| Document number: | PL22.16/09-0196 = WG21 N3006 |
| Date: | 2009-11-08 |
| Project: | Programming Language C++ |
| Reference: | ISO/IEC IS 14882:2003 |
| Reply to: | William M. Miller |
| Edison Design Group, Inc. | |
| wmm@edg.com |
This document contains the C++ core language issues on which the Committee (J16 + WG21) has not yet acted, that is, issues with status "Ready," "Tentatively Ready," "Review," "Drafting," and "Open."
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:
Section references in this document reflect the section numbering of document PL22.16/09-0150 = WG21 N2960.
The purpose of these documents is to record the disposition of issues that have come before the Core Language Working Group of the ANSI (INCITS PL22.16) and ISO (WG21) C++ Standard Committee.
Some issues represent potential defects in the ISO/IEC IS 14882:2003 document and corrected defects in the earlier ISO/IEC 14882:1998 document; others refer to text in the working draft for the next revision of the C++ language, informally known as C++0x, and not to any Standard text. Issues are not necessarily formal ISO Defect Reports (DRs). While some issues will eventually be elevated to DR status, others will be disposed of in other ways. (See Issue Status below.)
The most current public version of this document can be found at http://www.open-std.org/jtc1/sc22/wg21. Requests for further information about these documents should include the document number, reference ISO/IEC 14882:2003, and be submitted to the InterNational Committee for Information Technology Standards (INCITS), 1250 Eye Street NW, Suite 200, Washington, DC 20005, USA.
Information regarding how to obtain a copy of the C++ Standard, join the Standard Committee, or submit an issue can be found in the C++ FAQ at http://www.comeaucomputing.com/csc/faq.html. Public discussion of the C++ Standard and related issues occurs on newsgroup comp.std.c++.
Issues progress through various statuses as the Core Language Working Group and, ultimately, the full PL22.16 and WG21 committees deliberate and act. For ease of reference, issues are grouped in these documents by their status. Issues have one of the following statuses:
Open: The issue is new or the working group has not yet formed an opinion on the issue. If a Suggested Resolution is given, it reflects the opinion of the issue's submitter, not necessarily that of the working group or the Committee as a whole.
Drafting: Informal consensus has been reached in the working group and is described in rough terms in a Tentative Resolution, although precise wording for the change is not yet available.
Review: Exact wording of a Proposed Resolution is now available for an issue on which the working group previously reached informal consensus.
Ready: The working group has reached consensus that the issue is a defect in the Standard, the Proposed Resolution is correct, and the issue is ready to forward to the full Committee for ratification as a proposed defect report.
Tentatively Ready: Like "ready" except that the resolution was produced and approved by a subset of the working group membership between meetings. Persons not participating in these betwee-meeting activities are encouraged to review such resolutions carefully and to alert the working group with any problems that may be found.
DR: The full Committee has approved the item as a proposed defect report. The Proposed Resolution in an issue with this status reflects the best judgment of the Committee at this time regarding the action that will be taken to remedy the defect; however, the current wording of the Standard remains in effect until such time as a Technical Corrigendum or a revision of the Standard is issued by ISO.
TC1: A DR issue included in Technical Corrigendum 1. TC1 is a revision of the Standard issued in 2003.
CD1: A DR issue not resolved in TC1 but included in Committee Draft 1. CD1 was advanced for balloting at the September, 2008 WG21 meeting.
WP: A DR issue whose resolution is reflected in the current Working Paper. The Working Paper is a draft for a future version of the Standard.
Dup: The issue is identical to or a subset of another issue, identified in a Rationale statement.
NAD: The working group has reached consensus that the issue is not a defect in the Standard. A Rationale statement describes the working group's reasoning.
Extension: The working group has reached consensus that the issue is not a defect in the Standard but is a request for an extension to the language. The working group expresses no opinion on the merits of an issue with this status; however, the issue will be maintained on the list for possible future consideration as an extension proposal.
Concepts: The issue relates to the “Concepts” proposal that was removed from the working paper at the Frankfurt (July, 2009) meeting and hence is no longer under consideration.
1.10 [intro.multithread] paragraph 12 says,
A visible side effect A on an object M with respect to a value computation B of M satisfies the conditions:
A happens before B, and
there is no other side effect X to M such that A happens before X and X happens before B.
The value of a non-atomic scalar object M, as determined by evaluation B, shall be the value stored by the visible side effect A. [Note: If there is ambiguity about which side effect to a non-atomic object is visible, then there is a data race, and the behavior is undefined. —end note]
The note here suggests that, except in the case of a data race, visible side effects to value computation can always be determined. But unsequenced and indeterminately sequenced side effects on the same object create ambiguities with respect to a later value computation as well. So the wording needs to be revisited, see the following examples.
int main(){
int i = 0;
i = // unsequenced side effect A
i++; // unsequenced side effect B
return i; // value computation C
}
According to the definition in the draft, both A and B are visible side effects to C. However, there is no data race, because (paragraph 14) a race involves at least two threads. So the note in paragraph 12 is logically false.
The model introduces the special case of indeterminately sequenced side effects, that leave open what execution order is taken in a concrete situation. If the execution paths access the same data, unpredictable results are possible, just as it is the case with data races. Whereas data races constitute undefined behavior, indeterminatedly sequenced side effects on the same object do not. As a consequence of this disparity, indeterminately sequenced execution occasionally needs exceptional treatment.
int i = 0;
int f(){
return
i = 1; // side effect A
}
int g(){
return
i = 2; // side effect B
}
int h(int, int){
return i; // value computation C
}
int main(){
return h(f(),g()); // function call D returns 1 or 2?
}
Here, either A or B is the visible side effect on the value computation C, but you cannot tell which (cf. 1.9 [intro.execution] paragraph 16). Although an ambiguity is present, it is neither because of a data race, nor is the behavior undefined, in total contradiction to the note.
Proposed resolution (October, 2009):
Change 1.10 [intro.multithread] paragraph 12 as follows:
...The value of a non-atomic scalar object or bit-field M, as determined by evaluation B, shall be the value stored by the visible side effect A. [Note: If there is ambiguity about which side effect to a non-atomic object or bit-field is visible, then there is a data race, and the behavior is either unspecified or undefined. —end note]...
The specification of raw string literals interacts poorly with the specification of preprocessing tokens. The grammar in 2.5 [lex.pptoken] has a production reading
This is echoed in the max-munch rule in paragraph 3:
If the input stream has been parsed into preprocessing tokens up to a given character, the next preprocessing token is the longest sequence of characters that could constitute a preprocessing token, even if that would cause further lexical analysis to fail.
This raises questions about the handling of raw string literals. Consider, for instance,
#define R "x"
const char* s = R"y";
The character sequence R"y" does not satisfy the syntactic requirements for a raw string. Should it be diagnosed as an ill-formed attempt at a raw string, or should it be well-formed, interpreting R as a preprocessor token that is a macro name and thus initializing s with a pointer to the string "xy"?
For another example, consider:
#define R "]"
const char* x = R"foo[";
Presumably this means that the entire rest of the file must be scanned for the characters ]foo" and, if they are not found, macro-expand R and initialize x with a pointer to the string "]foo[". Is this the intended result?
Finally, does the requirement in 2.14.5 [lex.string] that
A d-char-sequence shall consist of at most 16 characters.
mean that
#define R "x"
const char* y = R"12345678901234567[y]12345678901234567";
is ill-formed, or a valid initialization of y with a pointer to the string "x12345678901234567[y]12345678901234567"?
Additional note, June, 2009:
The translation of characters that are not in the basic source character set into universal-character-names in translation phase 1 raises an additional problem: each such character will occupy at least six of the 16 r-chars that are permitted. Thus, for example, R"@@@[]@@@" is ill-formed because @@@ becomes \u0040\u0040\u0040, which is 18 characters.
One possibility for addressing this might be to disallow the \ character completely as an d-char, which would have the effect of restricting r-chars to the basic source character set.
Proposed resolution (October, 2009):
Change the grammar in 2.14.5 [lex.string] as follows:
Change 2.14.5 [lex.string] paragraph 2 as follows:
A string literal that has an R in the prefix is a raw string literal. The d-char-sequence serves as a delimiter. The terminating d-char-sequence of a raw-string is the same sequence of characters as the initial d-char-sequence. A d-char-sequence shall consist of at most 16 characters. If the input stream contains a sequence of characters that could be the prefix and initial double quote of a raw string literal, such as R", those characters are considered to begin a raw string literal even if that literal is not well-formed. [Example:
#define R "x" const char* s = R"y"; // ill-formed raw string, not "x" "y"—end example]
Since members of the basic source character set can be written inside a string using a universal character name, it is not clear whether a UCN that represents ']' or one of the characters in the terminating d-char-sequence should be interpreted as that character or as an attempt to “escape” that character and prevent its interpretation as part of the terminating sequence of a raw character string.
Notes from the July, 2009 meeting:
The CWG supported a resolution in which the d-char-sequence of a raw string literal is considered to be outside the literal and thus, by 2.3 [lex.charset] paragraph 2, could not contain a UCN designating a member of the basic source character set.
Proposed resolution (October, 2009):
Change 2.3 [lex.charset] paragraph 2 as follows:
Additionally, if the hexadecimal value for a universal-character-name outside the c-char-sequence, s-char-sequence, or r-char-sequence of a character or string literal corresponds to a control character (in either of the ranges 0x00-0x1F or 0x7F-0x9F, both inclusive) or to a character in the basic source character set, the program is ill-formed.
2.14.8 [lex.ext] paragraph 5 says,
If L is a user-defined-string-literal, let str be the literal without its ud-suffix and let len be the number of characters (or code points) in str (i.e., its length excluding the terminating null character).
The length of a null-terminated string is defined in 17.5.2.1.4.1 [byte.strings] as the number of bytes preceding the terminator, but a single code point in a UTF-8 string can require more than one byte, so this sentence is inconsistent and needs to be revised to make clear which definition is in view.
Proposed resolution (October, 2009):
Change 2.14.8 [lex.ext] paragraph 5 as follows:
If L is a user-defined-string-literal, let str be the literal without its ud-suffix and let len be the number of characters (or code points) code units in str (i.e., its length excluding the terminating null character)...
At least in the new wording for 5.1.2 [expr.prim.lambda] paragraph 10 as found in paper N2927, this is explicitly assumed to be an entity. It should be investigated whether this should be added to the list of entities found in 3 [basic] paragraph 3.
Proposed resolution (October, 2009):
Change 3 [basic] paragraph 3 as follows:
An entity is a value, object, variable, reference, function, enumerator, type, class member, template, template specialization, namespace, or parameter pack, or this.
Change 3.2 [basic.def.odr] paragraph 2 as follows:
...is immediately applied. this is used if it appears as a potentially-evaluated expression (including as the result of the implicit transformation in the body of a non-static member function (9.3.1 [class.mfct.non-static])). A virtual member function...
Delete 5.1.2 [expr.prim.lambda] paragraph 7:
For the purpose of describing the behavior of lambda-expressions below, this is considered to be “used” if replacing this by an invented variable v with automatic storage duration and the same type as this would result in v being used (3.2 [basic.def.odr]).
The Standard uses the terms “block scope” and “local scope” interchangeably, but the former is never formally defined. Would it be better to use only one term consistently? “Block scope” seems to be more frequently used.
Notes from the October, 2007 meeting:
The CWG expressed a preference for the term “local scope.”
Notes from the September, 2008 meeting:
Reevaluating the relative prevalence of the two terms (including the fact that new uses of “block scope” are being introduced, e.g., in both the lambda and thread-local wording) led to CWG reversing its previous preference for “local scope.” The resolution will need to add a definition of “block scope” and should change the title of 3.3.3 [basic.scope.local].
Proposed resolution (October, 2009):
Change 3.3.2 [basic.scope.pdecl] paragraph 2 as follows:
[Note: a nonlocal name from an outer scope remains visible up to the point of declaration of the local name that hides it. [Example:
const int i = 2; { int i[i]; }declares a local block-scope array of two integers. —end example] —end note]
Change the section heading of 3.3.3 [basic.scope.local] from “Local scope” to “Block scope.”
Change 3.3.3 [basic.scope.local] paragraph 1 as follows:
A name declared in a block (6.3 [stmt.block]) is local to that block; it has block scope. Its potential scope begins at its point of declaration (3.3.2 [basic.scope.pdecl]) and ends at the end of its declarative region block. A variable declared at block scope is a local variable.
Change 3.3.3 [basic.scope.local] paragraph 3 as follows:
The name in a catch exception-declaration declared in an exception-declaration is local to the handler handler and shall not be redeclared in the outermost block of the handler handler.
Change 3.3.11 [basic.scope.hiding] paragraph 3 as follows:
In a member function definition, the declaration of a local name at block scope hides the declaration of a member of the class with the same name...
Change 3.5 [basic.link] paragraph 8 as follows:
...Moreover, except as noted, a name declared in a local at block scope (3.3.3 [basic.scope.local]) has no linkage...
Change 3.6.3 [basic.start.term] paragraph 1 as follows:
...For an object of array or class type, all subobjects of that object are destroyed before any local block-scope object with static storage duration initialized during the construction of the subobjects is destroyed.
Change 3.6.3 [basic.start.term] paragraph 2 as follows:
If a function contains a local block-scope object of static or thread storage duration that has been destroyed and the function is called during the destruction of an object with static or thread storage duration, the program has undefined behavior if the flow of control passes through the definition of the previously destroyed local block-scope object. Likewise, the behavior is undefined if the function-local block-scope object is used indirectly (i.e., through a pointer) after its destruction.
Change 3.6.3 [basic.start.term] paragraph 3 as follows:
If the completion of the initialization of a non-local non-block-scope object with static storage duration is sequenced before a call to std::atexit (see <cstdlib>, 18.5 [support.start.term]), the call to the function passed to std::atexit is sequenced before the call to the destructor for the object. If a call to std::atexit is sequenced before the completion of the initialization of a non-local non-block-scope object with static storage duration, the call to the destructor...
[Editorial note: the occurrences of “non-local” in this change are removed by the proposed resolution for issue 946.]
Change 6.3 [stmt.block] paragraph 1 as follows:
...A compound statement defines a local block scope (3.3 [basic.scope])...
Change 6.4 [stmt.select] paragraph 1 as follows:
...The substatement in a selection-statement (each substatement, in the else form of the if statement) implicitly defines a local block scope (3.3 [basic.scope])...
Change 6.4 [stmt.select] paragraph 5 as follows:
If a condition can be syntactically resolved as either an expression or the declaration of a local block-scope name, it is interpreted as a declaration.
Change 6.5 [stmt.iter] paragraph 2 as follows:
The substatement in an iteration-statement implicitly defines a local block scope (3.3 [basic.scope]) which is entered and exited each time through the loop.
Change 6.7 [stmt.dcl] paragraph 3 as follows:
...A program that jumps84 from a point where a local variable with automatic storage duration...
Change 6.7 [stmt.dcl] paragraph 4 as follows:
The zero-initialization (8.5 [dcl.init]) of all local block-scope objects with static storage duration (3.7.1 [basic.stc.static]) or thread storage duration (3.7.2 [basic.stc.thread]) is performed before any other initialization takes place. Constant initialization (3.6.2 [basic.start.init]) of a local block-scope entity with static storage duration, if applicable, is performed before its block is first entered. An implementation is permitted to perform early initialization of other local block-scope objects...
Change 6.7 [stmt.dcl] paragraph 5 as follows:
The destructor for a local block-scope object with static or thread storage duration will be executed if and only if the variable was constructed. [Note: 3.6.3 [basic.start.term] describes the order in which local block-scope objects with static and thread storage duration are destroyed. —end note]
Change 8.4 [dcl.fct.def] paragraph 7 as follows:
In the function-body, a function-local predefined variable denotes a local block-scope object of static storage duration that is implicitly defined (see 3.3.3 [basic.scope.local]).
Change the example in 9.1 [class.name] paragraph 2 as follows:
...
void g() {
struct s; // hide global struct s
// with a local block-scope declaration
...
Change the example in 9.1 [class.name] paragraph 3 as follows:
...
void g(int s) {
struct s* p = new struct s; // global s
p->a = s; // local parameter s
}
18.5 [support.start.term] paragraph 7 says that the order of destruction of objects with static storage duration and calls to functions registered by calling std::atexit is given in 3.6.3 [basic.start.term]. Paragraph 1 of 3.6.3 [basic.start.term] says,
If the completion of the constructor or dynamic initialization of an object with static storage duration is sequenced before that of another, the completion of the destructor of the second is sequenced before the initiation of the destructor of the first.
This wording covers both local and namespace-scope objects, so it fixes the relative ordering of local object destructors with respect to those of namespace scope. Paragraph 3 says,
If the completion of the initialization of a non-local object with static storage duration is sequenced before a call to std::atexit (see <cstdlib>, 18.5 [support.start.term]), the call to the function passed to std::atexit is sequenced before the call to the destructor for the object. If a call to std::atexit is sequenced before the completion of the initialization of a non-local object with static storage duration, the call to the destructor for the object is sequenced before the call to the function passed to std::atexit.
This fixes the relative ordering of destructors for namespace scope objects with respect to calls of atexit functions. However, the relative ordering of local destructors and atexit functions is left unspecified.
In the 2003 Standard, this was clear: 18.3 paragraph 8 said,
A local static object obj3 is destroyed at the same time it would be if a function calling the obj3 destructor were registered with atexit at the completion of the obj3 constructor.
Proposed resolution (October, 2009):
Change 3.6.3 [basic.start.term] paragraph 3 as follows:
If the completion of the initialization of a non-local an object with static storage duration is sequenced before a call to std::atexit (see <cstdlib>, 18.5 [support.start.term]), the call to the function passed to std::atexit is sequenced before the call to the destructor for the object. If a call to std::atexit is sequenced before the completion of the initialization of a non-local an object with static storage duration, the call to the destructor for the object is sequenced before the call to the function passed to std::atexit...
According to 20.8.12.6 [util.dynamic.safety] paragraph 16, when std::get_pointer_safety() returns std::pointer_safety::relaxed,
pointers that are not safely derived will be treated the same as pointers that are safely derived for the duration of the program.
However, 3.7.4.3 [basic.stc.dynamic.safety] paragraph 4 says unconditionally that
If a pointer value that is not a safely-derived pointer value is dereferenced or deallocated, and the referenced complete object is of dynamic storage duration and has not previously been declared reachable (20.8.12.6 [util.dynamic.safety]), the behavior is undefined.
This is a contradiction: the library clause attempts to constrain undefined behavior, which by definition is unconstrained.
Proposed resolution (July, 2009):
Change 3.7.4.3 [basic.stc.dynamic.safety] paragraph 4 as follows to define the terms “strict pointer safety” and “relaxed pointer safety,” which could then be used by the library clauses to achieve the desired effect:
An implementation may have relaxed pointer safety, in which case the validity of a pointer value does not depend on whether it is a safely-derived pointer value or not. Alternatively, an implementation may have strict pointer safety, in which case if If a pointer value that is not a safely-derived pointer value is dereferenced or deallocated, and the referenced complete object is of dynamic storage duration and has not previously been declared reachable (20.8.12.6 [util.dynamic.safety]), the behavior is undefined. [Note: this is true even if the unsafely-derived pointer value might compare equal to some safely-derived pointer value. —end note] It is implementation-defined whether an implementation has relaxed or strict pointer safety.
Read literally, 3.8 [basic.life] paragraphs 1 and 5 would make any access to non-static members of a class from the class's destructor undefined behavior. This is clearly not the intent.
Proposed resolution (October, 2009):
Change 3.8 [basic.life] paragraphs 5-6 as follows:
...any pointer that refers to the storage location where the object will be or was located may be used but only in limited ways. Such For an object under construction or destruction, see 12.7 [class.cdtor]. Otherwise, such a pointer refers to allocated storage...
...any lvalue which refers to the original object may be used but only in limited ways. Such For an object under construction or destruction, see 12.7 [class.cdtor]. Otherwise, such an lvalue refers to allocated storage...
3.10 [basic.lval] paragraph 7 says,
Whenever an lvalue appears in a context where an rvalue is expected, the lvalue is converted to an rvalue
That is not correct in the context of an attempt to bind an rvalue reference to an lvalue (8.5.3 [dcl.init.ref]).
Proposed resolution (October, 2009):
Change 3.10 [basic.lval] paragraph 7 as follows:
Whenever an lvalue appears in a context where an rvalue is expected and an lvalue is not explicitly prohibited (as, for example, in 8.5.3 [dcl.init.ref]), the lvalue it is converted to an rvalue; see 4.1 [conv.lval], 4.2 [conv.array], and 4.3 [conv.func].
The grammar for nested-name-specifier in 5.1.1 [expr.prim.general] paragraph 7 does not allow decltype to be used in a qualified-id. This could be useful for cases like:
auto vec = get_vec(); decltype(vec)::value_type v = vec.first();(See also issue 950.)
Proposed resolution (September, 2009):
See paper PL22.16/09-0181 = WG21 N2991.
The following is not allowed by the current syntax of lambda-capture but would be useful:
template <typename ...Args> void f(Args... args) {
auto l = [&, args...] { return g(args...); };
}
Proposed resolution (October, 2009):
Change the grammar in 5.1.2 [expr.prim.lambda] paragraph 1 as follows:
Add a new paragraph at the end of 5.1.2 [expr.prim.lambda]:
A capture followed by an ellipsis is a pack expansion (14.6.3 [temp.variadic]). [Example:
template<typename ...Args> void f(Args... args) { auto l = [&, args...] { return g(args...); }; l(); }—end example]
Add a new bullet to the list in 14.6.3 [temp.variadic] paragraph 4:
[Editorial note: the editor may wish to consider sorting the bullets in this list in order of section reference.]
The specification of the members of a closure type does not rule out the possibility that its operator() might be virtual. It would be better to make it clear that it cannot.
Proposed resolution (October, 2009):
Change 5.1.2 [expr.prim.lambda] paragraph 5 as follows:
... followed by mutable. It is not neither virtual nor declared volatile. Default arguments...
The note in 5.2.10 [expr.reinterpret.cast] paragraph 2 says,
Subject to the restrictions in this section, an expression may be cast to its own type using a reinterpret_cast operator.
However, there is nothing in the normative text that permits this conversion, and paragraph 1 forbids any conversion not explicitly permitted.
(See also issue 944.)
Proposed resolution (October, 2009):
Change 5.2.10 [expr.reinterpret.cast] paragraph 2 as follows:
The reinterpret_cast operator shall not cast away constness (5.2.11 [expr.const.cast]). [Note: Subject to the restrictions in this section, an expression may be cast to its own type using a reinterpret_cast operator. —end note] An expression of integral, enumeration, pointer, or pointer-to-member type can be explictly converted to its own type; such a cast yields the value of its operand.
Change 5.2.10 [expr.reinterpret.cast] paragraph 10 as follows:
An rvalue of type “pointer to member of X of type T1” can be explicitly converted to an rvalue of a different type “pointer to member of Y of type T2” if T1 and T2 are both function types or both object types...
The resolution of issue 39 changed the diagnosis of ambiguity because of multiple subobjects from being a lookup error to being diagnosed where the result of the lookup is used. The formation of a pointer to member is one such context but was overlooked in the changes. 5.3.1 [expr.unary.op] paragraph 3 should have language similar to 5.2.5 [expr.ref] paragraph 5 and should be mentioned in the note in 10.2 [class.member.lookup] paragraph 13.
Proposed resolution (October, 2009):
Change 5.3.1 [expr.unary.op] paragraph 3 as follows:
...For a qualified-id, if the member is a static member of type “T”, the type of the result is plain “pointer to T.” If the member is a non-static member of class C of type T, the type of the result is “pointer to member of class C of type T.,” and the program is ill-formed if C is an ambiguous base (10.2 [class.member.lookup]) of the class designated by the nested-name-specifier of the qualified-id....
Change 10.2 [class.member.lookup] paragraph 13 as follows:
[Note: Even if the result of name lookup is unambiguous, use of a name found in multiple subobjects might still be ambiguous (4.11 [conv.mem], 5.2.5 [expr.ref], 5.3.1 [expr.unary.op], 11.2 [class.access.base]). —end note] [Example:...
The current wording of the draft does not indicate what is supposed to happen when an rvalue of type std::nullptr_t is compared with an integral null pointer constant. (This could occur, for example, in template code like
template<typename T> void f(T t) {
if (t == 0) // ...
}
in a call like f(nullptr) -- presumably the body of the template was written before nullptr became available and thus used an integral null pointer constant.) Because an integral null pointer constant can be converted to std::nullptr_t (4.10 [conv.ptr] paragraph 1), one might expect that 0 would be converted to std::nullptr_t and the two operands would compare equal, but 5.9 [expr.rel] paragraph 2 does not handle this case at all, leaving it as undefined behavior.
The current situation is more well-defined (but perhaps not better) with respect to the conditional operator. 5.16 [expr.cond] paragraphs 3-6 make it ill-formed to have std::nullptr_t and 0 as the second and third operands. Again, it's not too hard to imagine a legacy function template like
template<typename T> void f(T t, bool b) {
T t = b ? t : 0;
}
which would be ill-formed under the current wording of 5.16 [expr.cond].
Either 5.9 [expr.rel] and 5.10 [expr.eq] should be changed to make this combination of operands ill-formed, or those two sections should be changed to give the comparison defined semantics and 5.16 [expr.cond] should be changed to make those operands well-formed.
