination.]

In 14.8.1  temp.arg.explicit paragraph 2 insert before the first example:

In contexts where deduction is done and fails, or in contexts where deduction is not done, if a template argument list is specified and it, along with any default template arguments, identifies a single function template specialization, then the template-id is an lvalue for the function template specialization.

Change the first example of 14.8.1  temp.arg.explicit paragraph 2:

  template<class X, class Y> X f(Y);
  void g()
  {
    int i = f<int>(5.6);    // Y is deduced to be double
    int j = f(5.6);         // ill-formed: X cannot be deduced
  }

  template<class X, class Y> X f(Y);
  void g()
  {
    int i = f<int>(5.6);    // Y is deduced to be double
    int j = f(5.6);         // ill-formed: X cannot be deduced
    f<void>(f<int, bool>);  // Y for outer f deduced to be
                            //   int (*)(bool)
    f<void>(f<int>);        // ill-formed: f<int> does not denote a
                            //   single template function specialization
  }

Note: This interacts with the resolution of issue 226 (default template arguments for function templates).

221. Must compound assignment operators be member functions?

[Moved to DR at 4/01 meeting.]

Is the intent of 13.5.3  over.ass paragraph 1 that all assignment operators be non-static member functions (including operator+=, operator*=, etc.) or only simple assignment operators (operator=)?

Notes from 04/00 meeting:

Nearly all references to "assignment operator" in the IS mean operator= and not the compound assignment operators. The ARM was specific that this restriction applied only to operator=. If it did apply to compound assignment operators, it would be impossible to overload these operators for bool operands.

Proposed resolution (04/01):

1. Change the title of 5.17  expr.ass from "Assignment operators" to "Assignment and compound assignment operators."

2. Change the first sentence of 5.17  expr.ass paragraph 1 from

   There are several assignment operators, all of which group right-to-left. All require a modifiable lvalue as their left operand, and the type of an assignment expression is that of its left operand. The result of the assignment operation is the value stored in the left operand after the assignment has taken place; the result is an lvalue.

   to

   The assignment operator (=) and the compound assignment operators all group right-to-left. All require a modifiable lvalue as their left operand and return an lvalue with the type and value of the left operand after the assignment has taken place.

Additional note (10/00): Paragraphs 2-6 of 5.17  expr.ass should all be understood to apply to simple assignment only and not to compound assignment operators.

425. Set of candidates for overloaded built-in operator with float operand

[Voted into WP at March 2004 meeting.]

During a discussion over at the boost mailing list (www.boost.org), we came across the following "puzzle":

  struct A {
    template< typename T > operator T() const;
  } a;

  template<> A::operator float() const
  {
    return 1.0f;
  }

  int main()
  {
    float f = 1.0f * a;
  }

The code is compiled without errors or warnings from EDG-based compilers (Comeau, Intel), but rejected from others (GCC, MSVC [7.1]). The question: Who is correct? Where should I file the bug report?

To explain the problem: The EDG seems to see 1.0f*a as a call to the unambiguous operator*(float,float) and thus casts 'a' to 'float'. The other compilers have several operators (float*float, float*double, float*int, ...) available and thus can't decide which cast is appropriate. I think the latter is the correct behaviour, but I'd like to hear some comments from the language lawyers about the standard's point of view on this problem.

Andreas Hommel: Our compiler also rejects this code:

Error   : function call 'operator*(float, {lval} A)' is ambiguous
'operator*(float, unsigned long long)'
'operator*(float, int)'
'operator*(float, unsigned int)'
'operator*(float, long)'
'operator*(float, unsigned long)'
'operator*(float, float)'
'operator*(float, double)'
'operator*(float, long double)'
'operator*(float, long long)'
Test.cp line 12       float f = 1.0f * a;

Is this example really legal? It was my understanding that all candidates from 13.6  over.built participate in overload resolution.

Daveed Vandevoorde: I believe the EDG-based compiler is right. Note that the built-in operator* requires "usual arithmetic conversions" (see 5.6  expr.mul paragraph 2 and 5  expr paragaph 9). This means that there is no candidate taking (float, double) arguments: Only (float, float) or (double, double).

Since your first argument is of type float, the (float, float) case is preferred over the (double, double) case (the latter would require a floating-point promotion).

Stave Adamczyk: Daveed's statement is wrong; as Andreas says, the prototypes in 13.6  over.built paragraph 12 have pairs of types, not the same type twice. However, the list of possibilities considered in Andreas' message is wrong also: 13.6  over.built paragraph 12 calls for pairs of promoted arithmetic types, and float is not a promoted type (it promotes to double -- see 4.6  conv.fpprom).