Proposed resolution (October, 2009):
Change 5.9 [expr.rel] paragraph 2 as follows:
The usual arithmetic conversions are performed on operands of arithmetic or enumeration type. Pointer conversions (4.10 [conv.ptr]) and qualification conversions (4.4 [conv.qual]) are performed on pointer operands (or on a pointer operand and a null pointer constant, or on two null pointer constants, at least one of which is non-integral) to bring them to their composite pointer type. If one operand is a null pointer constant, the composite pointer type is std::nullptr_t if the other operand is also a null pointer constant or, if the other operand is a pointer, the type of the other operand. Otherwise...
Change 5.16 [expr.cond] paragraph 6 bullet 3 as follows:
According to 7.1.1 [dcl.stc] paragraph 4,
The thread_local specifier shall be applied only to the names of objects or references of namespace scope and to the names of objects or references of block scope that also specify static.
Why require two keywords, where one on its own becomes ill-formed? thread_local should imply static in this case, and the combination of keywords should be banned rather than required. This would also eliminate the one of two exceptions documented in paragraph 1.
Notes from the July, 2009 meeting:
The consensus of the CWG was that thread_local should imply static, as suggested, but that the combination should still be allowed (it is needed, for example, for thread-local static data members).
Proposed resolution (October, 2009):
Change 7.1.1 [dcl.stc] paragraph 4 as follows:
The thread_local specifier indicates that the named entity has thread storage duration (3.7.2 [basic.stc.thread]). It shall be applied only to the names of objects or references of namespace scope, to the names of objects or references of or block scope that also specify extern or static, and to the names of static data members. It specifies that the named object or reference has thread storage duration (3.7.2 [basic.stc.thread]). When thread_local is applied to a variable of block scope the storage-class-specifier static is implied if it does not appear explicitly.
The normative text in 7.1.6.1 [dcl.type.cv] paragraph 2 reads,
An object declared in namespace scope with a const-qualified type has internal linkage unless it is explicitly declared extern or unless it was previously declared to have external linkage. A variable of non-volatile const-qualified integral or enumeration type initialized by an integral constant expression can be used in integral constant expressions (5.19 [expr.const]).
These two sentences parallel the specifications of 7.1.1 [dcl.stc] paragraph 7 and 5.19 [expr.const]. However, the passages are not identical, leading to questions about whether the meanings are the same.
Proposed resolution (October, 2009):
Change 7.1.6.1 [dcl.type.cv] paragraph 2 as follows:
An object declared in namespace scope with a const-qualified type has internal linkage unless it is explicitly declared extern or unless it was previously declared to have external linkage. A variable of non-volatile const-qualified integral or enumeration type initialized by an integral constant expression can be used in integral constant expressions (5.19 [expr.const]). [Note: Declaring a variable const can affect its linkage (7.1.1 [dcl.stc]) and its usability in constant expressions (5.19 [expr.const]). As as described in 8.5 [dcl.init], the definition of an object or subobject of const-qualified type must specify an initializer or be subject to default-initialization. —end note]
References to references are ill-formed, but special provision is made in cases where this occurs via typedefs or template type parameters. A similar provision is probably needed for types resulting from decltype:
int x, *p = &x;
decltype(*p) &y = *p; // reference to reference is ill-formed
Proposed resolution (October, 2009):
Delete 7.1.3 [dcl.typedef] paragraph 9:
If a typedef TD names a type that is a reference to a type T, an attempt to create the type “lvalue reference to cv TD” creates the type “lvalue reference to T,” while an attempt to create the type “rvalue reference to cv TD” creates the type TD. [Example: ... —end example]
Delete 14.4.1 [temp.arg.type] paragraph 4:
If a template-argument for a template-parameter T names a type that is a reference to a type A, an attempt to create the type “lvalue reference to cv T” creates the type “lvalue reference to A,” while an attempt to create the type “rvalue reference to cv T” creates the type T [Example: ... —end example]
Add the following as a new paragraph at the end of 8.3.2 [dcl.ref]:
If a typedef (7.1.3 [dcl.typedef]), a type template-parameter (14.4.1 [temp.arg.type]), or a decltype-specifier (7.1.6.2 [dcl.type.simple]) denotes a type TR that is a reference to a type T, an attempt to create the type “lvalue reference to cv TR” creates the type “lvalue reference to T,” while an attempt to create the type “rvalue reference to cv TR” creates the type TR. [Example:
int i; typedef int& LRI; typedef int&& RRI; LRI& r1 = i; // r1 has the type int& const LRI& r2 = i; // r2 has the type int& const LRI&& r3 = i; // r3 has the type int& RRI& r4 = i; // r4 has the type int& RRI&& r5 = i; // r5 has the type int&& decltype(r2)& r6 = i; // r6 has the type int& decltype(r2)&& r7 = i; // r7 has the type int&—end example]
There is a lack of symmetry in the specification of attributes that apply to class and enum types. For example:
class X [[attr]]; // #1
typedef class Y [[attr]] YT; // #2
According to 7.1.6.3 [dcl.type.elab] paragraph 1, #1 associates the attr attribute with class X for all subsequent references. On the other hand, 8.3 [dcl.meaning] paragraph 5 says that #2 associates the attr attribute with the type but not with class Y.
Existing implementations (Microsoft, GNU, Sun) with attributes place an attribute that is intended to be associated with a class type between the class-key and the class name, and it would be preferable to adopt such an approach instead of the contextual approach in the current formulation.
Proposed resolution (October, 2009):
Change 3.3.2 [basic.scope.pdecl] paragraph 6 bullet 1 as follows:
for a declaration of the form
the identifier is declared...
Change 3.4.4 [basic.lookup.elab] paragraph 2 as follows:
...unless the elaborated-type-specifier appears in a declaration with the following form:
class-key attribute-specifieropt identifier attribute-specifieropt ;
the identifier is looked up... if the elaborated-type-specifier appears in a declaration with the form:
class-key attribute-specifieropt identifier attribute-specifieropt ;
the elaborated-type-specifier is a declaration...
In 7.1.6.3 [dcl.type.elab], change the grammar and paragraph 1 as follows:
elaborated-type-specifier:
class-key attribute-specifieropt ::opr nested-name-specifieropt identifier
...An attribute-specifier shall not appear in an elaborated-type-specifier unless the latter is the sole constituent of a declaration. If an elaborated-type-specifier is the sole constituent of a declaration, the declaration is ill-formed unless it is an explicit specialization (14.8.3 [temp.expl.spec]), an explicit instantiation (14.8.2 [temp.explicit]) or it has one of the following forms:
class-key attribute-specifieropt identifier attribute-specifieropt ;
...
Change the grammar in 7.2 [dcl.enum] paragraph 1 as follows:
Change the grammar in 9 [class] paragraph 1 as follows:
The auto specifier can be used only in certain contexts, as specified in 7.1.6.4 [dcl.spec.auto] paragraphs 2-3:
Otherwise (auto appearing with no type specifiers other than cv-qualifiers), the auto type-specifier signifies that the type of an object being declared shall be deduced from its initializer. The name of the object being declared shall not appear in the initializer expression.
This use of auto is allowed when declaring objects in a block (6.3 [stmt.block]), in namespace scope (3.3.6 [basic.scope.namespace]), and in a for-init-statement (6.5.3 [stmt.for]). The decl-specifier-seq shall be followed by one or more init-declarators, each of which shall have a non-empty initializer of either of the following forms:
= assignment-expression
( assignment-expression )
It was intended that auto could be used only at the top level of a declaration, but it is not clear whether this wording is sufficient to forbid usage like the following:
template <class T> struct A {};
template <class T> void f(A<T> x) {}
void g()
{
f(A<short>());
A<auto> x = A<short>();
}
Notes from the February, 2008 meeting:
It was agreed that the example should be ill-formed.
Proposed resolution (October, 2009):
Change 7.1.6.4 [dcl.spec.auto] paragraph 3 as follows:
...The auto shall appear as one of the decl-specifiers in the decl-specifier-seq and the decl-specifier-seq shall be followed by one or more init-declarators, each of which shall have a non-empty initializer.
7.1.6.4 [dcl.spec.auto] paragraph 6 says,
Once the type of a declarator-id has been determined according to 8.3 [dcl.meaning], the type of the declared variable using the declarator-id is determined from the type of its initializer using the rules for template argument deduction. Let T be the type that has been determined for a variable identifier d. Obtain P from T by replacing the occurrences of auto with either a new invented type template parameter U or, if the initializer is a braced-init-list (8.5.4 [dcl.init.list]), with std::initializer_list<U>. The type deduced for the variable d is then the deduced type determined using the rules of template argument deduction from a function call (14.9.2.1 [temp.deduct.call]), where P is a function template parameter type and the initializer for d is the corresponding argument.
The reference to “the deduced type” is unclear; it could be taken as referring either to the template parameter or to the function parameter type. 14.9.2.1 [temp.deduct.call] uses the term “deduced A,” and that usage should be repeated here.
Proposed resolution (October, 2009):
Change 7.1.6.4 [dcl.spec.auto] paragraph 6 as follows:
...The type deduced for the variable d is then the deduced type A determined using the rules of template argument deduction...
According to 7.3.4 [namespace.udir] paragraph 4,
The using-directive is transitive: if a scope contains a using-directive that nominates a second namespace that itself contains using-directives, the effect is as if the using-directives from the second namespace also appeared in the first.
This is true only for unqualified lookup; the algorithm in 3.4.3.2 [namespace.qual] paragraph 2 gives different results (the transitive closure terminates when a declaration of the name being looked up is found).
Proposed resolution (October, 2009):
Change 7.3.4 [namespace.udir] paragraph 4 as follows:
The For unqualified lookup (3.4.1 [basic.lookup.unqual]), the using-directive is transitive: if a scope contains a using-directive that nominates a second namespace that itself contains using-directives, the effect is as if the using-directives from the second namespace also appeared in the first. [Note: For qualified lookup, see 3.4.3.2 [namespace.qual]. —end note] [Example:...The terms “appertain” and “apply” are used in different ways in different subsections of 7.6 [dcl.attr]. A thorough editorial sweep of the entire section is needed to regularize their usage.
Proposed resolution (October, 2009):
Change 7.6.1 [dcl.attr.grammar] paragraph 4 as follows:
...If an attribute-specifier that appertains to some entity or statement contains an attribute that does not is not allowed to apply to that entity or statement, the program is ill-formed...
Change 7.6.2 [dcl.align] paragraph 1 as follows:
...The attribute can may be applied to a variable...
Change 7.6.3 [dcl.attr.noreturn] paragraph 1 as follows:
...The attribute applies may be applied to the declarator-id in a function declaration...
Change 7.6.4 [dcl.attr.final] paragraph 1 as follows:
...The attribute applies may be applied to class definitions...
Change 7.6.5 [dcl.attr.override] paragraph 1 as follows:
...The attribute applies may be applied to virtual member functions...
Change 7.6.5 [dcl.attr.override] paragraph 3 as follows:
...The attribute applies may be applied to class members...
Change 7.6.5 [dcl.attr.override] paragraph 5 as follows:
...The attribute applies may be applied to a class definition.
Change 7.6.6 [dcl.attr.depend] paragraph 1 as follows:
...The attribute applies may be applied to the declarator-id of a parameter-declaration... The attribute may also applies be applied to the declarator-id of a function declaration...
7.6.1 [dcl.attr.grammar] paragraph 3 specifies that keywords can be used as attribute-tokens. However, the alternative tokens in 2.6 [lex.digraph], such as bitor and compl, are not keywords. The text should be changed to make the alternative tokens acceptable as attribute-tokens as well.
Proposed resolution, October, 2009:
Change 7.6.1 [dcl.attr.grammar] paragraph 3 as follows:
...A If a keyword (2.12 [lex.key]) or an alternative token (2.6 [lex.digraph]) that satisfies the syntactic requirements of an identifier (2.11 [lex.name]) is contained in an attribute-token, it is considered an identifier...
According to 7.6.4 [dcl.attr.final] paragraph 1, the [[final]] attribute applied to a class is just a shorthand notation for marking each of the class's virtual functions as [[final]]. This is different from the similar usage in other languages, where it means that the class so marked cannot be used as a base class. This discrepancy is confusing, and the definition used by the other languages is more useful.
Notes from the March, 2009 meeting:
The intent of the [[final]] attribute is as an aid in optimization, to avoid virtual function calls when the final overrider is known. It is possible to use the [[final]] attribute to prevent derivation by marking the destructor as [[final]]; in fact, as most polymorphic classes will, as a matter of good programming practice, have a virtual destructor, marking the class as [[final]] will have the effect of preventing derivation.
Nonetheless, the general consensus of the CWG was to change the meaning of class [[final]] to parallel the usage in other languages.
Proposed resolution (October, 2009):
Change 7.6.4 [dcl.attr.final] paragraph 1 and add a new paragraph, as follows:
The attribute-token final specifies derivation semantics for a class and overriding semantics for a virtual function. It shall appear at most once in each attribute-list and no attribute-argument-clause shall be present. The attribute applies to class definitions and to virtual member functions being declared in a class definition. If the attribute is specified for a class definition, it is equivalent to being specified for each virtual member function of that class, including inherited member functions.
If some class B is marked final and a class D is derived from B the program is ill-formed.
Change the example in 7.6.4 [dcl.attr.final] paragraph 3 as follows:
struct B1 { virtual void f [[ final ]] (); }; struct D1 : B1 { void f(); // ill-formed }; struct [[ final ]] B2 { }; struct D2 : B2 { // ill-formed };
The current wording for the carries_dependency attribute does not limit it to value-returning functions (when applied to the declarator-id, indicating that the return value is affected), nor does it prohibit use in the declaration of a typedef or function pointer. Arguably these meaningless declarations should be prohibited.
Proposed resolution (October, 2009):
Change 7.6.6 [dcl.attr.depend] paragraph 1 as follows:
...The attribute applies to the declarator-id of a parameter-declaration in a function declaration or lambda, in which case it specifies that the initialization of the parameter carries a dependency to (1.10 [intro.multithread]) each lvalue-to-rvalue conversion (4.1 [conv.lval]) of that object. The attribute also applies to the declarator-id of a function declaration, in which case it specifies that the return value, if any, carries a dependency to the evaluation of the function call expression.
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.
It is not clear whether this applies to parameter packs or not. For example, is the following well-formed?
template <typename... T> void f(int i = 0, T ...args) { }
Note for comparison the corresponding wording in 14.2 [temp.param] paragraph 11 regarding template parameter packs:
If a template-parameter of a class template has a default template-argument, each subsequent template-parameter shall either have a default template-argument supplied or be a template parameter pack.
Proposed resolution (October, 2009):
Change 8.3.6 [dcl.fct.default] paragraph 4:
...In a given function declaration, all each parameters subsequent to a parameter with a default argument shall have a default arguments supplied in this or a previous declarations or shall be a function parameter pack. A default argument...
According to 8.4 [dcl.fct.def] paragraph 10,
A deleted definition of a function shall be the first declaration of the function.
The Standard is not currently clear about what the “first declaration” of an explicit specialization of a function template is. For example,
template<typename T> void f() { }
template<> void f<int>() = delete; // First declaration?
Proposed resolution (October, 2009):
A deleted definition of a function shall be the first declaration of the function or, for an explicit specialization of a function template, the first declaration of that specialization.
(This resolution also resolves issue 915.)
Notes from the October, 2009 meeting:
It was observed that this specification is complicated by the fact that the “first declaration” of a function might be in a block-extern declaration.
The only restriction placed on the use of “=default” in 8.4 [dcl.fct.def] paragraph 9 is that a defaulted function must be a special member function. However, there are many variations of declarations of special member functions, and it's not clear which of those should be able to be defaulted. Among the possibilities:
default arguments
by-value parameter for a copy assignment operator
exception specifications
arbitrary return values for copy assignment operators
a const reference parameter when the implicit function would have a non-const
Presumably, you should only be able to default a function if it is declared compatibly with the implicit declaration that would have been generated.
Proposed resolution (October, 2009):
Change 8.4 [dcl.fct.def] paragraph 9 as follows:
A function definition of the form:
decl-specifier-seqopt attribute-specifieropt declarator = default ;
is called an explicitly-defaulted definition. Only special member functions may be explicitly defaulted, and the implementation shall define them as if they had implicit definitions (12.1 [class.ctor], 12.4 [class.dtor], 12.8 [class.copy]). A function that is explicitly defaulted shall
be a special member function,
have the same declared function type (except for possibly-differing ref-qualifiers and except that in the case of a copy constructor or copy assignment operator, the parameter type may be “reference to non-const T,” where T is the name of the member function's class) as if it had been implicitly declared,
not have default arguments, and
not have an exception-specification.
[Note: This implies that parameter types, return type, and cv-qualifiers must match the hypothetical implicit declaration. —end note] An explicitly-defaulted function may be declared constexpr only if it would have been implicitly declared as constexpr. If it is explicitly defaulted on its first declaration,
it shall be public,
it shall not be explicit,
it shall not be virtual,
it is implicitly considered to have the same exception-specification as if it had been implicitly declared (15.4 [except.spec]), and
in the case of a copy constructor or copy assignment operator, it shall have the same parameter type as if it had been implicitly declared.
[Note: Such a special member function may be trivial, and thus its accessibility and explicitness should match the hypothetical implicit definition; see below. —end note] [Example:
struct S { S(int a = 0) = default; // ill-formed: default argument void operator=(const S&) = default; // ill-formed: non-matching return type ~S() throw() = default; // ill-formed: exception-specification private: S(S&); // OK: private copy constructor }; S::S(S&) = default; // OK: defines copy constructor—end example] Explicitly-defaulted functions and implicitly-declared functions are collectively called defaulted functions, and the implementation shall provide implicit definitions for them (12.1 [class.ctor], 12.4 [class.dtor], 12.8 [class.copy]), which might mean defining them as deleted.A special member function that would be implicitly defined as deleted may be explicitly defaulted only on its first declaration, in which case it is defined as deleted. A special member function is user-provided if it is user-declared and not explicitly defaulted on its first declaration. A user-provided explicitly-defaulted function is defined at the point where it is explicitly defaulted. [Note:...
[Editorial note: this change incorporates the overlapping portion of the resolution of issue 667.]
Change 12.1 [class.ctor] paragraph 6 as follows:
This resolution also resolves issue 905. See also issue 667.
The final set of declarations in the example following 8.5.4 [dcl.init.list] paragraph 3 bullet 3 is:
struct S2 {
int m1;
double m2,m3;
};
S2 s21 = { 1, 2, 3.0 }; // OK
S2 s22 { 1.0, 2, 3 }; // error: narrowing
S2 s23 {}; // OK: default to 0,0,0
However, S2 is an aggregate. Aggregates are handled in bullet 1, while bullet 3 deals with classes with constructors. This part of the example should be moved to the first bullet.
Proposed resolution (October, 2009):
Move the S2 example from bullet 3 to bullet 1 in 8.5.4 [dcl.init.list] paragraph 3:
If T is an aggregate, aggregate initialization is performed (8.5.1 [dcl.init.aggr]).
[Example:
double ad[] = { 1, 2.0 }; // OK
int ai[] = { 1, 2.0 }; // error: narrowing
struct S2 {
int m1;
double m2,m3;
};
S2 s21 = { 1, 2, 3.0 }; // OK
S2 s22 { 1.0, 2, 3 }; // error: narrowing
S2 s23 {}; // OK: default to 0,0,0
—end example]
Otherwise, if T is a specialization...
Otherwise, if T is a class type...
[Example:
...
S s3 { }; // OK: invoke #2
struct S2 {
int m1;
double m2,m3;
};
S2 s21 = { 1, 2, 3.0 }; // OK
S2 s22 { 1.0, 2, 3 }; // error: narrowing
S2 s23 {}; // OK: default to 0,0,0
—end example]
...
It is presumably possible to declare a defaulted copy constructor to be explicit. Should that render a class not trivially copyable, even though the copy constructor is trivial? That is, does being “trivally copyable” mean that copy initialization, and not just direct initialization, is possible?
A related question is whether the specification of triviality should require that the copy constructor and copy assignment operator must be public. (With the advent of “=default” it is possible to make them non-public, which was not the case when these definitions were crafted.)
Proposed resolution (October, 2009):
This issues is resolved by the resolution of issue 906.
(From message 14555.)
The reasons for which an implicitly-declared default constructor is defined as deleted, given in 12.1 [class.ctor] paragraph 4, all deal with cases in which a member cannot be default-initialized. Presumably a brace-or-equal-initializer for such a member would eliminate the need to define the constructor as deleted, but this case is not addressed by the current wording.
Proposed resolution (October, 2009):
Change 12.1 [class.ctor] paragraph 5, the second list, as follows:
An implicitly-declared default constructor for class X is defined as deleted if:
X is a union-like class that has a variant member with a non-trivial default constructor,
any non-static data member with no brace-or-equal-initializer is of reference type,
any non-static data member of const-qualified type (or array thereof) with no brace-or-equal-initializer does not have a user-provided default constructor, or
any direct or virtual base class, or non-static data member with no brace-or-qual-initializer, or direct or virtual base class has class type M (or array thereof) and either M has no default constructor, or if constructor or overload resolution (13.3 [over.match]) as applied to M's default constructor, results in an ambiguity or in a function that is deleted or inaccessible from the implicitly-declared default constructor.
Consider the following example:
struct A {
A() {
std::thread(&A::Func, this).detach();
}
virtual void Func() {
printf("In A");
}
};
struct B : public A {
virtual void Func() {
printf("In B");
}
};
struct C : public B {
virtual void Func() {
printf("In C");
}
};
C c;
What is the program allowed to print? Should it be undefined behavior or merely unspecified which of the Func()s is called?
There is a related question about which variables C::Func() can depend on having been constructed. Unless we want to require the equivalent of at least memory_order_consume on the presumed virtual function table pointer, I think the answer is just the members of A.
If I instead just have
A a;
I think the only reasonable behavior is to print In A.
Finally, given
struct F {
F() {
std::thread(&F::Func, this).detach();
}
virtual void Func() {
print("In F");
}
};
struct G : public F {
};
G g;
I can see the behavior being undefined, but I think a lot of people would be confused if it did anything other than print In F.
Suggested resolution:
I think the intent here is that an object should not be used in another thread until any non-trivial constructor has been called. One possible way of saying that would be to add a new paragraph at the end of 12.7 [class.cdtor]:
A constructor for a class with virtual functions or virtual base classes modifies a memory location in the object that is accessed by any access to a virtual function or virtual base class or by a dynamic_cast. [Note: This implies that access to an object by another thread while it is being constructed often introduces a data race (see 1.10 [intro.multithread]). —end note]
Proposed resolution (October, 2009):
Add the following as a new paragraph at the end of 3.8 [basic.life]:
In this section, “before” and “after” refer to the “happens before” relation (1.10 [intro.multithread]). [Note: Therefore, undefined behavior results if an object that is being constructed in one thread is referenced from a different thread without adequate synchronization. —end note]
Should the following class have a trivial copy assignment operator?
struct A {
int& m;
A();
A(const A&);
};
12.8 [class.copy] paragraph 11 does not mention whether the presence of reference members (or cv-qualifiers, etc.) should affect triviality. Should it?
One reason why this matters is that implementations have to make the builtin type trait operator __has_trivial_default_ctor(T) work so that they can support the type trait template std::has_trivial_default_constructor.
Assuming the answer is “yes,” it looks like we probably need similar wording for trivial default and trivial copy ctors.
Notes from the February, 2008 meeting:
Deleted special member functions are also not trivial. Resolution of this issue should be coordinated with the concepts proposal.
Notes from the June, 2008 meeting:
It appears that this issue will be resolved by the concepts proposal directly. The issue is in “review” status to check if that is indeed the case in the final version of the proposal.
Additional notes (May, 2009):
Consider the following example:
struct Base {
private:
~Base() = default;
};
struct Derived: Base {
};
The implicitly-declared destructor of Derived is defined as deleted because Base::~Base() is inaccessible, but it fulfills the requirements for being trivial. Presumably the Base destructor should be non-trivial, either by directly specifying that it is non-trivial or by specifying that it is user-provided. An alternative would be to make it ill-formed to attempt to declare a defaulted non-public special member function.
Any changes to the definition of triviality should be checked against 9 [class] paragraph 6 for any changes needed there to accommodate the new definitions.
Notes from the July, 2009 meeting:
The July, 2009 resolution of issue 906 addresses the example above (with an inaccessible defaulted destructor): a defaulted special member function can only have non-public access if the defaulted definition is outside the class, making it non-trivial. The example as written above would be ill-formed.
Proposed resolution (October, 2009):
Change 8.4 [dcl.fct.def] paragraph 9 as follows:
...Only special member functions may be explicitly defaulted. Explicitly-defaulted functions and implicitly-declared functions are collectively called defaulted functions, and the implementation shall define them as if they had provide implicit definitions for them (12.1 [class.ctor], 12.4 [class.dtor], 12.8 [class.copy]), which might mean defining them as deleted. A special member function that would be implicitly defined as deleted may be explicitly defaulted only on its first declaration, in which case it is defined as deleted. A special member function is user-provided if it is user-declared and not explicitly defaulted on its first declaration. A user-provided explicitly-defaulted function is defined at the point where it is explicitly defaulted. [Note:...
Change 12.1 [class.ctor] paragraphs 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 constructor for class X, a constructor having no parameters is implicitly declared as defaulted (8.4 [dcl.fct.def]). An implicitly-declared default constructor is an inline public member of its class. A default constructor is trivial if it is not user-provided (8.4 [dcl.fct.def]) and if:
its class has no virtual functions (10.3 [class.virtual]) and no virtual base classes (10.1 [class.mi]), and
no non-static data member of its class has a brace-or-equal-initializer, and
all the direct base classes of its class have trivial default constructors, and
for all the non-static data members of its class that are of class type (or array thereof), each such class has a trivial default constructor.
An implicitly-declared defaulted default constructor for class X is defined as deleted if:
X is a union-like class that has a variant member with a non-trivial default constructor,
any non-static data member is of reference type,
any non-static data member of const-qualified type (or array thereof) does not have a user-provided default constructor, or
any non-static data member or direct or virtual base class has class type M (or array thereof) and M has no default constructor, or if overload resolution (13.3 [over.match]) as applied to M's default constructor, results in an ambiguity or a function that is deleted or inaccessible from the implicitly-declared default constructor.
A default constructor is trivial if it is neither user-provided nor deleted and if:
its class has no virtual functions (10.3 [class.virtual]) and no virtual base classes (10.1 [class.mi]), and
no non-static data member of its class has a brace-or-equal-initializer, and
all the direct base classes of its class have trivial default constructors, and
for all the non-static 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.
A non-user-provided default constructor for a class that is defaulted and not deleted is implicitly defined when it is used (3.2 [basic.def.odr]) to create an object of its class type (1.8 [intro.object]), or when it is explicitly defaulted after its first declaration. The implicitly-defined or explicitly-defaulted default constructor performs the set of initializations of the class that would be performed by a user-written default constructor for that class with no ctor-initializer (12.6.2 [class.base.init]) and an empty compound-statement. If that user-written default constructor would be ill-formed, the program is ill-formed. If that user-written default constructor would satisfy the requirements of a constexpr constructor (7.1.5 [dcl.constexpr]), the implicitly-defined default constructor is constexpr. Before the non-user-provided defaulted default constructor for a class is implicitly defined, all the non-user-provided default constructors for its base classes and its non-static data members shall have been implicitly defined. [Note: an implicitly-declared default constructor has an exception-specification (15.4 [except.spec]). An explicitly-defaulted definition has no implicit exception-specification. —end note]
Change 12.4 [class.dtor] paragraphs 3-4 as follows:
If a class has no user-declared destructor, a destructor is declared implicitly declared as defaulted (8.4 [dcl.fct.def]). An implicitly-declared destructor is an inline public member of its class. If the class is a union-like class that has a variant member with a non-trivial destructor, an implicitly-declared destructor is defined as deleted (8.4 [dcl.fct.def]). A destructor is trivial if it is not user-provided and if:
the destructor is not virtual,
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.
An implicitly-declared defaulted destructor for a class X is defined as deleted if:
X is a union-like class that has a variant member with a non-trivial destructor,
any of the non-static data members has class type M (or array thereof) and M has an a deleted destructor or a destructor that is inaccessible from the implicitly-declared destructor, or
any direct or virtual base class has a deleted destructor or a destructor that is inaccessible from the implicitly-declared destructor.
A destructor is trivial if it is neither user-provided nor deleted and if:
the destructor is not virtual,
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.
A non-user-provided destructor that is defaulted and not defined as deleted is implicitly defined when it is used to destroy an object of its class type (3.7 [basic.stc]), or when it is explicitly defaulted after its first declaration. A program is ill-formed if the class for which a destructor is implicitly defined or explicitly defaulted has:
a non-static data member of class type (or array thereof) with an inaccessible destructor, or
a base class with an inaccessible destructor.
Before the non-user-provided defaulted destructor for a class is implicitly defined, all the non-user-defined non-user-provided destructors for its base classes and its non-static data members shall have been implicitly defined. [Note: an implicitly-declared destructor has an exception-specification (15.4 [except.spec]). An explictly defaulted definition has no implicit exception-specification. —end note]
Change 12.8 [class.copy] paragraphs 4-9 as follows:
If the class definition does not explicitly declare a copy constructor, one is declared implicitly implicitly declared as defaulted (8.4 [dcl.fct.def]). Thus...
...An implicitly-declared copy constructor is an inline public member of its class. An implicitly-declared defaulted copy constructor for a class X is defined as deleted if X has: ...
A copy constructor for class X is trivial trivial if it is not neither user-provided nor deleted (8.4 [dcl.fct.def]) and if...
A non-user-provided copy constructor that is defaulted and not defined as deleted is implicitly defined if it is used to initialize an object of its class type from a copy of an object of its class type or of a class type derived from its class type116, or when it is explicitly defaulted after its first declaration. [Note: the copy constructor is implicitly defined even if the implementation elided its use (12.2 [class.temporary]). —end note]
Before the non-user-provided defaulted copy constructor for a class is implicitly defined, all non-user-provided copy constructors...
The implicitly-defined or explicitly-defaulted copy constructor for a non-union class X performs...
The implicitly-defined or explicitly-defaulted copy constructor for a union X copies the object representation (3.9 [basic.types]) of X.
Change 12.8 [class.copy] paragraphs 11-15 as follows:
If the class definition does not explicitly declare a copy assignment operator, one is declared implicitly implicitly declared as defaulted (8.4 [dcl.fct.def])...
...An implicitly-declared defaulted copy assignment operator for class X is defined as deleted if X has:...
A copy assignment operator for class X is trivial if it is not neither user-provided nor deleted and if...
A non-user-provided copy assignment operator that is defaulted and not defined as deleted is implicitly defined when an object of its class type is assigned a value of its class type or a value of a class type derived from its class type, or when it is explicitly defaulted after its first declaration.
Before the non-user-provided defaulted copy assignment operator for a class is implicitly defined...
The implicitly-defined or explicitly-defaulted copy assignment operator for a non-union class X performs...
It is unspecified whether subobjects representing virtual base classes are assigned more than once by the implicitly-defined or explicitly-defaulted copy assignment operator. [Example:...
The implicitly-defined or explicitly-defaulted copy assignment operator for a union X copies the object representation (3.9 [basic.types]) of X.
12.8 [class.copy] paragraph 16 details the conditions under which a thrown object can be moved instead of copied. However, the optimization as currently described is unsafe. Consider the following example:
void f() {
X x;
try {
throw x;
} catch (...) {
}
// x may have been moved from but can still be accessed here
}
When the operation is a throw, as opposed to a return, there must be a restriction that the object potentially being moved be defined within the innermost enclosing try block.
Notes from the July, 2009 meeting:
It is not clear how important this optimization is in the context of throw: how often is a large object with substantial copying overhead thrown? Also, throwing an exception is already a heavyweight operation, so presumably moving instead of copying an object would not make much difference.
Proposed resolution (October, 2009):
Change 12.8 [class.copy] paragraph 17 second bullet as follows:
Consider the following example:
struct C { };
struct A {
explicit operator int() const;
explicit operator C() const;
};
struct B {
int i;
B(const A& a): i(a) { }
};
int main() {
A a;
int i = a;
int j(a);
C c = a;
C c2(a);
}
It's clear that the B constructor and the declaration of j are well-formed and the declarations of i and c are ill-formed. But what about the declaration of c2? This is supposed to work, but it doesn't under the current wording.
C c2(a) is direct-initialization of a class, so constructors are considered. The only possible candidate is the default copy constructor. So we look for a conversion from A to const C&. There is a conversion operator to C, but it is explicit and we are now performing copy-initialization of a reference temporary, so it is not a candidate, and the declaration of c2 is ill-formed.
Proposed resolution (October, 2009):
Change 13.3.1.4 [over.match.copy] paragraph 1 second bullet as follows:
13.3.3.1 [over.best.ics] paragraph 4 says,
However, when considering the argument of a user-defined conversion function that is a candidate by 13.3.1.3 [over.match.ctor] when invoked for the copying of the temporary in the second step of a class copy-initialization, by 13.3.1.7 [over.match.list] when passing the initializer list as a single argument or when the initializer list has exactly one element and a conversion to some class X or reference to (possibly cv-qualified) X is considered for the first parameter of a constructor of X, or by 13.3.1.4 [over.match.copy], 13.3.1.5 [over.match.conv], or 13.3.1.6 [over.match.ref] in all cases, only standard conversion sequences and ellipsis conversion sequences are allowed.
This is not quite right, as this applies to constructor arguments, not just arguments of user-defined conversion functions. Furthermore, the word “allowed” might be better replaced by something like,
considered (in particular, for the purposes of determining whether the candidate function (that is either a constructor or a conversion function) is viable)
Proposed resolution (October, 2009):
Change 13.3.3.1 [over.best.ics] paragraph 4 as follows:
However, when considering the argument of a constructor or user-defined conversion function that is a candidate by 13.3.1.3 [over.match.ctor] when invoked for the copying of the temporary in the second step of a class copy-initialization, by 13.3.1.7 [over.match.list] when passing the initializer list as a single argument or when the initializer list has exactly one element and a conversion to some class X or reference to (possibly cv-qualified) X is considered for the first parameter of a constructor of X, or by 13.3.1.4 [over.match.copy], 13.3.1.5 [over.match.conv], or 13.3.1.6 [over.match.ref] in all cases, only standard conversion sequences and ellipsis conversion sequences are allowed considered.
Conversion of a pointer or pointer to member to bool is given special treatment as a tiebreaker in overload resolution in 13.3.3.2 [over.ics.rank] paragraph 4, bullet 1:
It would be reasonable to expect a similar provision to apply to conversions of std::nullptr_t to bool.
Proposed resolution (October, 2009):
Change 13.3.3.2 [over.ics.rank] paragraph 4 bullet 1 as follows:
13.6 [over.built] paragraphs 24-25 describe the imaginary built-in conditional operator functions. However, neither paragraph 24 (promoted arithmetic types) nor 25 (pointer and pointer-to-member types) covers scoped enumerations, whose values should be usable in conditional expressions.
(See also issue 835.)
Proposed resolution (October, 2009):
Change 13.6 [over.built] paragraph 25 as follows:
For every type T, where T is a pointer, or pointer-to-member, or scoped enumeration type, there exist candidate operator functions of the form
T operator?(bool, T , T );
5.19 [expr.const] permits literal types with a constexpr conversion function to an integral type to be used in an integral constant expression. However, such conversions are not listed in 14.4.2 [temp.arg.nontype] paragraph 5 bullet 1 among the conversions applied to template-arguments for a non-type template-parameter of integral or enumeration type.
Notes from the March, 2009 meeting:
The original national body comment suggested allowing any literal type as a non-type template argument. The CWG was not in favor of this change, but in the course of discussing the suggestion discovered the problem with template-parameters of integral and enumeration type.
Proposed resolution (October, 2009):
Change 14.4.2 [temp.arg.nontype] paragraph 1 bullet 1 as follows:
Is this code well-formed?
template <typename T> struct A {
struct B;
};
class C {
template <typename T> friend struct A<T>::B;
static int bar;
};
template <> struct A<char> {
struct B {
int f() {
return C::bar; // Is A<char>::B a friend of C?
}
};
};
According to 14.6.4 [temp.friend] paragraph 5,
A member of a class template may be declared to be a friend of a non-template class. In this case, the corresponding member of every specialization of the class template is a friend of the class granting friendship.
This would tend to indicate that the example is well-formed. However, technically A<char>::B does not “correspond to” the same-named member of the class template: 14.8.3 [temp.expl.spec] paragraph 4 says,
The definition of an explicitly specialized class is unrelated to the definition of a generated specialization. That is, its members need not have the same names, types, etc. as the members of a generated specialization.
In other words, there are no “corresponding members” in an explicit specialization.
Is this the outcome we want for examples like the preceding? There is diversity among implementations on this question, with some accepting the example and others rejecting it as an access violation.
Notes from the July, 2009 meeting:
The consensus of the CWG was to allow the correspondence of similar members in explicit specializations.
Proposed resolution (October, 2009):
Change 14.6.4 [temp.friend] paragraph 5 as follows:
A member of a class template may be declared to be a friend of a non-template class. In this case, the corresponding member of every specialization of the class template is a friend of the class granting friendship. For explicit specializations the corresponding member is the member (if any) that has the same name, kind (type, function, class template or function template), template parameters, and signature as the member of the class template instantiation that would otherwise have been generated. [Example:
template<class T> struct A { struct B { }; void f(); struct D { void g(); }; }; template<> struct A<int> { struct B { }; int f(); struct D { void g(); }; }; class C { template<class T> friend struct A<T>::B; // grants friendship to A<int>::B even though // it is not a specialization of A<T>::B template<class T> friend void A<T>::f(); // does not grant friendship to A<int>::f() // because its return type does not match template<class T> friend void A<T>::D::g(); // does not grant friendship to A<int>::D::g() // because A<int>::D is not a specialization of A<T>::D };
14.7.2.2 [temp.dep.expr] paragraph 3 says,
An id-expression is type-dependent if it contains:
- an identifier that was declared with a dependent type...
This treatment seems inadequate with regard to id-expressions in function calls:
According to 14.7.2.1 [temp.dep.type] paragraph 6,
A type is dependent if it is
- ...
- a compound type constructed from any dependent type...
This would apply to the type of a member function of a class template if any of its parameters are dependent, even if the return type is not dependent. However, there is no need for a call to such a function to be a type-dependent expression because the type of the expression is known at definition time.
This wording does not handle the case of overloaded functions, some of which might have dependent types (however defined) and others not.
Notes from the October, 2009 meeting:
The consensus of the CWG was that the first point of the issue is not sufficiently problematic as to require a change.
Proposed resolution (October, 2009):
Change 14.7.2.2 [temp.dep.expr] paragraph 3 as follows:
An id-expression is type-dependent if it contains:
an identifier that was associated by name lookup with one or more declarations declared with a dependent type,
a template-id that is dependent,
a conversion-function-id that specifies a dependent type, or
a nested-name-specifier or a qualified-id that names a member of an unknown specialization.
According to 14.7.4.2 [temp.dep.candidate],
For a function call that depends on a template parameter, if the function name is an unqualified-id but not a template-id, the candidate functions are found using the usual lookup rules (3.4.1 [basic.lookup.unqual], 3.4.2 [basic.lookup.argdep]) except that:
For the part of the lookup using unqualified name lookup (3.4.1 [basic.lookup.unqual]), only function declarations with external linkage from the template definition context are found.
For the part of the lookup using associated namespaces (3.4.2 [basic.lookup.argdep]), only function declarations with external linkage found in either the template definition context or the template instantiation context are found.
It is not at all clear why a call using a template-id would be treated differently from one not using a template-id. Furthermore, is it really necessary to exclude internal linkage functions from the lookup? Doesn't the ODR give implementations sufficient latitude to handle this case without another wrinkle on name lookup?
(See also issue 524.)
Notes from the April, 2006 meeting:
The consensus of the group was that template-ids should not be treated differently from unqualified-ids (although it's not clear how argument-dependent lookup works for template-ids), and that internal-linkage functions should be found by the lookup (although they may result in errors if selected by overload resolution).
Note (June, 2006):
Although the notes from the Berlin meeting indicate that argument-dependent lookup for template-ids is under-specified in the Standard, further examination indicates that that is not the case: the note in 14.9.1 [temp.arg.explicit] paragraph 8 clearly indicates that argument-dependent lookup is to be performed for template-ids, and 3.4.2 [basic.lookup.argdep] paragraph 4 describes the lookup performed:
When considering an associated namespace, the lookup is the same as the lookup performed when the associated namespace is used as a qualifier (3.4.3.2 [namespace.qual]) except that:
Any using-directives in the associated namespace are ignored.
Any namespace-scope friend functions declared in associated classes are visible within their respective namespaces even if they are not visible during an ordinary lookup (11.4 [class.friend]).
Proposed resolution (October, 2009):
Change 14.7.2 [temp.dep] paragraph 1 as follows:
In an expression of the form:
postfix-expression ( expression-listopt )
where the postfix-expression is an unqualified-id but not a template-id, the unqualified-id denotes a dependent name if and only if any of the expressions in the expression-list is a type-dependent expression (14.7.2.2 [temp.dep.expr])...
Change 14.7.4.2 [temp.dep.candidate] paragraph 1 as follows:
For a function call that depends on a template parameter, if the function name is an unqualified-id but not a template-id, or if the function is called using operator notation, the candidate functions are found using the usual lookup rules (3.4.1 [basic.lookup.unqual], 3.4.2 [basic.lookup.argdep]) except that:
For the part of the lookup using unqualified name lookup (3.4.1 [basic.lookup.unqual]), only function declarations with external linkage from the template definition context are found.
For the part of the lookup using associated namespaces (3.4.2 [basic.lookup.argdep]), only function declarations with external linkage found in either the template definition context or the template instantiation context are found.
Consider this example:
template <class T> struct A {
virtual void f() {}
};
extern template struct A<int>;
int main() {
A<int> a;
a.f();
}
The intent is that the explicit instantiation declaration will suppress any compiler-generated machinery such as a virtual function table or typeinfo data in this translation unit, and that because of 14.8.2 [temp.explicit] paragraph 10,
An entity that is the subject of an explicit instantiation declaration and that is also used in the translation unit shall be the subject of an explicit instantiation definition somewhere in the program; otherwise the program is ill-formed, no diagnostic required.
the use of A<int> in declaring a requires an explicit instantiation definition in another translation unit that will provide the requisite compiler-generated data.
The existing wording of 14.8.2 [temp.explicit] does not express this intent clearly enough, however.
Suggested resolution:
Change 14.8.2 [temp.explicit] paragraph 7 as follows:
An explicit instantiation that names a class template specialization is also an explicit instantion of the same kind (declaration or definition) of each of its members (not including members inherited from base classes) that has not been previously explicitly specialized in the translation unit containing the explicit instantiation, except as described below.
Change 14.8.2 [temp.explicit] paragraph 9 as follows:
An explicit instantiation declaration that names a class template specialization has no effect on the class template specialization itself (except for perhaps resulting in its implicit instantiation). Except for inline functions and class template specializations, other explicit instantiation declarations have the effect of suppressing the implicit instantiation of the entity to which they refer...
Proposed resolution (October, 2009):
Change 14.8.2 [temp.explicit] paragraphs 7-9 as follows:
An explicit instantiation that names a class template specialization is also an explicit instantion of the same kind (declaration or definition) of each of its members (not including members inherited from base classes) that has not been previously explicitly specialized in the translation unit containing the explicit instantiation, except as described below. [Note: In addition, it will typically be an explicit instantiation of certain implementation-dependent data about the class. —end note]
An explicit instantiation definition that names a class template specialization explicitly instantiates the class template specialization and is only an explicit instantiation definition of only those members whose definition is visible at the point of instantiation.
An explicit instantiation declaration that names a class template specialization has no effect on the class template specialization itself (except for perhaps resulting in its implicit instantiation). Except for inline functions and class template specializations, other explicit instantiation declarations have the effect of suppressing the implicit instantiation of the entity to which they refer...
The adoption of paper N2844 made it ill-formed to attempt to bind an rvalue reference to an lvalue. However, the example in 14.9.2.1 [temp.deduct.call] paragraph 3 still reflects the previous specification:
template <typename T> int f(T&&);
int i;
int j = f(i); // calls f<int&>(i)
template <typename T> int g(const T&&);
int k;
int n = g(k); // calls g<int>(k)
The last line of that example is now ill-formed, attempting to bind the const int&& parameter of g to the lvalue k.
Proposed resolution (July, 2009):
Replace the example in 14.9.2.1 [temp.deduct.call] paragraph 3 with:
template<typename T> int f(T&&);
template<typename T> int g(const T&&);
int i;
int n1 = f(i); // calls f<int&>(int&)
int n2 = f(0); // calls f<int>(int&&)
int n3 = g(i); // error: would call g<int>(const int&&), which would
// bind an rvalue reference to an lvalue
(See also issue 858.)
15.1 [except.throw] paragraph 4 says,
When the last remaining active handler for the exception exits by any means other than throw; the temporary object is destroyed and the implementation may deallocate the memory for the temporary object...
With std::current_exception() (18.8.5 [propagation] paragraph 7), it might be possible to refer to the exception object after its last handler exits (if the exception object is not copied). The text needs to be updated to allow for that possibility.
Proposed resolution (September, 2009):
Change 15.1 [except.throw] paragraph 4 as follows:
The memory for the temporary copy of the exception being thrown exception object is allocated in an unspecified way, except as noted in 3.7.4.1 [basic.stc.dynamic.allocation]. The temporary persists as long as there is a handler being executed for that exception. In particular, if If a handler exits by executing a throw; statement, that passes control rethrowing, control is passed to another handler for the same exception, so the temporary remains. The exception object is destroyed after either When the last remaining active handler for the exception exits by any means other than throw; rethrowing, or the last object of type std::exception_ptr (18.8.5 [propagation]) that refers to the exception object is destroyed, whichever is later. In the former case, the destruction occurs when the handler exits, immediately after the destruction of the object declared in the exception-declaration in the handler, if any. In the latter case, the destruction occurs before the destructor of std::exception_ptr returns. the temporary object is destroyed and the The implementation may then deallocate the memory for the temporary exception object; any such deallocation is done in an unspecified way. The destruction occurs immediately after the destruction of the object declared in the exception-declaration in the handler.
There are a number of specifications in the Standard that should also apply to references. For example:
3 [basic] paragraphs 3-4 indicate that a reference cannot have a name because it is not an entity. (See also issue 485.)
3.4.1 [basic.lookup.unqual] paragraph 13 covers unqualified lookup in the initializer of a variable member of a namespace but not that of a reference member of a namespace. It would be very strange if the lookup in these two cases were different.
3.5 [basic.link] paragraph 8 prohibits use of a type without linkage as the type of a variable with linkage, but not as the type of a reference with linkage. (References with linkage are explicitly mentioned earlier in the section.)
3.7.1 [basic.stc.static] paragraph 3 permits local static variables but not local static references.
A number of other examples could be cited. A thorough review is needed to make sure that references are completely specified.
Notes from the September, 2008 meeting:
The CWG expressed interest in an approach that would define “variable” to include both objects and references and to use that for both this issue and issue 570.
Proposed resolution (October, 2009):
See paper PL22.16/09-0183 = WG21 N2993. This resolution also resolves issue 570.
3.2 [basic.def.odr] paragraph 1 says,
No translation unit shall contain more than one definition of any variable, function, class type, enumeration type or template.
This says nothing about references. Is it permitted to define a reference more than once in a single translation unit? (The list in paragraph 5 of things that can have definitions in multiple translation units does not include references.)
Notes from the September, 2008 meeting:
The CWG expressed interest in an approach that would define “variable” to include both objects and references and to use that for both this issue and issue 633.
Proposed resolution (October, 2009):
This issue is resolved by the resolution of issue 633.
Trigraphs are a complicated solution to an old problem, that cause more problems than they solve in the modern environment. Unexpected trigraphs in string literals and occasionally in comments can be very confusing for the non-expert. They should be deprecated.
Notes from the March, 2009 meeting:
IBM, at least, uses trigraphs in its header files in conditional compilation directives to select character-set dependent content in a character-set independent fashion and would thus be negatively affected by the removal of trigraphs. One possibility that was discussed was to avoid expanding trigraphs inside character string literals, which is the context that causes most surprise and confusion, but still to support them in the rest of the program text. Specifying that approach, however, would be challenging because trigraphs are replaced in phase 1, before character strings are recognized in phase 3. See also the similar discussion of universal-character-names in issue 787.
The consensus of the CWG was that trigraphs should be deprecated.
Proposed resolution (September, 2009):
See paper PL22.16/09-0168 = WG21 N2978.
Notes from the October, 2009 meeting:
The CWG is interested in exploring other alternatives that address the particular problem of trigraphs in raw strings but that do not require the grammar changes of the approach in N2978. One possibility might be to recognize raw strings in some way in translation phase 1.
The various uses of the term “declarative region” throughout the Standard indicate that the term is intended to refer to the entire block, class, or namespace that contains a given declaration. For example, 3.3 [basic.scope] paragraph 2 says, in part:
[Example: in
int j = 24; int main() { int i = j, j; j = 42; }The declarative region of the first j includes the entire example... The declarative region of the second declaration of j (the j immediately before the semicolon) includes all the text between { and }...
However, the actual definition given for “declarative region” in 3.3 [basic.scope] paragraph 1 does not match this usage:
Every name is introduced in some portion of program text called a declarative region, which is the largest part of the program in which that name is valid, that is, in which that name may be used as an unqualified name to refer to the same entity.
Because (except in class scope) a name cannot be used before it is declared, this definition contradicts the statement in the example and many other uses of the term throughout the Standard. As it stands, this definition is identical to that of the scope of a name.
The term “scope” is also misused. The scope of a declaration is defined in 3.3 [basic.scope] paragraph 1 as the region in which the name being declared is valid. However, there is frequent use of the phrase “the scope of a class,” not referring to the region in which the class's name is valid but to the declarative region of the class body, and similarly for namespaces, functions, exception handlers, etc. There is even a mention of looking up a name “in the scope of the complete postfix-expression” (3.4.5 [basic.lookup.classref] paragraph 3), which is the exact inverse of the scope of a declaration.
This terminology needs a thorough review to make it logically consistent. (Perhaps a discussion of the scope of template parameters could also be added to section 3.3 [basic.scope] at the same time, as all other kinds of scopes are described there.)