Nevertheless, the example is ambiguous. Let's look at the overload resolution costs. The right operand is always going to have a user-defined-conversion cost (the template conversion function will convert directly to the const version of the second parameter of the prototype). The left operand is always going to have a promotion (float --> double) or a standard conversion (anything else). So the cases with promotions are better than the others. However, there are several of those cases, with second parameters of type int, unsigned int, long, unsigned long, double, and long double, and all of those are equally good. Therefore the example is ambiguous.

Here's a reduced version that should be equivalent:

  struct A {
    template <typename T> operator T() const;
  } a;
  void f(double, int);
  void f(double, unsigned int);
  void f(double, long);
  void f(double, unsigned long);
  void f(double, double);
  void f(double, long double);
  int main() {
    f(1.0f, a);  // Ambiguous
  }

Personally, I think this is evidence that 13.6  over.built doesn't really do quite what it should. But the standard is clear, if possibly flawed.

Andreas Hommel: You are right, "float" is not a promoted arithmetic type, this is a bug in our compiler.

However, the usual arithmetic conversions (5  expr paragraph 9) do not promote the floating point types, so

  float operator+(float, float);

Steve Adamczyk: Hmm, the definition of the term in 13.6  over.built paragraph 2 is highly ambiguous:

Similarly, the term promoted arithmetic type refers to promoted integral types plus floating types.

Proposed Resolution (October 2003):

Change the second sentence of 13.6 paragraph 2 as follows:

Similarly, the term promoted arithmetic type refers to promoted integral types plus floating types floating types plus promoted integral types.

204. Exported class templates

[Voted into WP at April 2003 meeting.]

14  temp paragraph 7 allows class templates to be declared exported, including member classes and member class templates (implicitly by virtue of exporting the containing template class). However, paragraph 8 does not exclude exported class templates from the statement that

An exported template need only be declared (and not necessarily defined) in a translation unit in which it is instantiated.

If an implicit instantiation of a class template specialization is required and the template is declared but not defined, the program is ill-formed.

A definition of a class template or a class member template shall be in scope at the point of the explicit instantiation of the class template or class member template.

Suggested resolution:

• Change 14  temp paragraph 8 to say that "An exported non-class template need only be declared..."
• Change 14.7.1  temp.inst paragraph 6 to use wording similar to that of 14.7.2  temp.explicit paragraph 3 regarding the requirement for a definition of the class template to be in scope.

(See also issue 212.)

Notes from 04/00 meeting:

John Spicer opined that even though 14  temp paragraph 7 speaks of "declaring a class template exported," that does not mean that the class template is "an exported template" in the sense of paragraph 8. He suggested clarifying paragraph 7 to that effect instead of the change to paragraph 8 suggested above, and questioned the need for a change to 14.7.1  temp.inst.

Notes from the 4/02 meeting:

This is resolved by the proposed changes for issue 323.

323. Where must export appear?

[Voted into WP at April 2003 meeting.]

The standard doesn't seem to describe whether the keyword export should appear on exported template declarations that are not used or defined in that particular translation unit.

For example:

  // File 1:
  template<typename T> void f();  // export omitted

  // File 2:
  export template<typename T> void f() {}

  int main() { f<int>(); }

Another example is:

  // File 1:
  struct S {
     template<typename T> void m();
  };

  // File 2:
  struct S {
     template<typename T> void m();
  };

  export template<typename T> void S::m() {}

  int main() {
     S s;
     S.m<int>();
  }

I think both examples should be clarified to be invalid. If a template is exported in one translation unit, it should be declared export in all translation units in which it appears.

With the current wording, it seems that even the following is valid:

  // File 1:
  export template<typename T> void f();  // export effectively ignored

  // File 2:
  template<typename T> void f() {}  // Inclusion model
  void g() { f<int>(); }

  // File 3:
  void g();
  template<typename T> void f() {}  // Inclusion model

  int main() {
     g();
     f<int>();
  }

In fact, I think the declaration in "File 1" could be a definition and this would still satisfy the the requirements of the standard, which definitely seems wrong.

Proposed Resolution (revised October 2002):

Replace 14  temp paragraphs 6, 7, and 8 by the following text:

A template-declaration may be preceded by the export keyword. Such a template is said to be exported. Declaring exported a class template is equivalent to declaring exported all of its non-inline member functions, static data members, member classes, member class templates, and non-inline member function templates.