Proposed resolution (November, 2006):
Change 3.3 [basic.scope] paragraph 1 as follows:
Every name is introduced in some portion of program text called a declarative region, which is the largest part of the program in which that name is valid, that is, in which that name may be used as an unqualified name to refer to the same entity a statement, block, function declarator, function-definition, class, handler, template-declaration, template-parameter-list of a template template-parameter, or namespace. In general, each particular name is valid may be used as an unqualified name to refer to the entity of its declaration or to the label only within some possibly discontiguous portion of program text called its scope. To determine the scope of a declaration...
Change 3.3 [basic.scope] paragraph 3 as follows:
The names declared by a declaration are introduced into the scope in which the declaration occurs declarative region that directly encloses the declaration, except that declaration-statements, function parameter names in the declarator of a function-definition, exception-declarations (3.3.3 [basic.scope.local]), the presence of a friend specifier (11.4 [class.friend]), certain uses of the elaborated-type-specifier (7.1.6.3 [dcl.type.elab]), and using-directives (7.3.4 [namespace.udir]) alter this general behavior.
Change 3.3.3 [basic.scope.local] paragraphs 1-3 and add a new paragraph 4 before the existing paragraph 4 as follows:
A name declared in a block (6.3 [stmt.block]) is local to that block. Its potential scope begins at its point of declaration (3.3.2 [basic.scope.pdecl]) and ends at the end of its declarative region. The declarative region of a name declared in a declaration-statement is the directly enclosing block (6.3 [stmt.block]). Such a name is local to the block.
The potential scope declarative region of a function parameter name (including one appearing in the declarator of a function-definition or in a lambda-parameter-declaration-clause) or of a function-local predefined variable in a function definition (8.4 [dcl.fct.def]) begins at its point of declaration. If the function has a function-try-block the potential scope of a parameter or of a function-local predefined variable ends at the end of the last associated handler, otherwise it ends at the end of the outermost block of the function definition. A parameter name is the entire function definition or lambda-expression. Such a name is local to the function definition and shall not be redeclared in the any outermost block of the function definition nor in the outermost block of any handler associated with a function-try-block function-body (including handlers of a function-try-block) or lambda-expression.
The name in a catch exception-declaration The declarative region of a name declared in an exception-declaration is its entire handler. Such a name is local to the handler and shall not be redeclared in the outermost block of the handler.
The potential scope of any local name begins at its point of declaration (3.3.2 [basic.scope.pdecl]) and ends at the end of its declarative region.
Change 3.3.5 [basic.funscope] as indicated:
Labels (6.1 [stmt.label]) have function scope and may be used anywhere in the function in which they are declared except in members of local classes (9.8 [class.local]) of that function. Only labels have function scope.
Change 6.7 [stmt.dcl] paragraph 1 as follows:
A declaration statement introduces one or more new identifiers names into a block; it has the form
declaration-statement:
block-declaration
[Note: If an identifier a name introduced by a declaration was previously declared in an outer block, the outer declaration is hidden for the remainder of the block, after which it resumes its force (3.3.11 [basic.scope.hiding]). —end note]
[Drafting notes: This resolution deals almost exclusively with the unclear definition of “declarative region.” I've left the ambiguous use of “scope” alone for now. However sections 3.3.x all have headings reading “xxx scope,” but they don't mean the scope of a declaration but the different kinds of declarative regions and their effects on the scope of declarations contained therein. To me, it looks like most of 3.4 should refer to “declarative region” and not to “scope.”
The change to 6.7 fixes an “identifier” misuse (e.g., extern T operator+(T,T); at block scope introduces a name but not an identifier) and removes normative redundancy.]
The Standard does not completely specify how to look up the type-name(s) in a pseudo-destructor-name (5.2 [expr.post] paragraph 1, 5.2.4 [expr.pseudo]), and what information it does have is incorrect and/or in the wrong place. Consider, for instance, 3.4.5 [basic.lookup.classref] paragraphs 2-3:
If the id-expression in a class member access (5.2.5 [expr.ref]) is an unqualified-id, and the type of the object expression is of a class type C (or of pointer to a class type C), the unqualified-id is looked up in the scope of class C. If the type of the object expression is of pointer to scalar type, the unqualified-id is looked up in the context of the complete postfix-expression.
If the unqualified-id is ~type-name, and the type of the object expression is of a class type C (or of pointer to a class type C), the type-name is looked up in the context of the entire postfix-expression and in the scope of class C. The type-name shall refer to a class-name. If type-name is found in both contexts, the name shall refer to the same class type. If the type of the object expression is of scalar type, the type-name is looked up in the scope of the complete postfix-expression.
There are at least three things wrong with this passage with respect to pseudo-destructors:
A pseudo-destructor call (5.2.4 [expr.pseudo]) is not a “class member access”, so the statements about scalar types in the object expressions are vacuous: the object expression in a class member access is required to be a class type or pointer to class type (5.2.5 [expr.ref] paragraph 2).
On a related note, the lookup for the type-name(s) in a pseudo-destructor name should not be described in a section entitled “Class member access.”
Although the class member access object expressions are carefully allowed to be either a class type or a pointer to a class type, paragraph 2 mentions only a “pointer to scalar type” (disallowing references) and paragraph 3 deals only with a “scalar type,” presumably disallowing pointers (although it could possibly be a very subtle way of referring to both non-class pointers and references to scalar types at once).
The other point at which lookup of pseudo-destructors is mentioned is 3.4.3 [basic.lookup.qual] paragraph 5:
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.
Again, this specification is in the wrong location (a pseudo-destructor-name is not a qualified-id and thus should not be treated in the “Qualified name lookup” section).
Finally, there is no place in the Standard that describes the lookup for pseudo-destructor calls of the form p->T::~T() and r.T::~T(), where p and r are a pointer and reference to scalar, respectively. To the extent that it gives any guidance at all, 3.4.5 [basic.lookup.classref] deals only with the case where the ~ immediately follows the . or ->, and 3.4.3 [basic.lookup.qual] deals only with the case where the pseudo-destructor-name contains a nested-name-specifier that designates a scope in which names can be looked up.
See document J16/06-0008 = WG21 N1938 for further discussion of this and related issues, including 244, 305, 399, and 414.
Proposed resolution (June, 2008):
Add a new paragraph following 5.2 [expr.post] paragraph 2 as follows:
When a postfix-expression is followed by a dot . or arrow -> operator, the interpretation depends on the type T of the expression preceding the operator. If the operator is ., T shall be a scalar type or a complete class type; otherwise, T shall be a pointer to a scalar type or a pointer to a complete class type. When T is a (pointer to) a scalar type, the postfix-expression to which the operator belongs shall be a pseudo-destructor call (5.2.4 [expr.pseudo]); otherwise, it shall be a class member access (5.2.5 [expr.ref]).
Change 5.2.4 [expr.pseudo] paragraph 2 as follows:
The left-hand side of the dot operator shall be of scalar type. The left-hand side of the arrow operator shall be of pointer to scalar type. This scalar type The type of the expression preceding the dot operator, or the type to which the expression preceding the arrow operator points, is the object type...
Change 5.2.5 [expr.ref] paragraph 2 as follows:
For the first option (dot) the type of the first expression (the object expression) shall be “class object” (of a complete type) is a class type. For the second option (arrow) the type of the first expression (the pointer expression) shall be “pointer to class object” (of a complete type) is a pointer to a class type. In these cases, the id-expression shall name a member of the class or of one of its base classes.
Add a new paragraph following 3.4 [basic.lookup] paragraph 2 as follows:
In a pseudo-destructor-name that does not include a nested-name-specifier, the type-names are looked up as types in the context of the complete expression.
Delete the last sentence of 3.4.5 [basic.lookup.classref] paragraph 2:
If the id-expression in a class member access (5.2.5 [expr.ref]) is an unqualified-id, and the type of the object expression is of a class type C, the unqualified-id is looked up in the scope of class C. If the type of the object expression is of pointer to scalar type, the unqualified-id is looked up in the context of the complete postfix-expression.
Is this case valid? G++ compiles it.
namespace X {
namespace Y {
struct X {
void f()
{
using namespace X::Y;
namespace Z = X::Y;
}
};
}
}
The relevant citation from the standard is 3.4.6 [basic.lookup.udir]: "When looking up a namespace-name in a using-directive or namespace-alias-definition, only namespace names are considered." This statement could reasonably be interpreted to apply only to the last element of a qualified name, and that's the way EDG and Microsoft seem to interpret it.
However, since a class can't contain a namespace, it seems to me that this interpretation is, shall we say, sub optimal. If the X qualifiers in the above example are interpreted as referring to the struct X, an error of some sort is inevitable, since there can be no namespace for the qualified name to refer to. G++ apparently interprets 3.4.6 [basic.lookup.udir] as applying to nested-name-specifiers in those contexts as well, which makes a valid interpretation of the test possible.
I'm thinking it might be worth tweaking the words in 3.4.6 [basic.lookup.udir] to basically mandate the more useful interpretation. Of course a person could argue that the difference would matter only to a perverse program. On the other hand, namespaces were invented specifically to enable the building of programs that would otherwise be considered perverse. Where name clashes are concerned, one man's perverse is another man's real world.
Proposed Resolution (November, 2006):
Change 3.4.6 [basic.lookup.udir] paragraph 1 as follows:
When looking up a namespace-name in a using-directive or namespace-alias-definition, In a using-directive or namespace-alias-definition, during the lookup for a namespace-name or for a name in a nested-name-specifier, only namespace names are considered.
An lvalue referring to an out-of-lifetime non-POD class objects can be used in limited ways, subject to the restrictions in 3.8 [basic.life] paragraph 6:
if the original object will be or was of a non-POD class type, the program has undefined behavior if:
the lvalue is used to access a non-static data member or call a non-static member function of the object, or
the lvalue is implicitly converted (4.10 [conv.ptr]) to a reference to a base class type, 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
the lvalue is used as the operand of a dynamic_cast (5.2.7 [expr.dynamic.cast]) or as the operand of typeid.
There are at least a couple of questionable things in this list. First, there is no “implicit conversion to a reference to a base class,” as assumed by the second bullet. Presumably this is intended to say that the lvalue is bound to a reference to a base class, and the cross-reference should be to 8.5.3 [dcl.init.ref], not to 4.10 [conv.ptr] (which deals with pointer conversions). However, even given that adjustment, it is not clear why it is forbidden to bind a reference to a non-virtual base class of an out-of-lifetime object, as that is just an address offset calculation. (Binding to a virtual base, of course, would require access to the value of the object and thus cannot be done outside the object's lifetime.)
The third bullet also appears questionable. It's not clear why static_cast is discussed at all here, as the only permissible static_cast conversions involving reference types and non-POD classes are to references to base or derived classes and to the same type, modulo cv-qualification; if implicit “conversion” to a base class reference is forbidden in the second bullet, why would an explicit conversion be permitted in the third? Was this intended to refer to reinterpret_cast? Also, is there a reason to allow char types but disallow array-of-char types (which are more likely to be useful than a single char)?
Proposed resolution (March, 2008):
Change 3.8 [basic.life] paragraph 5 as follows:
...If the object will be or was of a non-trivial class type, the program has undefined behavior if:
the pointer is used to access a non-static data member or call a non-static member function of the object, or
the pointer is implicitly converted (
4.10 [conv.ptr] ) to a pointer to a virtual base class type, orthe pointer is used as the operand of a static_cast (5.2.9 [expr.static.cast]) (except when the conversion is to void*, or to void* and subsequently to char*, or unsigned char*). pointer to void, or to pointer to void and subsequently to pointer to cv char or pointer to cv unsigned char, or
the pointer is used as the operand of a dynamic_cast (5.2.7 [expr.dynamic.cast])...
Change 3.8 [basic.life] paragraph 6 as follows:
...if the original object will be or was of a non-trivial class type, the program has undefined behavior if:
the lvalue is used to access a non-static data member or call a non-static member function of the object, or
the lvalue is implicitly converted (4.10 [conv.ptr]) bound to a reference to a virtual base class type (8.5.3 [dcl.init.ref]), 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
the lvalue is used as the operand of a dynamic_cast (5.2.7 [expr.dynamic.cast]) or as the operand of typeid.
[Drafting notes: Paragraph 5 was changed to track the changes to paragraph 6. See also the resolution for issue 658.]
4 [conv] paragraph 1 says,
Standard conversions are implicit conversions defined for built-in types.
However, enumeration types (which take part in the integral promotions) and class types (which take part in the lvalue-to-rvalue conversion) are not “built-in” types, so the definition of “standard conversions” is wrong.
Proposed resolution (October, 2006):
Change 4 [conv] paragraph 1 as follows:
Standard conversions are implicit conversions defined for built-in types with built-in meaning...
4.1 [conv.lval] paragraph 1 says,
If the object to which the lvalue refers is not an object of type T and is not an object of a type derived from T, or if the object is uninitialized, a program that necessitates this conversion has undefined behavior.
I think there are at least three related issues around this specification:
Presumably assigning a valid value to an uninitialized object allows it to participate in the lvalue-to-rvalue conversion without undefined behavior (otherwise the number of programs with defined behavior would be vanishingly small :-). However, the wording here just says "uninitialized" and doesn't mention assignment.
There's no exception made for unsigned char types. The wording in 3.9.1 [basic.fundamental] was carefully crafted to allow use of unsigned char to access uninitialized data so that memcpy and such could be written in C++ without undefined behavior, but this statement undermines that intent.
It's possible to get an uninitialized rvalue without invoking the lvalue-to-rvalue conversion. For instance:
struct A {
int i;
A() { } // no init of A::i
};
int j = A().i; // uninitialized rvalue
There doesn't appear to be anything in the current IS wording that says that this is undefined behavior. My guess is that we thought that in placing the restriction on use of uninitialized objects in the lvalue-to-rvalue conversion we were catching all possible cases, but we missed this one.
In light of the above, I think the discussion of uninitialized objects ought to be removed from 4.1 [conv.lval] paragraph 1. Instead, something like the following ought to be added to 3.9 [basic.types] paragraph 4 (which is where the concept of "value" is introduced):
Any use of an indeterminate value (5.3.4 [expr.new], 8.5 [dcl.init], 12.6.2 [class.base.init]) of any type other than char or unsigned char results in undefined behavior.
John Max Skaller:
A().i had better be an lvalue; the rules are wrong. Accessing a member of a structure requires it be converted to an lvalue, the above calculation is 'as if':
struct A {
int i;
A *get() { return this; }
};
int j = (*A().get()).i;
and you can see the bracketed expression is an lvalue.
A consequence is:
int &j= A().i; // OK, even if the temporary evaporates
j now refers to a 'destroyed' value. Any use of j is an error. But the binding at the time is valid.
Proposed Resolution (November, 2006):
Add the indicated words to 3.9 [basic.types] paragraph 4:
... For trivial types, the value representation is a set of bits in the object representation that determines a value, which is one discrete element of an implementation-defined set of values. Any use of an indeterminate value (5.3.4 [expr.new], 8.5 [dcl.init], 12.6.2 [class.base.init]) of a type other than unsigned char results in undefined behavior.
Change 4.1 [conv.lval] paragraph 1 as follows:
If the object to which the lvalue refers is not an object of type T and is not an object of a type derived from T, or if the object is uninitialized, a program that necessitates this conversion has undefined behavior.
Additional note (May, 2008):
The C committee is dealing with a similar issue in their DR336. According to this analysis, they plan to take almost the opposite approach to the one described above by augmenting the description of their version of the lvalue-to-rvalue conversion. The CWG did not consider that access to an unsigned char might still trap if it is allocated in a register and needs to reevaluate the proposed resolution in that light. See also issue 129.
The adoption of paper N2844 made it ill-formed to attempt to bind an rvalue reference to an lvalue, but the example in 5 [expr] paragraph 6 was overlooked in making this change:
struct A { };
A&& operator+(A, A);
A&& f();
A a;
A&& ar = a;
The last line should be changed to use something like static_cast<A&&>(a).
(See also issue 847.)
Proposed resolution (July, 2009):
Change the example in 5 [expr] paragraph 6 as follows:
[Example:
struct A { }; A&& operator+(A, A); A&& f(); A a; A&& ar = static_cast<A&&>(a);The expressions f() and a + a are rvalues of type A. The expression ar is an lvalue of type A. —end example]
Split off from issue 315.
Incidentally, another thing that ought to be cleaned up is the inconsistent use of "indirection" and "dereference". We should pick one.
Proposed resolution (December, 2006):
Change 5.3.1 [expr.unary.op] paragraph 1 as follows:
The unary * operator performs indirection dereferences a pointer value: the expression to which it is applied shall be a pointer...
Change 8.3.4 [dcl.array] paragraph 8 as follows:
The results are added and indirection applied values are added and the result is dereferenced to yield an array (of five integers), which in turn is converted to a pointer to the first of the integers.
Change 8.3.5 [dcl.fct] paragraph 9 as follows:
The binding of *fpi(int) is *(fpi(int)), so the declaration suggests, and the same construction in an expression requires, the calling of a function fpi, and then using indirection through dereferencing the (pointer) result to yield an integer. In the declarator (*pif)(const char*, const char*), the extra parentheses are necessary to indicate that indirection through dereferencing a pointer to a function yields a function, which is then called.
Change the index for * and “dereferencing” no longer to refer to “indirection.”
[Drafting note: 26.6.9 [template.indirect.array] requires no change. Many more places in the current wording use “dereferencing” than “indirection.”]
According to the C++ Standard section 5.3.4 [expr.new] paragraph 21 it is unspecified whether the allocation function is called before evaluating the constructor arguments or after evaluating the constructor arguments but before entering the constructor.
On top of that paragraph 17 of the same section insists that
If any part of the object initialization described above [Footnote: This may include evaluating a new-initializer and/or calling a constructor.] terminates by throwing an exception and a suitable deallocation function is found, the deallocation function is called to free the memory in which the object was being constructed... If no unambiguous matching deallocation function can be found, propagating the exception does not cause the object's memory to be freed...
Now suppose we have:
struct copy_throw {
copy_throw(const copy_throw&)
{ throw std::logic_error("Cannot copy!"); }
copy_throw(long, copy_throw)
{ }
copy_throw()
{ }
};
int main()
try {
copy_throw an_object, /* undefined behaviour */
* a_pointer = ::new copy_throw(0, an_object);
return 0;
}
catch(const std::logic_error&)
{ }
Here the new-expression '::new copy_throw(0, an_object)' throws an exception when evaluating the constructor's arguments and before the allocation function is called. However, 5.3.4 [expr.new] paragraph 17 prescribes that in such a case the implementation shall call the deallocation function to free the memory in which the object was being constructed, given that a matching deallocation function can be found.
So a call to the Standard library deallocation function '::operator delete(void*)' shall be issued, but what argument is an implementation supposed to supply to the deallocation function? As per 5.3.4 [expr.new] paragraph 17 - the argument is the address of the memory in which the object was being constructed. Given that no memory has yet been allocated for the object, this will qualify as using an invalid pointer value, which is undefined behaviour by virtue of 3.7.4.2 [basic.stc.dynamic.deallocation] paragraph 4.
Suggested resolution:
Change the first sentence of 5.3.4 [expr.new] paragraph 17 to read:
If the memory for the object being created has already been successfully allocated and any part of the object initialization described above...
Proposed resolution (March, 2008):
Change 5.3.4 [expr.new] paragraph 18 as follows:
If any part of the object initialization described above [Footnote: ...] terminates by throwing an exception, storage has been obtained for the object, and a suitable deallocation function can be found, the deallocation function is called...
Does an explicit temporary of an integral type qualify as an integral constant expression? For instance,
void* p = int(); // well-formed?
It would appear to be, since int() is an explicit type conversion according to 5.2.3 [expr.type.conv] (at least, it's described in a section entitled "Explicit type conversion") and type conversions to integral types are permitted in integral constant expressions (5.19 [expr.const]). However, this reasoning is somewhat tenuous, and some at least have argued otherwise.
Note (March, 2008):
This issue should be closed as NAD as a result of the rewrite of 5.19 [expr.const] in conjunction with the constexpr proposal.
The constraints on type-specifiers given in 7.1.6 [dcl.type] paragraphs 2 and 3 (at most one type-specifier except as specified, at least one type-specifier, no redundant cv-qualifiers) are couched in terms of decl-specifier-seqs and declarations. However, they should also apply to constructs that are not syntactically declarations and that are defined to use type-specifier-seqs, including 5.3.4 [expr.new], 6.6 [stmt.jump], 8.1 [dcl.name], and 12.3.2 [class.conv.fct].
Proposed resolution (March, 2008):
Change 7.1.6 [dcl.type] paragraph 3 as follows:
At In a complete type-specifier-seq or in a complete decl-specifier-seq of a declaration, at least one type-specifier that is not a cv-qualifier is required in a declaration shall appear unless it the declaration declares a constructor, destructor or conversion function.
(Note: paper N2546, voted into the Working Draft in February, 2008, addresses part of this issue.)
A function with an exception-specification of throw() must be given a catch(...) clause to enforce its contract, i.e., to call std::unexpected() if it exits with an exception. It would be useful to have an attribute indicating that the function really does throw nothing and thus that the catch(...) clause need not be generated.
(See also issue 830.)
Proposed resolution (September, 2009):
See paper PL22.16/09-0162 = WG21 N2972.
According to 8.3 [dcl.meaning] paragraph 1,
A declarator-id shall not be qualified except for the definition of a member function (9.3 [class.mfct]) or static data member (9.4 [class.static]) outside of its class, the definition or explicit instantiation of a function or variable member of a namespace outside of its namespace, or the definition of a previously declared explicit specialization outside of its namespace, or the declaration of a friend function that is a member of another class or namespace (11.4 [class.friend]). When the declarator-id is qualified, the declaration shall refer to a previously declared member of the class or namespace to which the qualifier refers...
This restriction prohibits examples like the following:
void f();
void ::f(); // error: qualified declarator
namespace N {
void f();
void N::f() { } // error: qualified declarator
}
There doesn't seem to be any good reason for disallowing such declarations, and a number of implementations accept them in spite of the Standard's prohibition. Should the Standard be changed to allow them?
Notes from the April, 2006 meeting:
In discussing issue 548, the CWG agreed that the prohibition of qualified declarators inside their namespace should be removed.
Proposed resolution (October, 2006):
Remove the indicated words from 8.3 [dcl.meaning] paragraph 1:
...An unqualified-id occurring in a declarator-id shall be a simple identifier except for the declaration of some special functions (12.3 [class.conv], 12.4 [class.dtor], 13.5 [over.oper]) and for the declaration of template specializations or partial specializations (). A declarator-id shall not be qualified except for the definition of a member function (9.3 [class.mfct]) or static data member (9.4 [class.static]) outside of its class, the definition or explicit instantiation of a function or variable member of a namespace outside of its namespace, or the definition of a previously declared explicit specialization outside of its namespace, or the declaration of a friend function that is a member of another class or namespace (11.4 [class.friend]). When the declarator-id is qualified, the declaration shall refer to a previously declared member of the class or namespace to which the qualifier 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 of the declarator-id...
[Drafting note: The omission of “outside of its class” here does not give permission for redeclaration of class members; that is still prohibited by 9.2 [class.mem] paragraph 1. The removal of the enumeration of the kinds of declarations in which a qualified-id can appear does allow a typedef declaration to use a qualified-id, which was not permitted before; if that is undesirable, the prohibition can be reinstated here.]
The following example appears to be well-formed, with the partial specialization matching the type of Y::f(), even though it is rejected by many compilers:
template<class T> struct X;
template<class R> struct X< R() > {
};
template<class F, class T> void test(F T::* pmf) {
X<F> x;
}
struct Y {
void f() {
}
};
int main() {
test( &Y::f );
}
However, 8.3.5 [dcl.fct] paragraph 4 says,
A cv-qualifier-seq shall only be part of the function type for a non-static member function, the function type to which a pointer to member refers, or the top-level function type of a function typedef declaration. 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.
This specification makes it impossible to write a partial specialization for a const member function:
template<class R> struct X<R() const> {
};
A template argument is not one of the permitted contexts for cv-qualification of a function type. This restriction should be removed.
Notes from the April, 2006 meeting:
During the meeting the CWG was of the opinion that the “R() const” specialization would not match the const member function even if it were allowed and so classified the issue as NAD. Questions have been raised since the meeting, however, suggesting that the template argument in the partial specialization would, in fact, match the type of a const member function (see, for example, the very similar usage via typedefs in 9.3 [class.mfct] paragraph 9). The issue is thus being left open for renewed discussion at the next meeting.
Proposed resolution (June, 2008):
Change 8.3.5 [dcl.fct] paragraph 7 as follows:
A cv-qualifier-seq shall only be part of the function type for a non-static member function, the function type to which a pointer to member refers, or the top-level function type of a function typedef declaration, or the top-level function type of a type-id that is a template-argument for a type template-parameter. The effect... A ref-qualifier shall only be part of the function type for a non-static member function, the function type to which a pointer to member refers, or the top-level function type of a function typedef declaration, or the top-level function type of a type-id that is a template-argument for a type template-parameter. The return type...
In looking at a large handful of core issues related to elaborated-type-specifiers and the naming of classes in general, I discovered an odd fact. It turns out that there is exactly one place in the grammar where nested-name-specifier is not immediately preceded by "::opt": in class-head, which is used only for class definitions. So technically, this example is ill-formed, and should evoke a syntax error:
struct A;
struct ::A { };
However, all of EDG, GCC and Microsoft's compiler accept it without a qualm. In fact, I couldn't get any of them to even warn about it.
Suggested resolution:
It would simplify the grammar, and apparently better reflect existing practice, to factor the global-scope operator into the rule for nested-name-specifier.
Proposed resolution (November, 2006):
In 3.4.3 [basic.lookup.qual] paragraph 6, change the grammar snippet as follows:
Delete 5.1.1 [expr.prim.general] paragraph 4 (“The operator :: followed by...”). [Drafting note: It's covered by paragraph 8 (type, lvalue-ness, member-ness, reference to 3.4.3.2 [namespace.qual]) and 3.4.3.2 [namespace.qual] (qualified lookup for namespace members).]
Change the grammar in 5.1.1 [expr.prim.general] paragraph 7 as follows (deleting the :: forms from qualified-id and adding :: as a new production for nested-name-specifier):
Change 5.1.1 [expr.prim.general] paragraph 8 as follows:
A nested-name-specifier that names designates a namespace (7.3 [basic.namespace]), followed by the name of a member of that namespace...
Change 5.1.1 [expr.prim.general] paragraph 10 as follows:
In a qualified-id, if the id-expression unqualified-id is a conversion-function-id...
In 5.2 [expr.post] paragraph 1, change the grammar as follows:
In 5.2.4 [expr.pseudo] paragraph 2, change the grammar snippet as follows:
In 7.1.6.2 [dcl.type.simple] paragraph 1, change the grammar as follows:
In 7.1.6.3 [dcl.type.elab] before paragraph 1, change the grammar as follows:
In 7.1.6.3 [dcl.type.elab] paragraph 1, change the grammar snippet as follows:
In 7.3.2 [namespace.alias] paragraph 1, change the grammar as follows:
In 7.3.3 [namespace.udecl] paragraph 1, change the grammar as follows:
In 7.3.4 [namespace.udir] before paragraph 1, change the grammar as follows:
In 8 [dcl.decl] paragraph 4, change the grammar as follows:
In 8.3.3 [dcl.mptr] paragraph 1, change the grammar snippet as follows:
In 9.2 [class.mem] before paragraph 1, change the grammar as follows:
In 10 [class.derived] paragraph 1, change the grammar as follows:
In 12.6.2 [class.base.init] paragraph 1, change the grammar as follows:
In 14.7 [temp.res] paragraph 3, change the grammar as follows:
[Drafting notes: gcc 4.1.1 rejects the example in the issue description. I still think it's a good idea to make the grammar more uniform, and there ought to be nothing special about the global scope operator. However, there is a slight change in effective grammar with these modification: all places that require a non-optional nested-name-specifier used to required at least one named level of nesting. With these changes, "::" is a valid nested-name-specifier (that denotes the global scope). Any such use needed to protect against non-class (i.e. namespace) scopes in its semantic description anyway, which also covers the "::" case.]
Can a member of a union be of a class that has a user-declared non-default constructor? The restrictions on union membership in 9.5 [class.union] paragraph 1 only mention default and copy constructors:
An object of a class with a non-trivial default constructor (12.1 [class.ctor]), a non-trivial copy constructor (12.8 [class.copy]), a non-trivial destructor (12.4 [class.dtor]), or a non-trivial copy assignment operator (13.5.3 [over.ass], 12.8 [class.copy]) cannot be a member of a union...
(12.1 [class.ctor] paragraph 11 does say, “a non-trivial constructor,” but it's not clear whether that was intended to refer only to default and copy constructors or to any user-declared constructor. For example, 12.2 [class.temporary] paragraph 3 also speaks of a “non-trivial constructor,” but the cross-references there make it clear that only default and copy constructors are in view.)
Note (March, 2008):
This issue was resolved by the adoption of paper J16/08-0054 = WG21 N2544 (“Unrestricted Unions”) at the Bellevue meeting.
According to 12.1 [class.ctor] paragraph 5,
An implicitly-declared default constructor for class X is defined as deleted if: ... any non-static data member of const-qualified type (or array thereof) does not have a user-provided default constructor, or...
It is not clear if this adequately covers the case in which some variant members are const-qualified but others are not. The intent of the restriction is to prevent creation of an object with uninitialized members that would require a const_cast to set their value later, but const-qualified members of an anonymous union in which other members are not const do not seem to present that problem.
Proposed resolution (October, 2009):
Change 12.1 [class.ctor] paragraph 5 bullet 3 of the second list and add a fourth bullet as follows:
...
any non-variant non-static data member of const-qualified type (or array thereof) does not have a user-provided default constructor, or
all variant members are of const-qualified type (or array thereof), or
...
Split off from issue 86.
Should binding a reference to the result of a "," operation whose second operand is a temporary extend the lifetime of the temporary?
const SFileName &C = ( f(), SFileName("abc") );
Notes from the March 2004 meeting:
We think the temporary should be extended.
Proposed resolution (October, 2004):
Change 12.2 [class.temporary] paragraph 2 as indicated:
... In all these cases, the temporaries created during the evaluation of the expression initializing the reference, except the temporary that is the overall result of the expression [Footnote: For example, if the expression is a comma expression (5.18 [expr.comma]) and the value of its second operand is a temporary, the reference is bound to that temporary.] and to which the reference is bound, are destroyed at the end of the full-expression in which they are created and in the reverse order of the completion of their construction...
[Note: this wording partially resolves issue 86. See also issue 446.]
Notes from the April, 2005 meeting:
The CWG suggested a different approach from the 10/2004 resolution, leaving 12.2 [class.temporary] unchanged and adding normative wording to 5.18 [expr.comma] specifying that, if the result of the second operand is a temporary, that temporary is the result of the comma expression as well.
Proposed Resolution (November, 2006):
Add the indicated wording to 5.18 [expr.comma] paragraph 1:
... 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 an lvalue, and is a bit-field if its right operand is an lvalue and a bit-field. If the value of the right operand is a temporary (12.2 [class.temporary]), the result is that temporary.
Jack Rouse: In 12.8 [class.copy] paragraph 8, the standard includes the following about the copying of class subobjects in such a constructor:
Mike Miller: I'm more concerned about 12.8 [class.copy] paragraph 7, which lists the situations in which an implicitly-defined copy constructor can render a program ill-formed. Inaccessible and ambiguous copy constructors are listed, but not a copy constructor with a cv-qualification mismatch. These two paragraphs taken together could be read as requiring the calling of a copy constructor with a non-const reference parameter for a const data member.
Proposed Resolution (November, 2006):
This issue is resolved by the proposed resolution for issue 535.
Footnote 112 (12.8 [class.copy] paragraph 2) says,
Because a template constructor is never a copy constructor, the presence of such a template does not suppress the implicit declaration of a copy constructor. Template constructors participate in overload resolution with other constructors, including copy constructors, and a template constructor may be used to copy an object if it provides a better match than other constructors.
However, many of the stipulations about copy construction are phrased to refer only to “copy constructors.” For example, 12.8 [class.copy] paragraph 14 says,
A program is ill-formed if the copy constructor... for an object is implicitly used and the special member function is not accessible (clause 11 [class.access]).
Does that mean that using an inaccessible template constructor to copy an object is permissible, because it is not a “copy constructor?” Obviously not, but each use of the term “copy constructor” in the Standard should be examined to determine if it applies strictly to copy constructors or to any constructor used for copying. (A similar issue applies to “copy assignment operators,” which have the same relationship to assignment operator function templates.)
Proposed Resolution (February, 2008):
Change 3.2 [basic.def.odr] paragraph 2 as follows:
... [Note: this covers calls to named functions (5.2.2 [expr.call]), operator overloading (clause 13 [over]), user-defined conversions (12.3.2 [class.conv.fct]), allocation function for placement new (5.3.4 [expr.new]), as well as non-default initialization (8.5 [dcl.init]). A copy constructor selected to copy class objects is used even if the call is actually elided by the implementation (12.8 [class.copy]). —end note] ... A copy-assignment function for a class An assignment operator function in a class is used by an implicitly-defined copy-assignment function for another class as specified in 12.8 [class.copy]...
Delete 12.1 [class.ctor] paragraphs 10 and 11:
A copy constructor (12.8 [class.copy]) is used to copy objects of class type.
A union member shall not be of a class type (or array thereof) that has a non-trivial constructor.
Replace the “example” in 12.2 [class.temporary] paragraph 1 with a note as follows:
[Example: even if the copy constructor is not called, all the semantic restrictions, such as accessibility (clause 11 [class.access]), shall be satisfied. —end example] [Note: This includes accessibility (clause 11 [class.access]) for the constructor selected. —end note]
Change 12.8 [class.copy] paragraph 7 as follows:
A non-user-provided copy constructor is implicitly defined if it is used to initialize an object of its class type from a copy of an object of its class type or of a class type derived from its class type (3.2 [basic.def.odr]). [Footnote: See 8.5 [dcl.init] for more details on direct and copy initialization. —end footnote] [Note: the copy constructor is implicitly defined even if the implementation elided its use (12.2 [class.temporary]) the copy operation (12.8 [class.copy]). —end note] A program is ill-formed if the class for which a copy constructor is implicitly defined or explicitly defaulted has:
a non-static data member of class type (or array thereof) with an inaccessible or ambiguous copy constructor, or
a base class with an inaccessible or ambiguous copy constructor.
Before the non-user-provided copy constructor for a class is implicitly defined...
Change 12.8 [class.copy] paragraph 8 as follows:
...Each subobject is copied in the manner appropriate to its type:
if the subobject is of class type, the copy constructor for the class is used direct-initialization (8.5 [dcl.init]) is performed [Note: If overload resolution fails or the constructor selected by overload resolution is inaccessible (11 [class.access]) in the context of X, the program is ill-formed. —end note];
if the subobject is an array...
[Drafting note: 8.5 [dcl.init] paragraph 15 requires “unambiguous” and 13.3 [over.match] paragraph 3 requires “accessible,” thus no need for normative text here.]
Change 12.8 [class.copy] paragraph 12 as follows:
A non-user-provided copy assignment operator is implicitly defined when an object of its class type is assigned a value of its class type or a value of a class type derived from its class type it is used (3.2 [basic.def.odr]). A program is ill-formed if the class for which a copy assignment operator is implicitly defined or explicitly defaulted has: a non-static data member of const or reference type.
a non-static data member of const type, or
a non-static data member of reference type, or
a non-static data member of class type (or array thereof) with an inaccessible copy assignment operator, or
a base class with an inaccessible copy assignment operator.
Change 12.8 [class.copy] paragraph 13 as follows:
... Each subobject is assigned in the manner appropriate to its type:
if the subobject is of class type, the copy assignment operator for the class the assignment operator function selected by overload resolution (13.3 [over.match]) for that class is used (as if by explicit qualification; that is, ignoring any possible virtual overriding functions in more derived classes) [Note: If overload resolution fails or the assignment operator function selected by overload resolution is inaccessible (11 [class.access]) in the context of X, the program is ill-formed. —end note];
if the subobject is an array...
Delete 12.8 [class.copy] paragraph 14:
A program is ill-formed if the copy constructor or the copy assignment operator for an object is implicitly used and the special member function is not accessible (clause 11 [class.access]). [Note: Copying one object into another using the copy constructor or the copy assignment operator does not change the layout or size of either object. —end note]
Change 12.8 [class.copy] paragraph 15 as follows:
When certain criteria are met, an implementation is allowed to omit the copy construction of a class object, even if the copy constructor selected for the copy operation and/or the destructor for the object have side effects. In such cases, the implementation treats the source and target of the omitted copy operation as simply two different ways of referring to the same object, and the destruction of that object occurs at the later of the times when the two objects would have been destroyed without the optimization. [Footnote: Because only one object is destroyed instead of two, and one copy constructor is not executed, there is still one object destroyed for each one constructed. —end footnote] This elision...
Change 13.3.3.1.2 [over.ics.user] paragraph 4 as follows:
A conversion of an expression of class type to the same class type is given Exact Match rank, and a conversion of an expression of class type to a base class of that type is given Conversion rank, in spite of the fact that a copy constructor (i.e., a user-defined conversion function) is called for those cases.
Change 15.1 [except.throw] paragraph 3 as follows:
A throw-expression initializes a temporary object, called the exception object, the type of which by copy-initialization (8.5 [dcl.init]). The type of that temporary object is determined...
Change 15.1 [except.throw] paragraph 5 as follows:
When the thrown object is a class object, the copy constructor selected for the copy-initialization and the destructor shall be accessible, even if the copy operation is elided (12.8 [class.copy]).
Change 15.3 [except.handle] paragraphs 16-17 as follows:
When the exception-declaration specifies a class type, a copy constructor copy-initialization (8.5 [dcl.init]) is used to initialize either the object declared in the exception-declaration or, if the exception-declaration does not specify a name, a temporary object of that type. The object shall not have an abstract class type. The object is destroyed when the handler exits, after the destruction of any automatic objects initialized within the handler. The copy constructor selected for the copy-initialization and the destructor shall be accessible in the context of the handler, even if the copy operation is elided (12.8 [class.copy]). If the copy constructor and destructor are implicitly declared (12.8 [class.copy]), such a use in the handler causes these functions to be implicitly defined; otherwise, the program shall provide a definition for these functions.
The copy constructor and destructor associated with the object shall be accessible even if the copy operation is elided (12.8 [class.copy]).
Change the footnote in 15.5.1 [except.terminate] paragraph 1 as follows:
[Footnote: For example, if the object being thrown is of a class with a copy constructor type, std::terminate() will be called if that copy constructor the constructor selected to copy the object exits with an exception during a throw. —end footnote]
(This resolution also resolves issue 111.)
[Drafting note: The following do not require changes: 5.17 [expr.ass] paragraph 4; 9 [class] paragraph 5; 9.5 [class.union] paragraph 1; 12.2 [class.temporary] paragraph 2; 12.8 [class.copy] paragraphs 1-2; 15.4 [except.spec] paragraph 14.]
Notes from February, 2008 meeting:
These changes overlap those that will be made when concepts are added. This issue will be maintained in “review” status until the concepts proposal is adopted and any conflicts will be resolved at that point.
According to 1.3 [intro.defs], “dynamic type,”
The dynamic type of an rvalue expression is its static type.
This is not true of an rvalue reference, which can be bound to an object of a class type derived from the reference's static type.
Proposed resolution (June, 2008):
Change 1.3 [intro.defs], “dynamic type,” as follows:
the type of the most derived object (1.8 [intro.object]) to which the lvalue denoted by an lvalue or an rvalue-reference (clause 5 [expr]) expression refers. [Example: if a pointer (8.3.1 [dcl.ptr]) p whose static type is “pointer to class B” is pointing to an object of class D, derived from B (clause 10 [class.derived]), the dynamic type of the expression *p is “D.” References (8.3.2 [dcl.ref]) are treated similarly. —end example] The dynamic type of an rvalue expression that is not an rvalue reference is its static type.
Notes from the June, 2008 meeting:
Because expressions have an rvalue reference type only fleetingly, immediately becoming either lvalues or rvalues and no longer references, the CWG expressed a desire for a different approach that would somehow describe an rvalue that resulted from an rvalue reference instead of using the concept of an expression that is an rvalue reference, as above. This approach could also be used in the resolution of issue 664.
Additional note (August, 2008):
This issue, along with issue 664, indicates that rvalue references have more in common with lvalues than with other rvalues: they denote particular objects, thus allowing object identity and polymorphic behavior. That suggests that these issues may be just the tip of the iceberg: restrictions on out-of-lifetime access to objects, the aliasing rules, and many other specifications are written to apply only to lvalues, on the assumption that only lvalues refer to specific objects. That assumption is no longer valid with rvalue references.
This suggests that it might be better to classify all rvalue references, not just named rvalue references, as lvalues instead of rvalues, and then just change the reference binding, overload resolution, and template argument deduction rules to cater to the specific kind of lvalues that are associated with rvalue references.
Additional note, May, 2009:
Another place in the Standard where the assumption is made that only lvalues can have dynamic types that differ from their static types is 5.2.8 [expr.typeid] paragraph 2.
(See also issues 846 and 863.)
Additional note, September, 2009:
Yet another complication is the statement in 3.10 [basic.lval] paragraph 9 stating that “non-class rvalues always have cv-unqualified types.” If an rvalue reference is an rvalue, then the following example is well-formed:
void f(int&&); // reference to non-const
void g() {
const int i = 0;
f(static_cast<const int&&>(i));
}
The static_cast binds an rvalue reference to the const object i, but the fact that it's an rvalue means that the cv-qualification is lost, effectively allowing the parameter of f, a reference to non-const, to bind directly to the const object.
There are several instances of undefined behavior in lexical processing:
2.2 [lex.phases] paragraph 1, phase 2: a universal-character-name resulting from a line splice.
2.2 [lex.phases] paragraph 1, phase 2: a file ending without a new-line character or with a new-line character that is spliced away.
2.2 [lex.phases] paragraph 1, phase 4: a universal-character-name resulting from macro token concatenation.
2.9 [lex.header] paragraph 2: ', \, /*, //, or " appearing in a header-name.
These would be more appropriately handled as conditionally-supported behavior, requiring implementations either to document their handling of these constructs or to issue a diagnostic.
Additional note, March, 2009:
The undefined behavior referred to above regarding universal-character-names is the result of the considerations described in the C99 Rationale, section 5.2.1, in the part entitled “UCN models.” Three different models for support of UCNs are described, each involving different conversions between UCNs and wide characters and/or at different times during program translation. Implementations, as well as the specification in a language standard, can employ any of the three, but it must be impossible for a well-defined program to determine which model was actually employed by implementation. The implication of this “equivalence principle” is that any construct that would give different results under the different models must be classified as undefined behavior. For example, an apparent UCN resulting from a line-splice would be recognized as a UCN by an implementation in which all wide characters were translated immediately into UCNs, as described in C++ phase 1, but would not be recognized as a UCN by another implementation in which all UCNs were translated immediately into wide characters (a possibility mentioned parenthetically in C++ phase 1).
There are additional implications for this “equivalence principle” beyond the ones identified in the UK CD comments. See also issue 578; presumably a string like the one in that issue should also be described as having undefined behavior. Also, because C++'s model introduces backslash characters as part of UCNs for any character outside the basic source character set, any header-name that contains such a character (e.g., #include "@.h") will have undefined behavior in C++. This is also the reason that UCNs are translated into wide characters inside raw strings: two of the three models articulated in the C99 Rationale translate to or from UCNs in phase 1, before raw strings are recognized as tokens in phase 3, so raw strings cannot treat UCNs differently from the way they are treated in other contexts. See also issue 789 for similar points regarding trigraphs.
Notes from the October, 2009 meeting:
The CWG decided that the non-UCN aspects of this issue should be resolved, while the overall questions regarding trigraphs, UCNs, and raw strings will be investigated separately.
According to 2.3 [lex.charset] paragraph 3,
The values of the members of the execution character sets are implementation-defined, and any additional members are locale-specific.
This makes it sound as if the locale determines only whether an extended character (one not in the basic execution character set) exists, not its value (which is just implementation-defined, not locale-specific). The description should be clarified to indicate that the value of a given character can vary between locales, as well.
2.5 [lex.pptoken] paragraph 2 specifies that there are 5 categories of tokens in phases 3 to 6. With 2.13 [lex.operators] paragraph 1, it is unclear whether new is an identifier or a preprocessing-op-or-punc; likewise for delete. This is relevant to answer the question whether
#define delete foo
is a well-formed control-line, since that requires an identifier after the define token.
(See also issue 189.)
The nonterminals operator and punctuator in 2.7 [lex.token] are not defined. There is a definition of the nonterminal operator in 13.5 [over.oper] paragraph 1, but it is apparent that the two nonterminals are not the same: the latter includes keywords and multi-token operators and does not include the nonoverloadable operators mentioned in paragraph 3.
There is a definition of preprocessing-op-or-punc in 2.13 [lex.operators] , with the notation that
Each preprocessing-op-or-punc is converted to a single token in translation phase 7 (2.1).However, this list doesn't distinguish between operators and punctuators, it includes digraphs and keywords (can a given token be both a keyword and an operator at the same time?), etc.
Suggested resolution:
Additional note (April, 2005):
The resolution for this problem should also address the fact that sizeof and typeid (and potentially others like decltype that may be added in the future) are described in some places as “operators” but are not listed in 13.5 [over.oper] paragraph 3 among the operators that cannot be overloaded.
(See also issue 369.)
3.1 [basic.def] makes statements about declarations that do not appear to apply to static_assert-declarations. For example, paragraph 1 says,
A declaration (clause 7 [dcl.dcl]) introduces names into a translation unit or redeclares names introduced by previous declarations. A declaration specifies the interpretation and attributes of these names.
What name is being declared or described by a static_assert-declaration?
Also, paragraph 2 lists the kinds of declarations that are not definitions, and a static_assert-declaration is not among them. Is it intentional that static_assert-declarations are definitions?
I thought this case would result in undefined behavior according to 3.2 [basic.def.odr]:
// t.h:
struct A { void (*p)(); };
// t1.cpp:
#include "t.h" // A::p is a pointer to C++ func
// t2.cpp:
extern "C" {
#include "t.h" // A::p is a pointer to C func
}
...but I don't see how any of the bullets in the list in paragraph 5 apply.
In describing static data members initialized inside the class definition, 9.4.2 [class.static.data] paragraph 3 says,
The member shall still be defined in a namespace scope if it is used in the program...
The definition of “used” is in 3.2 [basic.def.odr] paragraph 1:
An object or non-overloaded function whose name appears as a potentially-evaluated expression is used unless it is an object that satisfies the requirements for appearing in a constant expression (5.19 [expr.const]) and the lvalue-to-rvalue conversion (4.1 [conv.lval]) is immediately applied.
Now consider the following example:
struct S {
static const int a = 1;
static const int b = 2;
};
int f(bool x) {
return x ? S::a : S::b;
}
According to the current wording of the Standard, this example requires that S::a and S::b be defined in a namespace scope. The reason for this is that, according to 5.16 [expr.cond] paragraph 4, the result of this conditional-expression is an lvalue and the lvalue-to-rvalue conversion is applied to that, not directly to the object, so this fails the “immediately applied” requirement. This is surprising and unfortunate, since only the values and not the addresses of the static data members are used. (This problem also applies to the proposed resolution of issue 696.)
When 3.4.1 [basic.lookup.unqual] paragraph 10 says,
In a friend declaration naming a member function, a name used in the function declarator and not part of a template-argument in a template-id is first looked up in the scope of the member function's class. If it is not found, or if the name is part of a template-argument in a template-id, the look up is as described for unqualified names in the definition of the class granting friendship.
what does “in the scope of the member function's class” mean? Does it mean that only members of the class and its base classes are considered? Or does it mean that the same lookup is to be performed as if the name appeared in the member function's class? Implementations vary in this regard. For example:
struct s1;
namespace ns {
struct s1;
}
struct s2 {
void f(s1 &);
};
namespace ns {
struct s3 {
friend void s2::f(s1 &);
};
}
Microsoft Visual C++ and Comeau C++ resolve s1 in the friend declaration to ns::s1 and issue an error, while g++ resolves it to ::s1 and accepts the code.
Notes from the April, 2005 meeting:
The phrase “looked up in the scope of [a] class” occurs frequently throughout the Standard and always refers to the member name lookup described in 10.2 [class.member.lookup]. This is the first interpretation mentioned above (“only members of the class and its base classes”), resolving s1 to ns::s1. A cross-reference to 10.2 [class.member.lookup] will be added to 3.4.1 [basic.lookup.unqual] paragraph 10 to make this clearer.
In discussing this question, the CWG noticed another problem: the text quoted above applies to all template-arguments appearing in the function declarator. The intention of this rule, however, is that only template-arguments in the declarator-id should ignore the member function's class scope; template-arguments used elsewhere in the function declarator should be treated like other names. For example:
template<typename T> struct S;
struct A {
typedef int T;
void foo(S<T>);
};
template <typename T> struct B {
friend void A::foo(S<T>); // i.e., S<A::T>
};
In discussing issue 197, the question arose as to whether the handling of fundamental types in argument-dependent lookup is actually what is desired. This question needs further discussion.
The algorithm for namespace-qualified lookup is given in 3.4.3.2 [namespace.qual] paragraph 2:
Given X::m (where X is a user-declared namespace), or given ::m (where X is the global namespace), let S be the set of all declarations of m in X and in the transitive closure of all namespaces nominated by using-directives in X and its used namespaces, except that using-directives that nominate non-inline namespaces (7.3.1 [namespace.def]) are ignored in any namespace, including X, directly containing one or more declarations of m.
Consider the following example:
namespace A {
inline namespace B {
namespace C {
int i;
}
using namespace C;
}
int i;
}
int j = A::i; // ambiguous
The transitive closure includes B because it is inline, and it includes C because there is no declaration of i in B. As a result, A::i finds both the i declared in A and the one declared in C, and the lookup is ambiguous.
This result is apparently unintended.
Proposed resolution (July, 2009):
Change 7.3.1 [namespace.def] paragraph 9 as follows:
These properties are transitive: if a namespace N contains an inline namespace M, which in turn contains an inline namespace O, then the members of O can be used as though they were members of M or N. The transitive closure of all inline namespaces in N is the inline namespace set of N. The set of namespaces consisting of the innermost non-inline namespace enclosing an inline namespace O, together with any intervening inline namespaces, is the enclosing namespace set of O.