If a template is exported in one translation unit, it shall be exported in all translation units in which it appears; no diagnostic is required. A declaration of an exported template shall appear with the export keyword before any point of instantiation (14.6.4.1  temp.point) of that template in that translation unit. In addition, the first declaration of an exported template containing the export keyword must not follow the definition of that template. The export keyword shall not be used in a friend declaration.

Templates defined in an unnamed namespace, inline functions, and inline function templates shall not be exported. An exported non-class template shall be defined only once in a program; no diagnostic is required. An exported non-class template need only be declared (and not necessarily defined) in a translation unit in which it is instantiated.

A non-exported non-class template must be defined in every translation unit in which it is implicitly instantiated (14.7.1  temp.inst), unless the corresponding specialization is explicitly instantiated (14.7.2  temp.explicit) in some translation unit; no diagnostic is required.

Note: This change also resolves issues 204 and 335.

335. Allowing export on template members of nontemplate classes

[Voted into WP at April 2003 meeting.]

The syntax for "export" permits it only on template declarations. Clause 14  temp paragraph 6 further restricts "export" to appear only on namespace scope declarations. This means that you can't export a member template of a non-template class, as in:

  class A {
    template <class T> void f(T);
  };

  export template <class T> void A<T>::f(T){}

There is also the issue of whether or not we should permit this usage:

  export struct A {
    template <class T> void f(T);
  };

If we eliminate the requirement that "export" precede "template" there is a similar issue regarding this case, which is currently prohibited:

  template <class T> struct B {
    export void f();
  };

Notes from the 4/02 meeting:

This is resolved by the proposed changes for issue 323.

184. Default arguments in template template-parameters

[Voted into WP at April 2003 meeting.]

John Spicer: Where can default values for the template parameters of template template parameters be specified and where so they apply?

For normal template parameters, defaults can be specified only in class template declarations and definitions, and they accumulate across multiple declarations in the same way that function default arguments do.

I think that defaults for parameters of template template parameters should be handled differently, though. I see no reason why such a default should extend beyond the template declaration with which it is associated. In other words, such defaults are a property of a specific template declaration and are not part of the interface of the template.

    template <class T = float> struct B {};

    template <template <class _T = float> class T> struct A {
        inline void f();
        inline void g();
    };

    template <template <class _T> class T> void A<T>::f() {
        T<> t;  // Okay? (proposed answer - no)
    }

    template <template <class _T = char> class T> // Okay? (proposed answer - yes)
    void A<T>::g() {
        T<> t;  // T<char> or T<float>?  (proposed answer - T<char>)
    }

    int main() {
        A<B> ab;
        ab.f();
    }

I don't think this is clear in the standard.

Gabriel Dos Reis: On the other hand I fail to see the reasons why we should introduce yet another special rule to handle that situation differently. I think we should try to keep rules as uniform as possible. For default values, it has been the case that one should look for any declaration specifying default values. Breaking that rules doesn't buy us anything, at least as far as I can see. My feeling is that [allowing different defaults in different declarations] is very confusing.

Mike Miller: I'm with John on this one. Although we don't have the concept of "prototype scope" for template parameter lists, the analogy with function parameters would suggest that the two declarations of T (in the template class definition and the template member function definition) are separate declarations and completely unrelated. While it's true that you accumulate default arguments on top-level declarations in the same scope, it seems to me a far leap to say that we ought also to accumulate default arguments in nested declarations. I would expect those to be treated as being in different scopes and thus not to share default argument information.

When you look up the name T in the definition of A<T>::f(), the declaration you find has no default argument for the parameter of T, so T<> should not be allowed.

Proposed Resolution (revised October 2002):

In 14.1  temp.param, add the following as a new paragraph at the end of this section:

A template-parameter of a template template-parameter is permitted to have a default template-argument. When such default arguments are specified, they apply to the template template-parameter in the scope of the template template-parameter. [Example:

    template <class T = float> struct B {};

    template <template <class TT = float> class T> struct A {
        inline void f();
        inline void g();
    };

    template <template <class TT> class T> void A<T>::f() {
        T<> t;  // error - TT has no default template argument
    }

    template <template <class TT = char> class T>void A<T>::g() {
        T<> t;  // OK - T<char>
    }

-- end example]

226. Default template arguments for function templates

[Voted into WP at April 2003 meeting.]

The prohibition of default template arguments for function templates is a misbegotten remnant of the time where freestanding functions were treated as second class citizens and required all template arguments to be deduced from the function arguments rather than specified.

The restriction seriously cramps programming style by unnecessarily making freestanding functions different from member functions, thus making it harder to write STL-style code.

Suggested resolution:

Replace

A default template-argument shall not be specified in a function template declaration or a function template definition, nor in the template-parameter-list of the definition of a member of a class template.

by

A default template-argument shall not be specified in the template-parameter-list of the definition of a member of a class template.

The actual rules are as stated for arguments to class templates.

Notes from 10/00 meeting:

The core language working group was amenable to this change. Questions arose, however, over the interaction between default template arguments and template argument deduction: should it be allowed or forbidden to specify a default argument for a deduced parameter? If it is allowed, what is the meaning: should one or the other have priority, or is it an error if the default and deduced arguments are different?

Notes from the 10/01 meeting:

It was decided that default arguments should be allowed on friend declarations only when the declaration is a definition. It was also noted that it is not necessary to insist that if there is a default argument for a given parameter all following parameters have default arguments, because (unlike in the class case) arguments can be deduced if they are not specified.

Note that there is an interaction with issue 115.

Proposed resolution (revised October 2002):

1. In 14.1  temp.param paragraph 9, replace

   A default template-argument may be specified in a class template declaration or a class template definition. A default template-argument shall not be specified in a function template declaration or a function template definition, nor in the template-parameter-list of the definition of a member of a class template.

   with

   A default template-argument may be specified in a template declaration. A default template-argument shall not be specified in the template-parameter-lists of the definition of a member of a class template that appears outside of the member's class.

2. In 14.1  temp.param paragraph 9, replace

   A default template-argument shall not be specified in a friend template declaration.

   with

   A default template-argument shall not be specified in a friend class template declaration. If a friend function template declaration specifies a default template-argument, that declaration shall be a definition and shall be the only declaration of the function template in the translation unit.

3. In 14.1  temp.param paragraph 11, replace

   If a template-parameter has a default template-argument, all subsequent template-parameters shall have a default template-argument supplied.

   with

   If a template-parameter of a class template has a default template-argument, all subsequent template-parameters shall have a default template-argument supplied. [Note: This is not a requirement for function templates because template arguments might be deduced (14.8.2  temp.deduct).]

4. In 14.8  temp.fct.spec paragraph 1, replace

   Template arguments can either be explicitly specified when naming the function template specialization or be deduced (14.8.2  temp.deduct) from the context, e.g. from the function arguments in a call to the function template specialization.

   with

   Template arguments can be explicitly specified when naming the function template specialization, deduced from the context (14.8.2  temp.deduct), e.g., deduced from the function arguments in a call to the function template specialization), or obtained from default template arguments.

5. In 14.8.1  temp.arg.explicit paragraph 2, replace

   Trailing template arguments that can be deduced (14.8.2  temp.deduct) may be omitted from the list of explicit template-arguments.

   with

   Trailing template arguments that can be deduced (14.8.2  temp.deduct) or obtained from default template-arguments may be omitted from the list of explicit template-arguments.

6. In 14.8.2  temp.deduct paragraph 1, replace

   The values can be either explicitly specified or, in some cases, deduced from the use.

   with

   The values can be explicitly specified or, in some cases, be deduced from the use or obtained from default template-arguments.

7. In 14.8.2  temp.deduct paragraph 4, replace

   The resulting substituted and adjusted function type is used as the type of the function template for template argument deduction. When all template arguments have been deduced, all uses of template parameters in nondeduced contexts are replaced with the corresponding deduced argument values. If the substitution results in an invalid type, as described above, type deduction fails.

   with

   The resulting substituted and adjusted function type is used as the type of the function template for template argument deduction. If a template argument has not been deduced, its default template argument, if any, is used. [Example:

       template <class T, class U = double>
       void f(T t = 0, U u = 0);
   
       void g()
       {
           f(1, 'c');         // f<int,char>(1,'c')
           f(1)               // f<int,double>(1,0)
           f();               // error: T cannot be deduced
           f<int>();          // f<int,double>(0,0)
           f<int,char>();     // f<int,char>(0,0)
       }
   

   ---end example]

   When all template arguments have been deduced or obtained from default template arguments, all uses of template parameters in nondeduced contexts are replaced with the corresponding deduced or default argument values. If the substitution results in an invalid type, as described above, type deduction fails.

228. Use of template keyword with non-member templates

[Voted into WP at April 2003 meeting.]