Change 3.4.3.2 [namespace.qual] paragraph 2 as follows:
Given X::m (where X is a user-declared namespace), or given ::m (where X is the global namespace), let S' be the set of all declarations of m in X and in the inline namespace set of X (7.3.1 [namespace.def]). If S' is not empty, S is S'; otherwise, let S be the set of all declarations of m in X and in the transitive closure of all namespaces nominated by using-directives in X and its used namespaces, except that using-directives that nominate non-inline namespaces (7.3.1 [namespace.def]) are ignored in any namespace, including X, directly containing one or more declarations of m. No namespace is searched more than once...
Paragraph 7 of 3.4.5 [basic.lookup.classref] says,
If the id-expression is a conversion-function-id, its conversion-type-id shall denote the same type in both the context in which the entire postfix-expression occurs and in the context of the class of the object expression (or the class pointed to by the pointer expression).Does this mean that the following example is ill-formed?
struct A { operator int(); } a;
void foo() {
typedef int T;
a.operator T(); // 1) error T is not found in the context
// of the class of the object expression?
}
The second bullet in paragraph 1 of
3.4.3.1 [class.qual]
says,
a conversion-type-id of an operator-function-id is looked up both in the scope of the class and in the context in which the entire postfix-expression occurs and shall refer to the same type in both contextsHow about:
struct A { typedef int T; operator T(); };
struct B : A { operator T(); } b;
void foo() {
b.A::operator T(); // 2) error T is not found in the context
// of the postfix-expression?
}
Is this interpretation correct? Or was the intent for
this to be an error only if
T was found in both scopes and referred to different entities?
If the intent was for these to be errors, how do these rules apply to template arguments?
template <class T1> struct A { operator T1(); }
template <class T2> struct B : A<T2> {
operator T2();
void foo() {
T2 a = A<T2>::operator T2(); // 3) error? when instantiated T2 is not
// found in the scope of the class
T2 b = ((A<T2>*)this)->operator T2(); // 4) error when instantiated?
}
}
(Note bullets 2 and 3 in paragraph 1 of 3.4.3.1 [class.qual] refer to postfix-expression. It would be better to use qualified-id in both cases.)
Erwin Unruh: The intent was that you look in both contexts. If you find it only once, that's the symbol. If you find it in both, both symbols must be "the same" in some respect. (If you don't find it, its an error).
Mike Miller: What's not clear to me in these examples is whether what is being looked up is T or int. Clearly the T has to be looked up somehow, but the "name" of a conversion function clearly involves the base (non-typedefed) type, not typedefs that might be used in a definition or reference (cf 3 [basic] paragraph 7 and 12.3 [class.conv] paragraph 5). (This is true even for types that must be written using typedefs because of the limited syntax in conversion-type-ids — e.g., the "name" of the conversion function in the following example
typedef void (*pf)();
struct S {
operator pf();
};
is S::operator void(*)(), even though you can't write its name
directly.)
My guess is that this means that in each scope you look up the type named in the reference and form the canonical operator name; if the name used in the reference isn't found in one or the other scope, the canonical name constructed from the other scope is used. These names must be identical, and the conversion-type-id in the canonical operator name must not denote different types in the two scopes (i.e., the type might not be found in one or the other scope, but if it's found in both, they must be the same type).
I think this is all very vague in the current wording.
3.4.5 [basic.lookup.classref] does not mention template aliases as the possible result of the lookup but should do so.
An example in 3.5 [basic.link] paragraph 6 creates two file-scope variables with the same name, one with internal linkage and one with external.
static void f();
static int i = 0; //1
void g() {
extern void f(); // internal linkage
int i; //2: i has no linkage
{
extern void f(); // internal linkage
extern int i; //3: external linkage
}
}
Is this really what we want? C99 has 6.2.2.7/7, which gives undefined behavior for having an identifier appear with internal and external linkage in the same translation unit. C++ doesn't seem to have an equivalent.
Notes from October 2003 meeting:
We agree that this is an error. We propose to leave the example but change the comment to indicate that line //3 has undefined behavior, and elsewhere add a normative rule giving such a case undefined behavior.
Proposed resolution (October, 2005):
Change 3.5 [basic.link] paragraph 6 as indicated:
...Otherwise, if no matching entity is found, the block scope entity receives external linkage. If, within a translation unit, the same entity is declared with both internal and external linkage, the behavior is undefined.
[Example:
static void f(); static int i = 0; // 1 void g () { extern void f (); // internal linkage int i; // 2: i has no linkage { extern void f (); // internal linkage extern int i; // 3: external linkage } }There are three objects named i in this program. The object with internal linkage introduced by the declaration in global scope (line //1 ), the object with automatic storage duration and no linkage introduced by the declaration on line //2, and the object with static storage duration and external linkage introduced by the declaration on line //3. Without the declaration at line //2, the declaration at line //3 would link with the declaration at line //1. But because the declaration with internal linkage is hidden, //3 is given external linkage, resulting in a linkage conflict. —end example]
Notes frum the April 2006 meeting:
According to 3.5 [basic.link] paragraph 9, the two variables with linkage in the proposed example are not “the same entity” because they do not have the same linkage. Some other formulation will be needed to describe the relationship between those two variables.
Notes from the October 2006 meeting:
The CWG decided that it would be better to make a program with this kind of linkage mismatch ill-formed instead of having undefined behavior.
The recent changes to allow use of unnamed types as template arguments require some rethinking of how unnamed types are treated in general. At least, a class-scope unnamed type should have the same linkage as its containing class. For example:
// File "hdr.h"
struct S {
static enum { No, Yes } locked;
};
template<class T> void f(T);
// File "impl1.c"
#include "hdr.h"
template void f(decltype(S::locked));
// File "impl2.c"
#include "hdr.h"
template void f(decltype(S::locked));
The two explicit instantiation directives should refer to the same specialization.
Sent in by David Abrahams:
Yes, and to add to this tangent, 3.9.1 [basic.fundamental] paragraph 1 states "Plain char, signed char, and unsigned char are three distinct types." Strangely, 3.9 [basic.types] paragraph 2 talks about how "... the underlying bytes making up the object can be copied into an array of char or unsigned char. 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." I guess there's no requirement that this copying work properly with signed chars!
Notes from October 2002 meeting:
We should do whatever C99 does. 6.5p6 of the C99 standard says "array of character type", and "character type" includes signed char (6.2.5p15), and 6.5p7 says "character type". But see also 6.2.6.1p4, which mentions (only) an array of unsigned char.
Proposed resolution (April 2003):
Change 3.8 [basic.life] paragraph 5 bullet 3 from
to
Change 3.8 [basic.life] paragraph 6 bullet 3 from
to
Change the beginning of 3.9 [basic.types] paragraph 2 from
For any object (other than a base-class subobject) of POD 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.
to
For any object (other than a base-class subobject) of POD 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 byte-character type.
Add the indicated text to 3.9.1 [basic.fundamental] paragraph 1:
Objects declared as characters (char) shall be large enough to store any member of the implementation's basic character set. If a character from this set is stored in a character object, the integral value of that character object is equal to the value of the single character literal form of that character. It is implementation-defined whether a char object can hold negative values. Characters can be explicitly declared unsigned or signed. Plain char, signed char, and unsigned char are three distinct types, called the byte-character types. A char, a signed char, and an unsigned char occupy the same amount of storage and have the same alignment requirements (3.9 [basic.types]); that is, they have the same object representation. For byte-character types, all bits of the object representation participate in the value representation. For unsigned byte-character types, all possible bit patterns of the value representation represent numbers. These requirements do not hold for other types. In any particular implementation, a plain char object can take on either the same values as a signed char or an unsigned char; which one is implementation-defined.
Change 3.10 [basic.lval] paragraph 15 last bullet from
to
Notes from October 2003 meeting:
It appears that in C99 signed char may have padding bits but no trap representation, whereas in C++ signed char has no padding bits but may have -0. A memcpy in C++ would have to copy the array preserving the actual representation and not just the value.
March 2004: The liaisons to the C committee have been asked to tell us whether this change would introduce any unnecessary incompatibilities with C.
Notes from October 2004 meeting:
The C99 Standard appears to be inconsistent in its requirements. For example, 6.2.6.1 paragraph 4 says:
The value may be copied into an object of type unsigned char [n] (e.g., by memcpy); the resulting set of bytes is called the object representation of the value.
On the other hand, 6.2 paragraph 6 says,
If a value is copied into an object having no declared type using memcpy or memmove, or is copied as an array of character type, then the effective type of the modified object for that access and for subsequent accesses that do not modify the value is the effective type of the object from which the value is copied, if it has one.
Mike Miller will investigate further.
Is the following example well-formed?
struct S {
static char a[5];
};
char S::a[]; // Unspecified bound in definition
3.5 [basic.link] paragraph 10 certainly makes allowance for declarations to differ in the presence or absence of a major array bound. However, 3.1 [basic.def] paragraph 5 says that
A program is ill-formed if the definition of any object gives the object an incomplete type (3.9 [basic.types]).
3.9 [basic.types] paragraph 7 says,
The declared type of an array object might be an array of unknown size and therefore be incomplete at one point in a translation unit and complete later on; the array types at those two points (“array of unknown bound of T” and “array of N T”) are different types.
This wording appears to make no allowance for the C concept of “composite type;” instead, each declaration is said to have its own type. By this interpretation, the example is ill-formed, because the type declared by the definition of S::a is incomplete.
If the example is intended to be well-formed, the Standard needs explicit wording stating that an omitted array bound in a declaration is implicitly taken from that of a visible declaration of that object, if any.
Notes from the April, 2007 meeting:
The CWG agreed that this usage should be permitted.
Proposed resolution (June, 2008):
Change 8.3.4 [dcl.array] paragraph 1 as follows:
...If Except as noted below, if the constant expression is omitted, the type of the identifier of D is “derived-declarator-type-list array of unknown bound of T,” an incomplete object type...
Change 8.3.4 [dcl.array] paragraph 3 as follows:
When several “array of” specifications are adjacent, a multidimensional array is created; only the first of the constant expressions that specify the bounds of the arrays can may be omitted only for the first member of the sequence. [Note: this elision is useful for function parameters of array types, and when the array is external and the definition, which allocates storage, is given elsewhere. —end note] In addition to declarations in which an incomplete object type is allowed, an array bound may be omitted in the declaration of a function parameter (8.3.5 [dcl.fct]). The first constant-expression can An array bound may also be omitted when the declarator is followed by an initializer (8.5 [dcl.init]). In this case the bound is calculated from the number of initial elements (say, N) supplied (8.5.1 [dcl.init.aggr]), and the type of the identifier of D is “array of N T.” Furthermore, if there is a visible declaration of the name declared by the declarator-id (if any) in which the bound was specified, an omitted array bound is taken to be the same as in that earlier declaration.
Notes from the September, 2008 meeting:
The proposed resolution does not capture the result favored by the CWG: array bound information should be accumulated across declarations within the same scope, but a block extern declaration in a nested scope should not inherit array bound information from the outer declaration. (This is consistent with the treatment of default arguments in function declarations.) For example:
int a[5];
void f() {
extern int a[];
sizeof(a);
}
Although there was some confusion about the C99 wording dealing with this case, it is probably well-formed in C99. However, it should be ill-formed in C++, because we want to avoid the concept of “compatible types” as it exists in C.
The aliasing rules given in 3.10 [basic.lval] paragraph 10 rely on the concept of “dynamic type.” The problem is that the dynamic type of an object often cannot be determined (or even sufficiently constrained) at the point at which an optimizer needs to be able to determine whether aliasing might occur or not. For example, consider the function
void foo(int* p, double* q) {
*p = 42;
*q = 3.14;
}
An optimizer, on the basis of the existing aliasing rules, might decide that an int* and a double* cannot refer to the same object and reorder the assignments. This reordering, however, could result in undefined behavior if the function foo is called as follows:
void goo() {
union {
int i;
double d;
} t;
t.i = 12;
foo(&t.i, &t.d);
cout << t.d << endl;
};
Here, the reference to t.d after the call to foo will be valid only if the assignments in foo are executed in the order in which they were written; otherwise, the union will contain an int object rather than a double.
One possibility would be to require that if such aliasing occurs, it be done only via member names and not via pointers.
Notes from the July, 2007 meeting:
This is the same issue as C's DR236. The CWG expressed a desire to address the issue the same way C99 does. The issue also occurs in C++ when placement new is used to end the lifetime of one object and start the lifetime of a different object occupying the same storage.
According to 4.1 [conv.lval] paragraph 1, applying the lvalue-to-rvalue conversion to any uninitialized object results in undefined behavior. However, character types are intended to allow any data, including uninitialized objects and padding, to be copied (hence the statements in 3.9.1 [basic.fundamental] paragraph 1 that “For character types, all bits of the object representation participate in the value representation” and in 3.10 [basic.lval] paragraph 15 that char and unsigned char types can alias any object). The lvalue-to-rvalue conversion should be permitted on uninitialized objects of character type without evoking undefined behavior.
The descriptions of explicit (5.2.9 [expr.static.cast] paragraph 9) and implicit (4.11 [conv.mem] paragraph 2) pointer-to-member conversions differ in two significant ways:
(This situation cannot arise in an implicit pointer-to-member conversion where the source value is something like &X::f, since you can only implicitly convert from pointer-to-base-member to pointer-to-derived-member. However, if the source value is the result of an explicit "up-cast," the target type of the conversion might still not contain the member referred to by the source value.)
The first difference seems like an oversight. It is not clear whether the latter difference is intentional or not.
(See also issue 794.)
There are at least a couple of problems in the description of the various id-expressions in 5.1.1 [expr.prim.general]:
Paragraph 4 embodies an incorrect assumption about the syntax of qualified-ids:
The operator :: followed by an identifier, a qualified-id, or an operator-function-id is a primary-expression.
The problem here is that the :: is actually part of the syntax of qualified-id; consequently, “:: followed by... a qualified-id” could be something like “:: ::i,” which is ill-formed. Presumably this should say something like, “A qualified-id with no nested-name-specifier is a primary-expression.”
More importantly, some kinds of id-expressions are not described by 5.1.1 [expr.prim.general]. The structure of this section is that the result, type, and lvalue-ness are specified for each of the cases it covers:
paragraph 4 deals with qualified-ids that have no nested-name-specifier
paragraph 7 deals with bare identifiers and with qualified-ids containing a nested-name-specifier that names a class
paragraph 8 deals with qualified-ids containing a nested-name-specifier that names a namespace
This treatment leaves unspecified all the non-identifier unqualified-ids (operator-function-id, conversion-function-id, and template-id), as well as (perhaps) “:: template-id” (it's not clear whether the “:: followed by a qualified-id” case is supposed to apply to template-ids or not). Note also that the proposed resolution of issue 301 slightly exacerbates this problem by removing the form of operator-function-id that contains a tmeplate-argument-list; as a result, references like “::operator+<X>” are no longer covered in 5.1.1 [expr.prim.general].
this is a keyword and thus not subject to ordinary name lookup. That makes the interpretation of examples like the following somewhat unclear:
struct outer {
void f() {
struct inner {
int a[sizeof(*this)]; // #1
};
}
};
According to 5.1.1 [expr.prim.general] paragraph 3,
The keyword this shall be used only inside a non-static class member function body (9.3 [class.mfct]) or in a brace-or-equal-initializer for a non-static data member.
Should the use of this at #1 be interepreted as a well-formed reference to outer::f()'s this or as an ill-formed attempt to refer to a this for outer::inner?
One possible interpretation is that the intent is as if this were an ordinary identifier appearing as a parameter in each non-static member function. (This view applies to the initializers of non-static data members as well if they are considered to be rewritten as mem-initializers in the constructor body.) Under this interpretation, the prohibition against using this in other contexts simply falls out of the fact that name lookup would fail to find this anywhere else, so the reference in the example is well-formed. (Implementations vary in their treatment of this example, so clearer wording is needed, whichever way the interpretation goes.)
The current wording of 5.2.2 [expr.call] paragraph 7 is:
After these conversions, if the argument does not have arithmetic, enumeration, pointer, pointer to member, or effective class type, the program is ill-formed.
It's not clear whether this is intended to exclude anything other than void, but the effect is to disallow passing nullptr to ellipsis. That seems unnecessary.
Notes from the October, 2009 meeting:
The CWG agreed that this should be supported and the effect should be like passing (void*)nullptr.
The resolution to issue 195 makes “converting a pointer to a function into a pointer to an object type or vice versa” conditionally-supported behavior. In doing so, however, it overlooked the fact that void is not an “object type” (3.9 [basic.types] paragraph 9). The wording should be amended to allow conversion to and from void* types.
Proposed resolution (June, 2008):
Change 3.9.2 [basic.compound] paragraph 4 as follows:
Objects of cv-qualified (3.9.3 [basic.type.qualifier]) or cv-unqualified type void* (pointer to void) A pointer to cv-qualified or cv-unqualified void can be used to point to objects of unknown type. A void* shall be able to hold any object pointer and is thus considered to be an object pointer type, although it is not a pointer to object type (because void is not an object type). A cv-qualified or cv-unqualified (3.9.3 [basic.type.qualifier]) An object of type cv void* shall have the same representation and alignment requirements as a cv-qualified or cv-unqualified cv char*.
Change 4.10 [conv.ptr] paragraph 1 as follows:
...A null pointer constant can be converted to a pointer type; the result is the null pointer value of that type and is distinguishable from every other value of pointer to object or pointer to function object pointer or function pointer type...
Change 5.2.10 [expr.reinterpret.cast] paragraph 7 as follows:
A pointer to an object An object pointer can be explicitly converted to a pointer to an object an object pointer of different type. Except that converting an rvalue of type “pointer to T1” to the type “pointer to T2” (where T1 and T2 are object types or void and where the alignment requirements of T2 are no stricter than those of T1) and back to its original type yields the original pointer value, the result of such a pointer conversion is unspecified.
Change 5.2.10 [expr.reinterpret.cast] paragraph 8 as follows:
Converting a pointer to a function into a pointer to an object type a function pointer to an object pointer or vice versa is conditionally-supported...
[Drafting note: 14.2 [temp.param] paragraph 4 was not changed and thus continues to allow only pointers to objects, not object pointers, as non-type template parameters.]
Consider the following example:
static const char test1 = 'x';
static const char test2 = 'x';
bool f() {
return &test1 != &test2;
}
Is f() allowed to return false? Can a smart optimizer alias these two variables, taking advantage of the fact that they are const, initialized to the same value, and thus can never be different in a well-defined program?
The C++ Standard doesn't explicitly specify address allocation of objects except as members of arrays and classes, so the answer would appear to be that such an implementation would be conforming.
This situation appears to have been the inadvertent result of the resolution of issue 73. Prior to that change, 5.10 [expr.eq] said,
Two pointers of the same type compare equal if and only if they... both point to the same object...
That resolution introduced the current wording,
Notes from the March, 2009 meeting:
The CWG agreed that this aliasing should not be permitted in a conforming implementation.
Two pointers of the same type compare equal if and only if... both represent the same address.
5.2.11 [expr.const.cast] paragraph 4 says,
...Similarly, for two effective object types T1 and T2, an expression of type T1 can be explicitly converted to an rvalue of type T2 using the cast const_cast<T2&&> if a pointer to T1 can be explicitly converted to the type “pointer to T2” using a const_cast. The result of a reference const_cast refers to the original object.
However, in some rvalue-reference const_casts there is no “original object,” e.g.,
const_cast<int&&>(2)
Notes from the July, 2009 meeting:
The coresponding cast to an lvalue reference to const is ill-formed: in such cases, the operand must be an lvalue. The consensus of the CWG was that a cast to an rvalue reference should only accept an rvalue that is an rvalue reference (i.e., an object).
At least a couple of places in the IS state that indirection through a null pointer produces undefined behavior: 1.9 [intro.execution] paragraph 4 gives "dereferencing the null pointer" as an example of undefined behavior, and 8.3.2 [dcl.ref] paragraph 4 (in a note) uses this supposedly undefined behavior as justification for the nonexistence of "null references."
However, 5.3.1 [expr.unary.op] paragraph 1, which describes the unary "*" operator, does not say that the behavior is undefined if the operand is a null pointer, as one might expect. Furthermore, at least one passage gives dereferencing a null pointer well-defined behavior: 5.2.8 [expr.typeid] paragraph 2 says
If the lvalue expression is obtained by applying the unary * operator to a pointer and the pointer is a null pointer value (4.10 [conv.ptr]), the typeid expression throws the bad_typeid exception (18.7.4 [bad.typeid]).
This is inconsistent and should be cleaned up.
Bill Gibbons:
At one point we agreed that dereferencing a null pointer was not undefined; only using the resulting value had undefined behavior.
For example:
char *p = 0;
char *q = &*p;
Similarly, dereferencing a pointer to the end of an array should be allowed as long as the value is not used:
char a[10];
char *b = &a[10]; // equivalent to "char *b = &*(a+10);"
Both cases come up often enough in real code that they should be allowed.
Mike Miller:
I can see the value in this, but it doesn't seem to be well reflected in the wording of the Standard. For instance, presumably *p above would have to be an lvalue in order to be the operand of "&", but the definition of "lvalue" in 3.10 [basic.lval] paragraph 2 says that "an lvalue refers to an object." What's the object in *p? If we were to allow this, we would need to augment the definition to include the result of dereferencing null and one-past-the-end-of-array.
Tom Plum:
Just to add one more recollection of the intent: I was very happy when (I thought) we decided that it was only the attempt to actually fetch a value that creates undefined behavior. The words which (I thought) were intended to clarify that are the first three sentences of the lvalue-to-rvalue conversion, 4.1 [conv.lval]:
An lvalue (3.10 [basic.lval]) of a non-function, non-array type T can be converted to an rvalue. If T is an incomplete type, a program that necessitates this conversion is ill-formed. If the object to which the lvalue refers is not an object of type T and is not an object of a type derived from T, or if the object is uninitialized, a program that necessitates this conversion has undefined behavior.
In other words, it is only the act of "fetching", of lvalue-to-rvalue conversion, that triggers the ill-formed or undefined behavior. Simply forming the lvalue expression, and then for example taking its address, does not trigger either of those errors. I described this approach to WG14 and it may have been incorporated into C 1999.
Mike Miller:
If we admit the possibility of null lvalues, as Tom is suggesting here, that significantly undercuts the rationale for prohibiting "null references" -- what is a reference, after all, but a named lvalue? If it's okay to create a null lvalue, as long as I don't invoke the lvalue-to-rvalue conversion on it, why shouldn't I be able to capture that null lvalue as a reference, with the same restrictions on its use?
I am not arguing in favor of null references. I don't want them in the language. What I am saying is that we need to think carefully about adopting the permissive approach of saying that it's all right to create null lvalues, as long as you don't use them in certain ways. If we do that, it will be very natural for people to question why they can't pass such an lvalue to a function, as long as the function doesn't do anything that is not permitted on a null lvalue.
If we want to allow &*(p=0), maybe we should change the definition of "&" to handle dereferenced null specially, just as typeid has special handling, rather than changing the definition of lvalue to include dereferenced nulls, and similarly for the array_end+1 case. It's not as general, but I think it might cause us fewer problems in the long run.
Notes from the October 2003 meeting:
See also issue 315, which deals with the call of a static member function through a null pointer.
We agreed that the approach in the standard seems okay: p = 0; *p; is not inherently an error. An lvalue-to-rvalue conversion would give it undefined behavior.
Proposed resolution (October, 2004):
(Note: the resolution of issue 453 also resolves part of this issue.)
Add the indicated words to 3.10 [basic.lval] paragraph 2:
An lvalue refers to an object or function or is an empty lvalue (5.3.1 [expr.unary.op]).
Add the indicated words to 5.3.1 [expr.unary.op] paragraph 1:
The unary * operator performs indirection: the expression to which it is applied shall be a pointer to an object type, or a pointer to a function type and the result is an lvalue referring to the object or function to which the expression points, if any. If the pointer is a null pointer value (4.10 [conv.ptr]) or points one past the last element of an array object (5.7 [expr.add]), the result is an empty lvalue and does not refer to any object or function. An empty lvalue is not modifiable. If the type of the expression is “pointer to T,” the type of the result is “T.” [Note: a pointer to an incomplete type (other than cv void) can be dereferenced. The lvalue thus obtained can be used in limited ways (to initialize a reference, for example); this lvalue must not be converted to an rvalue, see 4.1 [conv.lval].—end note]
Add the indicated words to 4.1 [conv.lval] paragraph 1:
If the object to which the lvalue refers is not an object of type T and is not an object of a type derived from T, or if the object is uninitialized, or if the lvalue is an empty lvalue (5.3.1 [expr.unary.op]), a program that necessitates this conversion has undefined behavior.
Change 1.9 [intro.execution] as indicated:
Certain other operations are described in this International Standard as undefined (for example, the effect of dereferencing the null pointer division by zero).
Note (March, 2005):
The 10/2004 resolution interacts with the resolution of issue 73. We added wording to 3.9.2 [basic.compound] paragraph 3 to the effect that a pointer containing the address one past the end of an array is considered to “point to” another object of the same type that might be located there. The 10/2004 resolution now says that it would be undefined behavior to use such a pointer to fetch the value of that object. There is at least the appearance of conflict here; it may be all right, but it at needs to be discussed further.