Consider the following example:

    template<class T>
    struct X {
       virtual void f();
    };

    template<class T>
    struct Y {
       void g(X<T> *p) {
	  p->template X<T>::f();
       }
    };

This is an error because X is not a member template; 14.2  temp.names paragraph 5 says:

If a name prefixed by the keyword template is not the name of a member template, the program is ill-formed.

In a way this makes perfect sense: X is found to be a template using ordinary lookup even though p has a dependent type. However, I think this makes the use of the template prefix even harder to teach.

Was this intentionally outlawed?

Proposed Resolution (4/02):

Elide the first use of the word "member" in 14.2  temp.names paragraph 5 so that its first sentence reads:

If a name prefixed by the keyword template is not the name of a member template, the program is ill-formed.

246. Jumps in function-try-block handlers

[Moved to DR at 4/01 meeting.]

Is it permitted to jump from a handler of a function-try-block into the body of the function?

15  except paragraph 2 would appear to disallow such a jump:

A goto, break, return, or continue statement can be used to transfer control out of a try block or handler, but not into one.

However, 15.3  except.handle paragraph 14 mentions only constructors and destructors for the prohibition:

If the handlers of a function-try-block contain a jump into the body of a constructor or destructor, the program is ill-formed.

Is this paragraph simply reemphasizing the more general restriction, or does it assume that such a jump would be permitted for functions other than constructors or destructors? If the former interpretation is correct, it is confusing and should be either eliminated or turned into a note. If the latter interpretation is accurate, 15  except paragraph 2 must be revised.

(See also issue 98.)

Proposed resolution (04/01):

Delete 15.3  except.handle paragraph 14.

62. Unnamed members of classes used as type parameters

[Moved to DR at 4/01 meeting.]

Section 14.3.1  temp.arg.type paragraph 2 says

A local type, a type with no linkage, an unnamed type or a type compounded from any of these types shall not be used as a template-argument for a template type-parameter.

It probably wasn't intended that classes with unnamed members should be included in this list, but they are arguably compounded from unnamed types.

Proposed resolution (04/01):

In 14.3.1  temp.arg.type paragraph 2, change

A local type, a type with no linkage, an unnamed type or a type compounded from any of these types shall not be used as a template-argument for a template type-parameter.

to

The following types shall not be used as a template-argument for a template type-parameter:

• a type whose name has no linkage
• an unnamed class or enumeration type that has no name for linkage purposes (7.1.3  dcl.typedef)
• a cv-qualified version of one of the types in this list
• a type created by application of declarator operators to one of the types in this list
• a function type that uses one of the types in this list

329. Evaluation of friends of templates

[Voted into WP at October 2003 meeting.]

14.5.3  temp.friend paragraph 5 says:

When a function is defined in a friend function declaration in a class template, the function is defined at each instantiation of the class template. The function is defined even if it is never used. The same restrictions on multiple declarations and definitions which apply to non-template function declarations and definitions also apply to these implicit definitions. [Note: if the function definition is ill-formed for a given specialization of the enclosing class template, the program is ill-formed even if the function is never used. ]

This means that the following program is invalid, even without the call of f(ai):

  template <class T> struct A {
    friend void f(A a) {
      g(a);
    }
  };
  int main()
  {
    A<int> ai;
  // f(ai);  // Error if f(ai) is actually called
  }

The EDG front end issues an error on this case even if f(ai) is never called. Of the compilers I tried (g++, Sun, Microsoft, Borland) we are the only ones to issue such an error.

This issue came up because there is a library that either deliberately or accidentally makes use of friend functions that are not valid for certain instantiations.

The wording in the standard is the result of a deliberate decision made long ago, but given the fact that most implementations do otherwise it raises the issue of whether we did the right thing.

Upon further investigation, the current rule was adopted as the resolution to issue 6.47 in my series of template issue papers. At the time the issue was discussed (7/96) most compilers did evaluate such friends. So it seems that a number of compilers have changed their behavior since then.

Based on current practice, I think the standard should be changed to evaluate such friends only when used.

Proposed resolution (October 2002):

Change section 14.5.3  temp.friend paragraph 5 from:

When a function is defined in a friend function declaration in a class template, the function is defined at each instantiation of the class template. The function is defined even if it is never used. The same restrictions on multiple declarations and definitions which apply to non-template function declarations and definitions also apply to these implicit definitions. [Note: if the function definition is ill-formed for a given specialization of the enclosing class template, the program is ill-formed even if the function is never used. ]

When a function is defined in a friend function declaration in a class template, the function is instantiated when the function is used. The same restrictions on multiple declarations and definitions that apply to non-template function declarations and definitions also apply to these implicit definitions.

410. Paragraph missed in changes for issue 166

[Voted into WP at October 2004 meeting.]

14.5.3  temp.friend paragraph 2 was overlooked when the changes for issue 166 were made.

The friend declaration of f<>(int) is now valid.

A friend function declaration that is not a template declaration and in which the name of the friend is an unqualified template-id shall refer to a specialization of a function template declared in the nearest enclosing namespace scope. [Example:

  namespace N {
          template <class T> void f(T);
          void g(int);
          namespace M {
                  template <class T> void h(T);
                  template <class T> void i(T);
                  struct A {
                          friend void f<>(int);   // ill-formed - N::f
                          friend void h<>(int);   // OK - M::h
                          friend void g(int);     // OK - new decl of M::g
                          template <class T> void i(T);
                          friend void i<>(int);   // ill-formed - A::i
                  };
          }
  }

--end example]

Proposed Resolution (October 2003):

Remove 14.5.3  temp.friend paragraph 2:

A friend function declaration that is not a template declaration and in which the name of the friend is an unqualified template-id shall refer to a specialization of a function template declared in the nearest enclosing namespace scope. [Example:

  namespace N {
          template <class T> void f(T);
          void g(int);
          namespace M {
                  template <class T> void h(T);
                  template <class T> void i(T);
                  struct A {
                          friend void f<>(int);   // ill-formed - N::f
                          friend void h<>(int);   // OK - M::h
                          friend void g(int);     // OK - new decl of M::g
                          template <class T> void i(T);
                          friend void i<>(int);   // ill-formed - A::i
                  };
          }
  }

--end example]

286. Incorrect example in partial specialization

[Moved to DR at 4/02 meeting.]

The example in 14.5.4  temp.class.spec paragraph 6 is incorrect. It reads,

    template<class T> struct A {
        class C {
            template<class T2> struct B { };
        };
    };

    // partial specialization of A<T>::C::B<T2>
    template<class T> template<class T2>
        struct A<T>::C::B<T2*> { };

    A<short>::C::B<int*> absip; // uses partial specialization

Because C is a class rather than a struct, the use of the name B is inaccessible.

Proposed Resolution (10/01):

Change class C to struct C in the example in 14.5.4  temp.class.spec paragraph 6. The example becomes

    template<class T> struct A {
        struct C {
            template<class T2> struct B { };
        };
    };

    // partial specialization of A<T>::C::B<T2>
    template<class T> template<class T2>
        struct A<T>::C::B<T2*> { };

    A<short>::C::B<int*> absip; // uses partial specialization

214. Partial ordering of function templates is underspecified

[Voted into WP at October 2003 meeting.]

In 14.5.5.2  temp.func.order, partial ordering is explained in terms of template argument deduction. However, the exact procedure for doing so is not specified. A number of details are missing, they are explained as sub-issues below.

1. 14.5.5.2  temp.func.order paragraph 2 refers to 14.8.2  temp.deduct for argument deduction. This is the wrong reference; it explains how explicit arguments are processed (paragraph 2) and how function parameter types are adjusted (paragraph 3). Neither of these steps is meaningful in the context of partial ordering. Next in deduction follows one of the steps in 14.8.2.1  temp.deduct.call, 14.8.2.2  temp.deduct.funcaddr, 14.8.2.3  temp.deduct.conv, or 14.8.2.5  temp.deduct.type. The standard does not specify which of these contexts apply to partial ordering.
2. Because 14.8.2.1  temp.deduct.call and 14.8.2.3  temp.deduct.conv both start with actual function parameters, it is meaningful to assume that partial ordering uses 14.8.2.5  temp.deduct.type, which only requires types. With that assumption, the example in 14.5.5.2  temp.func.order paragraph 5 becomes incorrect, considering the two templates

       template<class T> void g(T);  // #1
       template<class T> void g(T&); // #2
   

   Here, #2 is at least as specialized as #1: With a synthetic type U, #2 becomes g(U&); argument deduction against #1 succeeds with T=U&. However, #1 is not at least as specialized as #2: Deducing g(U) against g(T&) fails. Therefore, the second template is more specialized than the first, and the call g(x) is not ambiguous.
3. According to John Spicer, the intent of the partial ordering was that it uses deduction as in a function call (14.8.2.1  temp.deduct.call), which is indicated by the mentioning of "exact match" in 14.5.5.2  temp.func.order paragraph 4. If that is indeed the intent, it should be specified how values are obtained for the step in 14.8.2.1  temp.deduct.call paragraph 1, where the types of the arguments are determined. Also, since 14.8.2.1  temp.deduct.call paragraph 2 drops references from the parameter type, symmetrically, references should be dropped from the argument type (which is done in 5  expr paragraph 2, for a true function call).
4. 14.5.5.2  temp.func.order paragraph 4 requires an "exact match" for the "deduced parameter types". It is not clear whether this refers to the template parameters, or the parameters of the template function. Considering the example

       template<class S> void g(S);  // #1
       template<class T> void g(T const &); // #3
   