Notes from the April, 2005 meeting:
The CWG agreed that there is no contradiction between this direction and the resolution of issue 73. However, “not modifiable” is a compile-time concept, while in fact this deals with runtime values and thus should produce undefined behavior instead. Also, there are other contexts in which lvalues can occur, such as the left operand of . or .*, which should also be restricted. Additional drafting is required.
It is not clear from 5.3.4 [expr.new] whether a deleted operator delete is referenced by a new-expression in which there is no initialization or in which the initialization cannot throw an exception, rendering the program ill-formed. (The question also arises as to whether such a new-expression constitutes a “use” of the deallocation function in the sense of 3.2 [basic.def.odr].)
Notes from the July, 2009 meeting:
The rationale for defining a deallocation function as deleted would presumably be to prevent such objects from being freed. Treating the new-expression as a use of such a deallocation function would mean that such objects could not be created in the first place. There is already an exemption from freeing an object if “a suitable deallocation function [cannot] be found;” a deleted deallocation function should be treated similarly.
The grammar for condition in 6.4 [stmt.select] paragraph 1 does not allow for the constexpr specifier. This was not intended by the original proposal.
The intent is that the range-based for statement should be able to be used with a braced-init-list as the range over which to iterate. However, this does not work grammatically: a braced-init-list is not an expression, as required by the syntax in 6.5.4 [stmt.ranged] paragraph 1:
Even if this were resolved, the “equivalent to” code is not correct. It contains the declaration,
This has a similar problem, in that 7.1.6.4 [dcl.spec.auto] paragraph 3 requires that the initializer have one of the forms
which does not allow for a braced-initializer-list. In addition, although not allowed by the grammar, 7.1.6.4 [dcl.spec.auto] paragraph 6 treats the braced-init-list specially, in order for the type deduction to work correctly:
Obtain P from T by replacing the occurrences of auto with either a new invented type template parameter U or, if the initializer is a braced-init-list (8.5.4 [dcl.init.list]), with std::initializer_list<U>.
The problem here is that a parenthesized initializer, as in the code expansion of the range-based for statement, is not a braced-init-list.
According to 7.1 [dcl.spec] paragraph 2,
The longest sequence of decl-specifiers that could possibly be a type name is taken as the decl-specifier-seq of a declaration.
However, there are many decl-specifiers that cannot appear in a type name that are, nonetheless, part of a declaration's decl-specifier-seq, such as typedef, friend, static, etc.
7.1.2 [dcl.fct.spec] paragraph 4 specifies that local static variables and string literals appearing in the body of an inline function with external linkage must be the same entities in every translation unit in the program. Nothing is said, however, about whether local types are likewise required to be the same.
Although a conforming program could always have determined this by use of typeid, recent changes to C++ (allowing local types as template type arguments, lambda expression closure classes) make this question more pressing.
Notes from the July, 2009 meeting:
The types are intended to be the same.
Here's an example:
typedef struct S { ... } S;
void fs(S *x) { ... }
The big question is, to what declaration does the reference to identifier S actually refer? Is it the S that's declared as a typedef name, or the S that's declared as a class name (or in C terms, as a struct tag)? (In either case, there's clearly only one type to which it could refer, since a typedef declaration does not introduce a new type. But the debugger apparently cares about more than just the identity of the type.)
Here's a classical, closely related example:
struct stat { ... };
int stat();
... stat( ... ) ...
Does the identifier stat refer to the class or the function? Obviously, in C, you can't refer to the struct tag without using the struct keyword, because it is in a different name space, so the reference must be to the function. In C++, the reference is also to the function, but for a completely different reason.
Now in C, typedef names and function names are in the same name space, so the natural extrapolation would be that, in the first example, S refers to the typedef declaration, as it would in C. But C++ is not C. For the purposes of this discussion, there are two important differences between C and C++
The first difference is that, in C++, typedef names and class names are not in separate name spaces. On the other hand, according to section 3.3.11 [basic.scope.hiding] (Name hiding), paragraph 2:
A class name (9.1) or enumeration name (7.2) can be hidden by the name of an object, function, or enumerator declared in the same scope. If a class or enumeration name and an object, function, or enumerator are declared in the same scope (in any order) with the same name, the class or enumeration name is hidden wherever the object, function, or enumerator name is visible.
Please consider carefully the phrase I have highlighted, and the fact that a typedef name is not the name of an object, function or enumerator. As a result, this example:
struct stat { ... };
typedef int stat;
Which would be perfectly legal in C, is disallowed in C++, both implicitly (see the above quote) and explicitly (see section 7.1.3 [dcl.typedef] (The typedef specifier), paragraph 3):
In a given scope, a typedef specifier shall not be used to redefine the name of any type declared in that scope to refer to a different type. Similarly, in a given scope, a class or enumeration shall not be declared with the same name as a typedef-name that is declared in that scope and refers to a type other than the class or enumeration itself.
From which we can conclude that in C++ typedef names do not hide class names declared in the same scope. If they did, the above example would be legal.
The second difference is that, in C++, a typedef name that refers to a class is a class-name; see 7.1.3 [dcl.typedef] paragraph 4:
A typedef-name that names a class is a class-name(9.1). If a typedef-name is used following the class-key in an elaborated-type-specifier (7.1.5.3) or in the class-head of a class declaration (9), or is used as the identifier in the declarator for a constructor or destructor declaration (12.1, 12.4), the program is ill-formed.
This implies, for instance, that a typedef-name referring to a class can be used in a nested-name-specifier (i.e. before :: in a qualified name) or following ~ to refer to a destructor. Note that using a typedef-name as a class-name in an elaborated-type-specifier is not allowed. For example:
struct X { };
typedef struct X X2;
X x; // legal
X2 x2; // legal
struct X sx; // legal
struct X2 sx2; // illegal
The final relevant piece of the standard is 7.1.3 [dcl.typedef] paragraph 2:
In a given scope, a typedef specifier can be used to redefine the name of any type declared in that scope to refer to the type to which it already refers.
This of course is what allows the original example, to which let us now return:
typedef struct S { ... } S;
void fs(S *x) { ... }
The question, again is, to which declaration of S does the reference actually refer? In C, it would clearly be to the second, since the first would be accessible only by using the struct keyword. In C++, if typedef names hid class names declared in the same scope, the answer would be the same. But we've already seen that typedef names do not hide class names declared in the same scope.
So to which declaration does the reference to S refer? The answer is that it doesn't matter. The second declaration of S, which appears to be a declaration of a typedef name, is actually a declaration of a class name (7.1.3 [dcl.typedef] paragraph 4), and as such is simply a redeclaration. Consider the following example:
typedef int I, I; extern int x, x; void f(), f();
To which declaration would a reference to I, x or f refer? It doesn't matter, because the second declaration of each is really just a redeclaration of the thing declared in the first declaration. So to save time, effort and complexity, the second declaration of each doesn't add any entry to the compiler's symbol table.
Note (March, 2005):
Matt Austern: Is this legal?
struct A { };
typedef struct A A;
struct A* p;
Am I right in reading the standard [to say that this is ill-formed]? On the one hand it's a nice uniform rule. On the other hand, it seems likely to confuse users. Most people are probably used to thinking that 'typedef struct A A' is a null operation, and, if this code really is illegal, it would seem to be a gratuitous C/C++ incompatibility.
Mike Miller: I think you're right. 7.1.3 [dcl.typedef] paragraph 1:
A name declared with the typedef specifier becomes a typedef-name.
7.1.3 [dcl.typedef] paragraph 2:
In a given non-class scope, a typedef specifier can be used to redefine the name of any type declared in that scope to refer to the type to which it already refers.
After the typedef declaration in the example, the name X has been “redefined” — it is no longer just a class-name, it has been “redefined” to be a typedef-name (that, by virtue of the fact that it refers to a class type, is also a class-name).
John Spicer: In C, and originally in C++, an elaborated-type-specifier did not consider typedef names, so “struct X* x” would find the class and not the typedef.
When C++ was changed to make typedefs visible to elaborated-type-specifier lookups, I believe this issue was overlooked and inadvertantly made ill-formed.
I suspect we need add text saying that if a given scope contains both a class/enum and a typedef, that an elaborated type specifier lookup finds the class/enum.
Mike Miller: I'm a little uncomfortable with this approach. The model we have for declaring a typedef in the same scope as a class/enum is redefinition, not hiding (like the “struct stat” hack). This approach seems to assume that the typedef hides the class/enum, which can then be found by an elaborated-type-specifier, just as if it were hidden by a variable, function, or enumerator.
Also, this approach reduces but doesn't eliminate the incompatibility with C. For example:
struct S { };
{
typedef struct S S;
struct S* p; // still ill-formed
}
My preference would be for something following the basic principle that declaring a typedef-name T in a scope where T already names the type designated by the typedef should have no effect on whether an elaborated-type-specifier in that or a nested scope is well-formed or not. Another way of saying that is that a typedef-name that designates a same-named class or enumeration in the same or a containing scope is transparent with respect to elaborated-type-specifiers.
John Spicer: This strikes me as being a rather complicated solution. When we made the change to make typedefs visible to elaborated-type-specifiers we did so knowing it would make some C cases ill-formed, so this does not bother me. We've lived with the C incompatibility for many years now, so I don't personally feel a need to undo it. I also don't like the fact that you have to essentially do the old-style elaborated-type-specifier lookup to check the result of the lookup that found the typedef.
I continue to prefer the direction I described earlier where if a given scope contains both a class/enum and a typedef, that an elaborated-type-specifier lookup finds the class/enum.
Notes from the April, 2005 meeting:
The CWG agreed with John Spicer's approach, i.e., permitting a typedef-name to be used in an elaborated-type-specifier only if it is declared in the same scope as the class or enumeration it names.
7.1.5 [dcl.constexpr] paragraph 5 applies only to “the instantiated template specialization of a constexpr function template;” it should presumably apply to non-template member functions of a class template, as well.
Notes from the September, 2008 meeting:
This question is more involved than it might appear. For example, a constexpr member function is implicitly const; if the constexpr specifier is ignored, does that make the member function non-const? Also, should this provision apply only to dependent expressions in the function? Should it be an error if no constexpr function can be instantiated from the template, along the lines of the permission given in 14.7 [temp.res] paragraph 8 for an implementation to diagnose a template definition from which no valid specialization can be instantiated?
Notes from the July, 2009 meeting:
The consensus of the CWG was that an “ignored” constexpr specifier in this case simply means that the specialization is not constexpr, not that it is not const. The CWG also decided not to address the question of non-dependent expressions that render a function template specialization non-constexpr, leaving it to quality of implementation whether a (warning) diagnostic is issued in such cases.
The body of a constexpr function is required by 7.1.5 [dcl.constexpr] paragraph 3 to be of the form
However, there does not seem to be any good reason for prohibiting the alternate return syntax involving a braced-init-list. The restriction should be removed.
7.1.5 [dcl.constexpr] paragraph 6 says,
A constexpr specifier for a non-static member function that is not a constructor declares that member function to be const (9.3.1 [class.mfct.non-static]).
Is a const qualifier on such a member function redundant or ill-formed?
Notes from the July, 2009 meeting:
The CWG agreed that a const qualifier on a constexpr member function is simply redundant and not an error.
The rules for constexpr constructors are missing some necessary requirements. In particular, there is no requirement that a brace-or-equal-initializer for a non-static data member be a constant expression, and the requirement for constexpr constructors for initializing non-static data members applies only to members named in a mem-initializer, allowing a non-constexpr default constructor to be invoked.
It would be useful if constexpr functions and constructors could take
arguments via reference-to-const parameters.
In the current specification, a decltype resulting in a class type is not a class-name, meaning that it cannot be used as a base-specifier. There doesn't seem to be any reason not to allow that, and it would be consistent with the proposed outcome of issue 743.
It is not clear from the specification in 7.3.1 [namespace.def] paragraph 8 how a declaration in an inline namespace should be handled if the name is the same as one in the containing namespace or in an parallel inline namespace. For example:
namespace Q {
inline namespace V1 {
int i;
int j;
}
inline namespace V2 {
int j;
}
int i;
}
int Q::i = 1; // Q::i or Q::V1::i?
int Q::j = 2; // Q::V1::j or Q::V2::j?
Proposed resolution (July, 2009):
This issue is resolved by the resolution of issue 861.
According to 7.3.1 [namespace.def] paragraph 8,
Members of an inline namespace can be used in most respects as though they were members of the enclosing namespace... Furthermore, each member of the inline namespace can subsequently be explicitly instantiated (14.8.2 [temp.explicit]) or explicitly specialized (14.8.3 [temp.expl.spec]) as though it were a member of the enclosing namespace.
However, that assertion is contradicted for class template specializations by 9 [class] paragraph 11:
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...
It is also contradicted for function template specializations by 3.4.3.2 [namespace.qual] paragraph 6:
In a declaration for a namespace member in which the declarator-id is a qualified-id, given that the qualified-id for the namespace member has the formnested-name-specifier unqualified-id
the unqualified-id shall name a member of the namespace designated by the nested-name-specifier.
Proposed resolution (July, 2009):
Change 9 [class] paragraph 11 as follows:
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, or in an element of the inline namespace set (7.3.1 [namespace.def]) of that namespace (i.e., neither not merely inherited nor or introduced by a using-declaration), and the class-specifier shall appear in a namespace enclosing the previous declaration.
Change 3.4.3.2 [namespace.qual] paragraph 6 as follows:
In a declaration for a namespace member in which the declarator-id is a qualified-id, given that the qualified-id for the namespace member has the form
nested-name-specifier unqualified-id
the unqualified-id shall name a member of the namespace designated by the nested-name-specifier, or of an element of the inline namespace (7.3.1 [namespace.def]) of that namespace. [Example:...
(Note: this resolution depends on the resolution for issue 861.)
7.3.1.2 [namespace.memdef] paragraph 3 says,
If a friend declaration in a non-local class first declares a class or function the friend class or function is a member of the innermost enclosing namespace... When looking for a prior declaration of a class or a function declared as a friend, scopes outside the innermost enclosing namespace scope are not considered.It is not clear from this passage how to determine whether an entity is "first declared" in a friend declaration. One question is whether a using-declaration influences this determination. For instance:
void foo();
namespace A{
using ::foo;
class X{
friend void foo();
};
}
Is the friend declaration a reference to ::foo or
a different foo?
Part of the question involves determining the meaning of the word "synonym" in 7.3.3 [namespace.udecl] paragraph 1:
A using-declaration introduces a name into the declarative region in which the using-declaration appears. That name is a synonym for the name of some entity declared elsewhere.Is "using ::foo;" the declaration of a function or not?
More generally, the question is how to describe the lookup of the name in a friend declaration.
John Spicer: When a declaration specifies an unqualified name, that name is declared, not looked up. There is a mechanism in which that declaration is linked to a prior declaration, but that mechanism is not, in my opinion, via normal name lookup. So, the friend always declares a member of the nearest namespace scope regardless of how that name may or may not already be declared there.
Mike Miller: 3.4.1 [basic.lookup.unqual] paragraph 7 says:
A name used in the definition of a class X outside of a member function body or nested class definition shall be declared in one of the following ways:... [Note: when looking for a prior declaration of a class or function introduced by a friend declaration, scopes outside of the innermost enclosing namespace scope are not considered.]The presence of this note certainly implies that this paragraph describes the lookup of names in friend declarations.
John Spicer: It most certainly does not. If that section described the friend lookup it would yield the incorrect results for the friend declarations of f and g below. I don't know why that note is there, but it can't be taken to mean that that is how the friend lookup is done.
void f(){}
void g(){}
class B {
void g();
};
class A : public B {
void f();
friend void f(); // ::f not A::f
friend void g(); // ::g not B::g
};
Mike Miller: If so, the lookups for friend functions and classes behave differently. Consider the example in 3.4.4 [basic.lookup.elab] paragraph 3:
struct Base {
struct Data; // OK: declares nested Data
friend class Data; // OK: nested Data is a friend
};
If the friend declaration is not a reference to ::foo, there is a related but separate question: does the friend declaration introduce a conflicting (albeit "invisible") declaration into namespace A, or is it simply a reference to an as-yet undeclared (and, in this instance, undeclarable) A::foo? Another part of the example in 3.4.4 [basic.lookup.elab] paragraph 3 is related:
struct Data {
friend struct Glob; // OK: Refers to (as yet) undeclared Glob
// at global scope.
};
John Spicer: You can't refer to something that has not yet been declared. The friend is a declaration of Glob, it just happens to declare it in a such a way that its name cannot be used until it is redeclared.
(A somewhat similar question has been raised in connection with issue 36. Consider:
namespace N {
struct S { };
}
using N::S;
struct S; // legal?
According to 9.1 [class.name] paragraph 2,
A declaration consisting solely of class-key identifier ; is either a redeclaration of the name in the current scope or a forward declaration of the identifier as a class name.
Should the elaborated type declaration in this example be considered a redeclaration of N::S or an invalid forward declaration of a different class?)
(See also issues 95, 136, 139, 143, 165, and 166, as well as paper J16/00-0006 = WG21 N1229.)
Here's an interesting case:
int f;
namespace N {
extern "C" void f () {}
}
As far as I can tell, this is not precluded by the ODR section
(3.2 [basic.def.odr])
or the extern "C" section (7.5 [dcl.link]).
However, I believe many compilers
do not do name mangling on variables and (more-or-less by definition)
on extern "C" functions. That means the variable and the function
in the above end up having the same name at link time. EDG's front
end, g++, and the Sun compiler all get essentially the same error,
which is a compile-time assembler-level error because of the
duplicate symbols (in other words, they fail to check for this, and the
assembler complains). MSVC++ 7 links the program without error,
though I'm not sure how it is interpreted.
Do we intend for this case to be valid? If not, is it a compile time error (required), or some sort of ODR violation (no diagnostic required)? If we do intend for it to be valid, are we forcing many implementations to break binary compatibility by requiring them to mangle variable names?
Personally, I favor a compile-time error, and an ODR prohibition on such things in separate translation units.
Notes from the 4/02 meeting:
The working group agreed with the proposal. We feel a diagnostic should be required for declarations within one translation unit. We also noted that if the variable in global scope in the above example were declared static we would still expect an error.
Relevant sections in the standard are 7.5 [dcl.link] paragraph 6 and 3.5 [basic.link] paragraph 9. We feel that the definition should be written such that the entities in conflict are not "the same entity" but merely not allowed together.
Additional note (September, 2004)
This problem need not involve a conflict between a function and a variable; it can also arise with two variable declarations:
int x;
namespace N {
extern "C" int x;
}
Proposed resolution (March, 2008):
Change 7.5 [dcl.link] paragraph 6 as follows:
At most one function with a particular name can have C language linkage. Two declarations for a function with C language linkage with the same function name (ignoring the namespace names that qualify it) that appear in different namespace scopes refer to the same function. Two declarations for an object with C language linkage with the same name (ignoring the namespace names that qualify it) that appear in different namespace scopes refer to the same object. A function or object with C linkage shall not be declared with the same name (clause 3 [basic]) as an object or reference declared in global scope, unless both declarations denote the same object; no diagnostic is required if the declarations appear in different translation units. [Note: because of the one definition rule (3.2 [basic.def.odr]), only Only one definition for a function or object with C linkage may appear in the program (see 3.2 [basic.def.odr]); that is, implies that such a function or object must not be defined in more than one namespace scope. For example,
int x; namespace A { extern "C" int f(); extern "C" int g() { return 1; } extern "C" int h(); extern "C" int x(); // ill-formed: same name as global-scope object x } namespace B { extern "C" int f(); // A::f and B::f refer // to the same function extern "C" int g() { return 1; } // ill-formed, the function g // with C language linkage // has two definitions } int A::f() { return 98; } // definition for the function f // with C language linkage extern "C" int h() { return 97; } // definition for the function h // with C language linkage // A::h and ::h refer to the same function—end note]
Notes from the September, 2008 meeting:
It should also be possible to declare references with C name linkage (although the meaning the first sentence of 7.5 [dcl.link] paragraph 1 with respect to the meaning of such a declaration is not clear), which would mean that the changed wording should refer to declaring “the same entity” instead of “the same object.” The formulation here would probably benefit from the approach currently envisioned for issues 570 and 633, in which “variable” is defined as being either an object or a reference.
There are a number of problems with the treatment of attributes in the current draft. One issue is the failure to permit attributes to appear at various points in the grammar at which one might plausibly expect them:
In a new-type-id (5.3.4 [expr.new])
Preceding the type-specifier-seq in a condition (6.4 [stmt.select])
In a for-init-statement that is an expression-statement (6.5 [stmt.iter])
Preceding the type-specifier-seq in a for-range-declaration (6.5 [stmt.iter])
In a reference ptr-operator (8 [dcl.decl])
Preceding the type-specifier-seq in a type-id (8.1 [dcl.name])
Preceding the decl-specifier-seq in a parameter-declaration (8.3.5 [dcl.fct])
In a function-definition (8.4 [dcl.fct.def]) at any of the three locations where they might be expected:
preceding the decl-specifier-seq
following the parameter list (paragraph 2 repeats the syntax from 8.3.5 [dcl.fct] with the conspicuous omission of the attribute-specifier)
preceding the compound-statement of the function-body (this would introduce an ambiguity with the attribute-specifier following the parameter list that would need to be addressed)
Preceding the decl-specifier-seq of a member-declaration (9.2 [class.mem])
Preceding the compound-statement of a try-block or handler (15 [except])
Preceding the type-specifier-seq of an exception-declaration (15 [except])
Another group of problems is the failure to specify to what a given attribute-specifier appertains:
In a condition (6.4 [stmt.select])
In a for-range-declaration (6.5.4 [stmt.ranged])
In a parameter-declaration (8.3.5 [dcl.fct])
In a conversion-type-id (12.3.2 [class.conv.fct])
There is also a problem in the specification of the interpretation of an initial attribute-specifier. 8.3 [dcl.meaning] paragraph 5 says,
In a declaration attribute-specifieropt T attribute-specifieropt D where D is an unadorned identifier the type of this identifier is “T”. The first optional attribute-specifier appertains to the entity being declared.
This wording only covers the case where the declarator is a simple identifier. It leaves unspecified the meaning of the initial attribute-specifier with more complex declarators for pointers, references, functions, and arrays.
Finally, something needs to be said about the case where attribute-specifiers occur in both the initial position and following the declarator-id: is this permitted, and if so, under what constraints?
(See also issue 968.)
The [[ ... ]] notation for attributes was thought to be completely unambiguous. However, it turns out that two [ characters can be adjacent and not be an attribute-introducer: the first could be the beginning of an array bound or subscript operator and the second could be the beginning of a lambda-introducer. This needs to be explored and addressed.
(See also issue 951.)
According to 7.6.2 [dcl.align] paragraph 1, an alignment attribute can be specified only for a variable or a class data member. The corresponding Microsoft and GNU attributes can be also specified for a class type, and this usage seems to be widespread. It should be permitted with the standard attribute and there seems no good reason not to do so for enumeration types, as well.
Notes from the October, 2009 meeting:
Although there was initial concern for how to integrate the suggested change into the type system, it was observed that current practice is to have the attribute affect only the layout, not the type.
This case is nonstandard by 8.3 [dcl.meaning] paragraph 1 (there is a requirement that the specialization first be declared within the namespace before being defined outside of the namespace), but probably should be allowed:
namespace NS1 {
template<class T>
class CDoor {
public:
int mtd() { return 1; }
};
}
template<> int NS1::CDoor<char>::mtd()
{
return 0;
}
Notes from October 2002 meeting:
There was agreement that we wanted to allow this.
According to 8.3 [dcl.meaning] paragraph 1,
When the declarator-id is qualified, the declaration shall refer to a previously declared member of the class or namespace to which the qualifier refers (or of an inline namespace within that scope (7.3.1 [namespace.def])), 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 of the declarator-id.
This would appear to make the following example ill-formed, even though it would be well-formed if the using-declaration were omitted:
namespace A {
inline namespace B {
template <class T> void foo() { }
}
using B::foo;
}
template void A::foo<int>();
This seems strange.
Proposed resolution (July, 2009):
Change 8.3 [dcl.meaning] paragraph 1 as follows:
...When the declarator-id is qualified, the declaration shall refer to a previously declared member of the class or namespace to which the qualifier refers (or, in the case of a namespace, of an element of the inline namespace within that scope set of that namespace (7.3.1 [namespace.def])), and; the member shall not merely have been introduced by a using-declaration in the scope of the class or namespace nominated by the nested-name-specifier of the declarator-id. [Note:...
(Note: this resolution depends on the resolution of issue 861.)
8.3.2 [dcl.ref] paragraph 4 says:
A reference shall be initialized to refer to a valid object or function. [Note: in particular, a null reference cannot exist in a well-defined program, because the only way to create such a reference would be to bind it to the "object" obtained by dereferencing a null pointer, which causes undefined behavior ...]
What is a "valid" object? In particular the expression "valid object" seems to exclude uninitialized objects, but the response to Core Issue 363 clearly says that's not the intent. This is an example (overloading construction on constness of *this) by John Potter, which I think is supposed to be legal C++ though it binds references to objects that are not initialized yet:
struct Fun {
int x, y;
Fun (int x, Fun const&) : x(x), y(42) { }
Fun (int x, Fun&) : x(x), y(0) { }
};
int main () {
const Fun f1 (13, f1);
Fun f2 (13, f2);
cout << f1.y << " " << f2.y << "\n";
}
Suggested resolution: Changing the final part of 8.3.2 [dcl.ref] paragraph 4 to:
A reference shall be initialized to refer to an object or function.