   Here, #3 is clearly at least as specialized as #1. To determine whether #1 is at least as specialized as #3, a unique type U is synthesized, and deduction of g<U>(U) is performed against #3. Following the rules in 14.8.2.1  temp.deduct.call, deduction succeeds with T=U. Since the template argument is U, and the deduced template parameter is also U, we have an exact match between the template parameters. Even though the conversion from U to U const & is an exact match, it is not clear whether the added qualification should be taken into account, as it is in other places.

Issue 200 covers a related issue, illustrated by the following example:

    template <class T> T f(int);
    template <class T, class U> T f(U);
    void g() {
        f<int>(1);
    }

Even though one template is "obviously" more specialized than the other, deduction fails in both directions because neither function parameter list allows template parameter T to be deduced.

(See also issue 250.)

Nathan Sidwell:

14.5.5.2  temp.func.order describes the partial ordering of function templates. Paragraph 5 states,

A template is more specialized than another if, and only if, it is at least as specialized as the other template and that template is not at least as specialized as the first.

  template <typename T, class U> T checked_cast(U from); //#1
  template <typename T, class U> T checked_cast(U * from); //#2
  class C {};

  void foo (int *arg)
  {
    checked_cast <C const *> (arg);
  }

#1 can be deduced with T = 'C const *' and U = 'int *'

#2 can be deduced with T = 'C const *' and U = 'int'

It looks like #2 is more specialized that #1, but 14.5.5.2  temp.func.order does not make it so, as neither template can deduce 'T' from the other template's function parameters.

Possible Resolutions:

There are several possible solutions, however through experimentation I have discounted two of them.

Option 1:

When deducing function ordering, if the return type of one of the templates uses a template parameter, then return types should be used for deduction. This, unfortunately, makes existing well formed programs ill formed. For example

  template <class T> class X {};

  template <class T> X<T> Foo (T *);	// #1
  template <class T> int Foo (T const *); // #2

  void Baz (int *p1, int const *p2)
  {
    int j = Foo (p2); //#3
  }

Option 2:

As option 1, but only consider the return type when deducing the template whose return type involves template parameters. This has the same flaw as option 1, and that example is similarly ill formed, as #1's return type 'X<T,0>' fails to match 'int' so #1 cannot deduce #2. In the converse direction, return types are not considered, but the function parameters fail to deduce.

Option 3:

It is observed that the original example is only callable with a template-id-expr to supply a value for the first, undeducible, parameter. If that parameter were deducible it would also appear within at least one of the function parameters. We can alter paragraph 4 of [temp.func.order] to indicate that it is not necessary to deduce the parameters which are provided explicitly, when the call has the form of a template-id-expr. This is a safe extension as it only serves to make ill formed programs well formed. It is also in line with the concept that deduction for function specialization order should proceed in a similar manner to function calling, in that explicitly provided parameter values are taken into consideration.

Suggested resolution:

Insert after the first sentence of paragraph 4 in  

Should any template parameters remain undeduced, and the function call be of the form of a template-id-expr, those template parameters provided in the template-id-expr may be arbitrarily synthesized prior to determining whether the deduced arguments generate a valid function type.

See also issue 200.

(April 2002) John Spicer and John Wiegley have written a paper on this. See 02-0051/N1393.

Proposed resolution (October 2002):

Change 14.5.5.2  temp.func.order paragraph 2 to read:

Partial ordering selects which of two function templates is more specialized than the other by transforming each template in turn (see next paragraph) and performing template argument deduction using the function parameter types, or in the case of a conversion function the return type. The deduction process determines whether one of the templates is more specialized than the other. If so, the more specialized template is the one chosen by the partial ordering process.

Change 14.5.5.2  temp.func.order paragraph 3 to read:

To produce the transformed template, for each type, non-type, or template template parameter synthesize a unique type, value, or class template respectively and substitute it for each occurrence of that parameter in the function type of the template.

Change 14.5.5.2  temp.func.order paragraph 4 to read (note: the section reference should refer to the section added to 14.8.2  temp.deduct below):

Using the transformed function template's function parameter list, or in the case of a conversion function its transformed return type, perform type deduction against the function parameter list (or return type) of the other function. The mechanism for performing these deductions is given in 14.8.2.x.

Remove the text of 14.5.5.2  temp.func.order paragraph 5 but retain the example. The removed text is:

A template is more specialized than another if, and only if, it is at least as specialized as the other template and that template is not at least as specialized as the first.

Insert the following section before 14.8.2.5 (Note that this would either be a new 14.8.2.4, or would be given a number like 14.8.2.3a. If neither of these is possible from a troff point of view, this could be made 14.8.2.5. )

Deducing template arguments when determining the partial ordering of function templates (temp.deduct.partial)

Template argument deduction is done by comparing certain types associated with the two function templates being compared.

Two sets of types are used to determine the partial ordering. For each of the templates involved there is the original function type and the transformed function type. [Note: The creation of the transformed type is described in 14.5.5.2  temp.func.order.] The deduction process uses the transformed type as the argument template and the original type of the other template as the parameter template. This process is done twice for each type involved in the partial ordering comparison: once using the transformed template-1 as the argument template and template-2 as the parameter template and again using the transformed template-2 as the argument template and template-1 as the parameter template.

The types used to determine the ordering depend on the context in which the partial ordering is

• In the context of a function call, the function parameter types are used.
• In the context of a call to a conversion operator, the return types of the conversion function templates are used.
• In other contexts (14.5.5.2  temp.func.order), the function template's function type is used.

Each type from the parameter template and the corresponding type from the argument template are used as the types of P and A.

Before the partial ordering is done, certain transformations are performed on the types used for partial ordering:

• If P is a reference type, P is replaced by the type referred to.
• If A is a reference type, A is replaced by the type referred to.

If both P and A were reference types (before being replaced with the type referred to above), determine which of the two types (if any) is more cv-qualified than the other; otherwise the types are considered to be equally cv-qualified for partial ordering purposes. The result of this determination will be used below.

Remove any top-level cv-qualifiers:

• If P is a cv-qualified type, P is replaced by the cv-unqualified version of P.
• If A is a cv-qualified type, A is replaced by the cv-unqualified version of A.

Using the resulting types P and A the deduction is then done as described in (14.8.2.5  temp.deduct.type). If deduction succeeds for a given type, the type from the argument template is considered to be at least as specialized as the type from the parameter template.

If, for a given type, deduction succeeds in both directions (i.e., the types are identical after the transformations above) if the type from the argument template is more cv-qualified than the type from the parameter template (as described above) that type is considered to be more specialized than the other. If neither type is more cv-qualified than the other then neither type is more specialized than the other.

If for each type being considered a given template is at least as specialized for all types and more specialized for some set of types and the other template is not more specialized for any types or is not at least as specialized for any types, then the given template is more specialized than the other template. Otherwise, neither template is more specialized than the other.

In most cases, all template parameters must have values in order for deduction to succeed, but for partial ordering purposes a template parameter may remain without a value provided it is not used in the types being used for partial ordering. [Note: A template parameter used in a non-deduced context is considered used.]

[Example:

template <class T> T f(int);        // #1
template <class T, class U> T f(U); // #2
void g() {
    f<int>(1);  // Calls #1
}

--end example]

180. typename and elaborated types

[Moved to DR at 4/02 meeting.]

Mike Miller: A question about typename came up in the discussion of issue 68 that is somewhat relevant to the idea of omitting typename in contexts where it is clear that a type is required: consider something like

        template <class T>
        class X {
            friend class T::nested;
        };

Bill Gibbons: The class applies to the last identifier in the qualified name, since all the previous names must be classes or namespaces. Since the name is specified to be a class it does not need typename. [However,] it looks like 14.6  temp.res paragraph 3 requires typename and the following paragraphs do not exempt this case. This is not what we agreed on.

Proposed resolution (04/01):

In 14.6  temp.res paragraph 5, change

The keyword typename is not permitted in a base-specifier or in a mem-initializer; in these contexts a qualified-name that depends on a template-parameter (14.6.2  temp.dep) is implicitly assumed to be a type name.

to

A qualified name used as the name in a mem-initializer-id, a base-specifier, or an elaborated-type-specifier (in the class-key and enum forms) is implicitly assumed to name a type, without the use of the typename keyword. [Note: the typename keyword is not permitted by the syntax of these constructs.]