From its point of declaration on (see 3.3.2 [basic.scope.pdecl])
its name is an lvalue
which refers to that object or function. The reference may be
initialized to refer to an uninitialized object but, in that case,
it is usable in limited ways (3.8 [basic.life], paragraph 6)
[Note: On the other hand, a declaration like this:
int & ref = *(int*)0;
is ill-formed because ref will not refer to any object or function
]
I also think a "No diagnostic is required." would better be added (what about something like int& r = r; ?)
Proposed Resolution (October, 2004):
(Note: the following wording depends on the proposed resolution for issue 232.)
Change 8.3.2 [dcl.ref] paragraph 4 as follows:
A reference shall be initialized to refer to a valid object or function. If an lvalue to which a reference is directly bound designates neither an existing object or function of an appropriate type (8.5.3 [dcl.init.ref]), nor a region of memory of suitable size and alignment to contain an object of the reference's type (1.8 [intro.object], 3.8 [basic.life], 3.9 [basic.types]), the behavior is undefined. [Note: in particular, a null reference cannot exist in a well-defined program, because the only way to create such a reference would be to bind it to the “object” empty lvalue obtained by dereferencing a null pointer, which causes undefined behavior. As does not designate an object or function. Also, as described in 9.6 [class.bit], a reference cannot be bound directly to a bit-field. ]
The name of a reference shall not be used in its own initializer. Any other use of a reference before it is initialized results in undefined behavior. [Example:
int& f(int&); int& g(); extern int& ir3; int* ip = 0; int& ir1 = *ip; // undefined behavior: null pointer int& ir2 = f(ir3); // undefined behavior: ir3 not yet initialized int& ir3 = g(); int& ir4 = f(ir4); // ill-formed: ir4 used in its own initializer—end example]
Rationale: The proposed wording goes beyond the specific concerns of the issue. It was noted that, while the current wording makes cases like int& r = r; ill-formed (because r in the initializer does not "refer to a valid object"), an inappropriate initialization can only be detected, if at all, at runtime and thus "undefined behavior" is a more appropriate treatment. Nevertheless, it was deemed desirable to continue to require a diagnostic for obvious compile-time cases.
It was also noted that the current Standard does not say anything about using a reference before it is initialized. It seemed reasonable to address both of these concerns in the same wording proposed to resolve this issue.
Notes from the April, 2005 meeting:
The CWG decided that whether to require an implementation to diagnose initialization of a reference to itself should be handled as a separate issue (504) and also suggested referring to “storage” instead of “memory” (because 1.8 [intro.object] defines an object as a “region of storage”).
Proposed Resolution (April, 2005):
(Note: the following wording depends on the proposed resolution for issue 232.)
Change 8.3.2 [dcl.ref] paragraph 4 as follows:
A reference shall be initialized to refer to a valid object or function. If an lvalue to which a reference is directly bound designates neither an existing object or function of an appropriate type (8.5.3 [dcl.init.ref]), nor a region of storage of suitable size and alignment to contain an object of the reference's type (1.8 [intro.object], 3.8 [basic.life], 3.9 [basic.types]), the behavior is undefined. [Note: in particular, a null reference cannot exist in a well-defined program, because the only way to create such a reference would be to bind it to the “object” empty lvalue obtained by dereferencing a null pointer, which causes undefined behavior. As does not designate an object or function. Also, as described in 9.6 [class.bit], a reference cannot be bound directly to a bit-field. ]
Any use of a reference before it is initialized results in undefined behavior. [Example:
int& f(int&); int& g(); extern int& ir3; int* ip = 0; int& ir1 = *ip; // undefined behavior: null pointer int& ir2 = f(ir3); // undefined behavior: ir3 not yet initialized int& ir3 = g(); int& ir4 = f(ir4); // undefined behavior: ir4 used in its own initializer—end example]
Note (February, 2006):
The word “use” in the last paragraph of the proposed resolution was intended to refer to the description in 3.2 [basic.def.odr] paragraph 2. However, that section does not define what it means for a reference to be “used,” dealing only with objects and functions. Additional drafting is required to extend 3.2 [basic.def.odr] paragraph 2 to apply to references.
Additional note (May, 2008):
The proposed resolution for issue 570 adds wording to define “use” for references.
Paragraph 7 of 8.3.4 [dcl.array] says,
If E is an n-dimensional array of rank i × j × ... × k, then E appearing in an expression is converted to a pointer to an (n - 1)-dimensional array with rank j × ... × k.
This formulation does not allow for the existence of expressions in which the array-to-pointer conversion does not occur (as specified in clause 5 [expr] paragraph 9). This paragraph should be no more than a note, if it appears at all, and the wording should be corrected.
EDG rejects this code:
template <typename T>
struct S {};
void f (S<int (*)[]>);
G++ accepts it.
This is another case where the standard isn't very clear:
The language from 8.3.5 [dcl.fct] is:
If the type of a parameter includes a type of the form "pointer to array of unknown bound of T" or "reference to array of unknown bound of T," the program is ill-formed.Since "includes a type" is not a term defined in the standard, we're left to guess what this means. (It would be better if this were a recursive definition, the way a type theoretician would do it:
Notes from April 2003 meeting:
We agreed that the example should be allowed.
8.3.5 [dcl.fct] paragraph 13 requires that a parameter pack, if present, must appear at the end of the parameter list. This restriction is not necessary when template argument deduction is not needed and is inconsistent with the way pack expansions are handled. It should be removed.
(See also issue 692.)
According to 3.3.4 [basic.scope.proto] paragraph 1,
In a function declaration, or in any function declarator except the declarator of a function definition (8.4 [dcl.fct.def]), names of parameters (if supplied) have function prototype scope, which terminates at the end of the nearest enclosing function declarator.
Happily, this permits the use of parameter names with decltype in a late-specified return type, because the return type is part of the function's declarator. However, the note in 8.3.5 [dcl.fct] paragraph 11 is now inaccurate and should be updated:
[Note: ...If a parameter name is present in a function declaration that is not a definition, it cannot be used outside of the parameter-declaration-clause since it goes out of scope at the end of the function declarator (3.3 [basic.scope]). —end note]
According to the definition of value initialization (8.5 [dcl.init] paragraph 5), non-union class types without user-declared constructors are value-initialized by value-initializing each of their members rather than by executing the (generated) default constructor. However, a number of other items in the Standard are described in relationship to the execution of the constructor:
12.4 [class.dtor] paragraph 6: “Bases and members are destroyed in the reverse order of the completion of their constructor.” If a given base or member is value-initialized without running its constructor, is it destroyed? (For that matter, paragraph 10 refers to “constructed” objects; is an object that is value-initialized without invoking a constructor “constructed?”)
15.2 [except.ctor] paragraph 2: “An object that is partially constructed or partially destroyed will have destructors executed for all of its fully constructed subobjects, that is, for subobjects for which the constructor has completed execution...”
3.8 [basic.life] paragraph 1: The lifetime of an object begins when “the constructor call has completed.” (In the TC1 wording — “if T is a class type with a non-trivial constructor (12.1 [class.ctor]), the constructor call has completed” — the lifetime of some value-initialized objects never began; in the current wording — “the constructor invoked to create the object is non-trivial” — the lifetime begins before any of the members are initialized.)
Proposed resolution (October, 2005):
Add the indicated words to 8.5 [dcl.init] paragraph 6:
A program that calls for default-initialization or value-initialization of an entity of reference type is ill-formed. If T is a cv-qualified type, the cv-unqualified version of T is used for these definitions of zero-initialization, default-initialization, and value-initialization. Even when value-initialization of an object does not call that object's constructor, the object is deemed to have been fully constructed once its initialization is complete and thus subject to provisions of this International Standard applying to “constructed” objects, objects “for which the constructor has completed execution,” etc.
Notes from April, 2006 meeting:
There was some concern about whether this wording covered (or needed to cover) cases where an object is “partially constructed.” Another approach might be simply to define value initialization to be “construction.” Returned to “drafting” status for further investigation.
8.5 [dcl.init] paragraph 2 reads,
Automatic, register, static, and external variables of namespace scope can be initialized by arbitrary expressions involving literals and previously declared variables and functions.
Both “automatic” and “static” are used to describe storage durations, “register” is a storage class specifier which indicates the object has automatic storage duration, “external” describes linkage, and “namespace scope” is a kind of scope. Automatic, register, static and external, together with namespace scope, are used to restrict the “variables.”
Register objects are only a sub-set of automatic objects and thus the word “register” is redundant and should be elided. If register objects are to be emphasized, they should be mentioned like “Automatic (including register)...”
Variables having namespace scope can never be automatic; they can only be static, with either external or internal linkage. Therefore, there are in fact no “automatic variables of namespace scope,” and the “static” in “static variables of namespace scope” is useless.
In fact, automatic and static variables already compose all variables with either external linkage or not, and thus the “external” becomes redundant, too, and the quoted sentence seems to mean that all variables of namespace scope can be initialized by arbitrary expressions. But this is not true because not all internal variables of namespace scope can. Therefore, the restrictive “external” is really necessary, not redundant.
As a result, the erroneous restrictive “automatic, register, static” should be removed and the quoted sentence may be changed to:
External variables of namespace scope can be initialized by arbitrary expressions involving literals and previously declared variables and functions.
Notes from the April, 2007 meeting:
This sentence is poorly worded, but the analysis given in the issue description is incorrect. The intent is simply that the storage class of a variable places no restrictions on the kind of expression that can be used to initialize it (in contrast to C, where variables of static storage duration can only be initialized by constant expressions).
Proposed resolution (June, 2008):
Change 8.5 [dcl.init] paragraph 2 as follows:
Automatic, register, static, and external variables of namespace scope Variables of automatic, thread, and static storage duration can be initialized by arbitrary expressions involving literals and previously declared variables and functions...
Notes from the September, 2008 meeting:
The existing wording is intended to exclude block-scope extern declarations but to allow initializers in all other forms of variable declarations. The best way to phrase that is probably to say that all variable definitions (except for function parameters, where the initializer syntax is used for default arguments) can have arbitrary expressions as initializers, regardless of storage duration.
The C committee is considering changing the definition of zero-initialization of unions to guarantee that the bytes of the entire union are set to zero before assigning 0, converted to the appropriate type, to the first member. The argument (summarized here) is for backward compatibility. The C++ Committee may want to consider the same change.
Proposed resolution (August, 2008):
Change bullet 3 of 8.5 [dcl.init] paragraph 5 (in the first list, dealing with zero-initialization) as follows:
[Drafting notes: Ask a C liaison about the progress of WG14 paper N1311, which deals with this issue. Since the adoption of WG21 paper N2544, unions may have static data members, hence the change to refer to the first non-static data member and the deletion of the footnote.]
Notes from the September, 2008 meeting:
It was observed that padding bytes in structs are zero-initialized in C, so if we are changing the treatment of unions in this way we should consider adding the C behavior for padding bytes at the same time. In particular, using memcmp to compare structs only works reliably if the padding bytes are zero-initialized.
8.5 [dcl.init] paragraph 11 says,
If no initializer is specified for an object, the object is default-initialized; if no initialization is performed, a non-static object has indeterminate value.
This is inaccurate, because objects with thread storage duration are zero-initialized (3.6.2 [basic.start.init] paragraph 2).
The current wording of 8.5.1 [dcl.init.aggr] paragraph 1 does not consider brace-or-equal-initializers on members as affecting whether a class type is an aggregate or not. Because in-class member initializers are essentially syntactic sugar for mem-initializers, and the presence of a user-provided constructor disqualifies a class from being an aggregate, presumably the same should hold true of member initializers.
There is an inconsistency in the handling of references vs pointers in user defined conversions and overloading. The reason for that is that the combination of 8.5.3 [dcl.init.ref] and 4.4 [conv.qual] circumvents the standard way of ranking conversion functions, which was probably not the intention of the designers of the standard.
Let's start with some examples, to show what it is about:
struct Z { Z(){} };
struct A {
Z x;
operator Z *() { return &x; }
operator const Z *() { return &x; }
};
struct B {
Z x;
operator Z &() { return x; }
operator const Z &() { return x; }
};
int main()
{
A a;
Z *a1=a;
const Z *a2=a; // not ambiguous
B b;
Z &b1=b;
const Z &b2=b; // ambiguous
}
So while both classes A and B are structurally equivalent, there is a difference in operator overloading. I want to start with the discussion of the pointer case (const Z *a2=a;): 13.3.3 [over.match.best] is used to select the best viable function. Rule 4 selects A::operator const Z*() as best viable function using 13.3.3.2 [over.ics.rank] since the implicit conversion sequence const Z* -> const Z* is a better conversion sequence than Z* -> const Z*.
So what is the difference to the reference case? Cv-qualification conversion is only applicable for pointers according to 4.4 [conv.qual]. According to 8.5.3 [dcl.init.ref] paragraphs 4-7 references are initialized by binding using the concept of reference-compatibility. The problem with this is, that in this context of binding, there is no conversion, and therefore there is also no comparing of conversion sequences. More exactly all conversions can be considered identity conversions according to 13.3.3.1.4 [over.ics.ref] paragraph 1, which compare equal and which has the same effect. So binding const Z* to const Z* is as good as binding const Z* to Z* in terms of overloading. Therefore const Z &b2=b; is ambiguous. [13.3.3.1.4 [over.ics.ref] paragraph 5 and 13.3.3.2 [over.ics.rank] paragraph 3 rule 3 (S1 and S2 are reference bindings ...) do not seem to apply to this case]
There are other ambiguities, that result in the special treatment of references: Example:
struct A {int a;};
struct B: public A { B() {}; int b;};
struct X {
B x;
operator A &() { return x; }
operator B &() { return x; }
};
main()
{
X x;
A &g=x; // ambiguous
}
Since both references of class A and B are reference compatible with references of class A and since from the point of ranking of implicit conversion sequences they are both identity conversions, the initialization is ambiguous.
So why should this be a defect?
So overall I think this was not the intention of the authors of the standard.
So how could this be fixed? For comparing conversion sequences (and only for comparing) reference binding should be treated as if it was a normal assignment/initialization and cv-qualification would have to be defined for references. This would affect 8.5.3 [dcl.init.ref] paragraph 6, 4.4 [conv.qual] and probably 13.3.3.2 [over.ics.rank] paragraph 3.
Another fix could be to add a special case in 13.3.3 [over.match.best] paragraph 1.
According to 8.5.3 [dcl.init.ref] paragraph 5, a reference initialized with a reference-compatible rvalue of class type binds directly to the object. A reference-compatible non-class rvalue reference, however, is first copied to a temporary and the reference binds to that temporary, not to the target of the rvalue reference. This can cause problems when the result of a forwarding function is used in such a way that the address of the result is captured. For example:
struct ref {
explicit ref(int&& i): p(&i) { }
int* p;
};
int&& forward(int&& i) {
return i;
}
void f(int&& i) {
ref r(forward(i));
// Here r.p is a dangling pointer, pointing to a defunct int temporary
}
A formulation is needed so that rvalue references are treated like class and array rvalues.
Notes from the February, 2008 meeting:
You can't just treat scalar rvalues like class and array rvalues, because they might not have an associated object. However, if you have an rvalue reference, you know that there is an object, so probably the best way to address this issue is to specify somehow that binding a reference to an rvalue reference does not introduce a new temporary.
(See also issues 690 and 846.)
It should always be possible to use the new brace syntax to value-initialize an object. However, the current rules make the following example ill-formed because of ambiguity:
struct S {
S();
S(std::initializer_list<int>);
S(std::initializer_list<double>);
};
S s{}; // Ambiguous initializer-list constructor reference,
// not value initialization.
The grammar for member-declaration in 9.2 [class.mem] does not include a production for the alias-declaration form of typedef declarations, meaning that something like
struct S {
using UINT = unsigned int;
};
is ill-formed. This seems like an oversight.
The type long long is missing from the list of bit-field types in 9.6 [class.bit] paragraph 3 for which the implementation can choose the signedness. This was presumably an oversight. (If that is the case, we may want to reconsider the handling of 4.5 [conv.prom] paragraph 3: a long long bit-field that the implementation treats as unsigned will — pending the outcome of issue 739 — still promote to signed long long, which can lead to unexpected results for bit-fields with the same number of bits as long long.)
According to 9.8 [class.local] paragraph 1,
Declarations in a local class can use only type names, static variables, extern variables and functions, and enumerators from the enclosing scope.
This would presumably make both of the members of S2 below ill-formed:
void test () {
const int local_const = 7;
struct S2 {
int member:local_const;
void f() { int j = local_const; }
};
}
Should there be an exception to this rule for constant values? Current implementations seem to accept the reference to local_const in the bit-field declaration but not in the member function definition. Should they be the same or different?
Notes from the September, 2008 meeting:
The CWG agreed that both uses of local_const in the example above should be accepted. The intent of the restriction was to avoid the need to pass a frame pointer into local class member functions, so uses of local const variables as values should be permitted.
Proposed resolution (March, 2009):
Change 9.8 [class.local] paragraph 1 as follows:
Declarations in a local class can use only type names, static
variables, extern variables and functions, and enumerators
shall not use (3.2 [basic.def.odr]) an automatic variable or
reference from the enclosing scope. [Example:
int x;
void f() {
static int s ;
int x;
extern int g();
const int c = 42;
struct local {
int g() { return x; } // error: x has automatic storage duration
int h() { return s; } // OK
int k() { return ::x; } // OK
int l() { return g(); } // OK
int m() { return c; } // OK
};
}
local* p = 0; // error: local not in scope
—end example]
Notes from the July, 2009 meeting:
This proposed resolution relies on the definition of “use” in 3.2 [basic.def.odr]. The CWG was concerned about cases in which it might not be possible to immediately determine whether a reference to a local automatic variable constitutes a “use” or not, such as in overload resolution, conditional expressions, dependent contexts, etc. To address this concern, the CWG expressed support for an approach in which a reference to a local automatic variable in a nested class or lambda body would enter the expression as an rvalue, which would reduce the complexity of the problem.
Proposed resolution (September, 2009):
Change 5.1.1 [expr.prim.general] paragraph 6 as follows and add a new paragraph immediately following:
...The type of the expression is the type of the identifier. The result is the entity denoted by the identifier. The result is an lvalue if the entity is a function, variable, or data member.
Certain contexts (9.8 [class.local], 5.1.2 [expr.prim.lambda]) are called restricted automatic variable contexts because they permit the use of automatic variables declared in enclosing scopes only under certain conditions. In these contexts (only), an identifier denoting a variable with automatic storage duration declared in an enclosing scope that satisfies the requirements for appearing in a constant expression (5.19 [expr.const]) is called an r-variable expression. The result of an r-variable expression is an rvalue whose value is that of the variable. In all other cases, the result of an identifier expression denoting a variable is an lvalue.
Change 5.1.2 [expr.prim.lambda] paragraph 9 as follows:
A The compound-statement of a lambda-expression is a restricted automatic variable context (5.1.1 [expr.prim.general]), and r-variable expressions within its scope are permitted. In addition, a lambda-expression's compound-statement can use (see above) this from an immediately-enclosing member function definition, as well as variables and references with automatic storage duration from an immediately-enclosing function definition or lambda-expression, provided these entities are captured (as described below). Any other use (3.2 [basic.def.odr]) of a variable or reference with automatic storage duration declared outside the lambda-expression is ill-formed. [Example:
void f1(int i) { int const N = 20; [=]{ int const M = 30; int j; [=]{ int x[N][M]; // OK: N and M are not used r-variable expressions x[0][0] = i; // error: i is not declared in the immediately }; // enclosing lambda-expression int y = N+M; // OK: N and M are r-variable expressions &M; // error: taking the address of an rvalue y = i; // error: i is not an r-variable expression and variable i // is declared outside the immediately-enclosing // lambda-expression and thus not captured sizeof(i); // OK: i is not “used” int z = j; // OK: variable j is implicitly captured }; [M]{ int a[M]; // error: variable M explicitly captured, so M //refers to a member of the closure type }; }; }—end example]
Change 5.1.2 [expr.prim.lambda] paragraph 11 as follows:
If a lambda-expression has an associated capture-default and its compound-statement uses (3.2 [basic.def.odr]) this or a variable or reference with automatic storage duration declared in an enclosing function or lambda-expression and the used entity is not explicitly captured, then the used entity is said to be implicitly captured. An entity that is used (3.2 [basic.def.odr]) in the compound-statement of a lambda-expression but not explicitly captured will be implicitly captured if:
the lambda-expression has an associated capture-default, and
the entity is this or a variable or reference with automatic storage duration declared in an enclosing function or lambda-expression, and
the use is not an r-variable expression.
[Note: Implicit uses of this can result in implicit capture. —end note]
Change 5.1.2 [expr.prim.lambda] paragraph 16 as follows:
Every id-expression that is a use (3.2 [basic.def.odr]) of an entity captured by copy If an id-expression denotes an entity that is explicitly captured by copy, or it it is a use (3.2 [basic.def.odr]) of an entity that is implicitly captured by copy, that id-expression is transformed into an access to the corresponding unnamed data member of the closure type. [Note: r-variable expressions designating variables that are explicitly captured by copy are thus not id-expressions after this transformation and consequently are no longer r-variable expressions. If a variable is not explicitly captured, however, the fact that it is used as an r-variable expression does not cause that variable to be implicitly captured, either. As a result, such an id-expression will not be transformed to a member access, and it will therefore be treated as a constant rvalue. —end note] If this is captured, each use of this is transformed into an access to the corresponding unnamed data member of the closure type cast (5.4 [expr.cast]) to the type of this. [Note: The cast ensures that the transformed expression is an rvalue. —end note]
Change 9.8 [class.local] paragraph 1 as follows:
A class can be 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. A local class definition is a restricted automatic variable context (5.1.1 [expr.prim.general]). Declarations in a local class can use only type names, static variables, extern variables and functions, and enumerators shall not use (3.2 [basic.def.odr]) a variable or reference with automatic storage duration from the enclosing scope except as an r-variable expression. [Example:int x; void f() { static int s ; int x; extern int g(); const int c = 42; struct local { int g() { return x; } // error: x has automatic storage duration is not an r-variable expression int h() { return s; } // OK int k() { return ::x; } // OK int l() { return g(); } // OK int m() { return c; } // OK: c is an r-variable expression int* n() { return &c; } // error: taking the address of an rvalue }; } local* p = 0; // error: local not in scope—end example]
Drafting note: The change to 5.1.2 [expr.prim.lambda] paragraph 16 has the effect of making explicitly-captured variables no longer usable in constant expressions in lambda bodies. This change facilitates the design goal to be able to determine in a context-free manner whether a given id-expression should be transformed to a member access expression or not.
Notes from the October, 2009 meeting:
There was interest in an approach that would allow explicitly-captured constants to appear in constant expressions but also to be “used.” Another suggestion was to have variables captured if they appear in either “use” or “non-use” contexts.
10.3 [class.virtual] paragraph 5 requires that covariant return types be either both pointers or both references, but it does not specify that references must be both lvalue references or both rvalue references. Presumably this is an oversight.
The resolution of issue 372 leaves unclear whether the following are well-formed or not:
class C {
typedef int I; // private
template <int> struct X;
template <int> friend struct Y;
}
template <C::I> struct C::X { }; // C::I accessible to member?
template <C::I> struct Y { }; // C::I accessible to friend?
Presumably the answer to both questions is “yes,” but the new wording does not address template-parameters.
Proposed resolution (June, 2008):
Change 11 [class.access] paragraph 6 as follows:
...For purposes of access control, the base-specifiers of a class, the template-parameters of a template-declaration, and the definitions of class members that appear outside of the class definition are considered to be within the scope of that class...
Notes from the September, 2008 meeting:
The proposed resolution preserves the word “scope” as a holdover from the original specification prior to issue 372, which intended to change access determination from a scope-based model to an entity-based model. The resolution should eliminate all references to scope and simply use the entity-based model.
(See also issue 718.)
The access rules in 11.2 [class.access.base] do not appear to handle references in nested classes and outside of nonstatic member functions correctly. For example,
struct A {
typedef int I; // public
};
struct B: private A { };
struct C: B {
void f() {
I i1; // error: access violation
}
I i2; // OK
struct D {
I i3; // OK
void g() {
I i4; // OK
}
};
};
The reason for this discrepancy is that the naming class in the reference to I is different in these cases. According to 11.2 [class.access.base] paragraph 5,
The access to a member is affected by the class in which the member is named. This naming class is the class in which the member name was looked up and found.
In the case of i1, the reference to I is subject to the transformation described in 9.3.1 [class.mfct.non-static] paragraph 3:
Similarly during name lookup, when an unqualified-id (5.1 [expr.prim]) used in the definition of a member function for class X resolves to a static member, an enumerator or a nested type of class X or of a base class of X, the unqualified-id is transformed into a qualified-id (5.1 [expr.prim]) in which the nested-name-specifier names the class of the member function.
As a result, the reference to I in the declaration of i1 is transformed to C::I, so that the naming class is C, and I is inacessible in C. In the remaining cases, however, the transformation does not apply. Thus, the naming class of I in these references is A, and I is publicly accessible in A.
Presumably either the definition of “naming class” must be changed or the transformation of unqualified-ids must be broadened to include all uses within the scope of a class and not just within nonstatic member functions (and following the declarator-id in the definition of a static member, per 9.4 [class.static] paragraph 4).
Does the restriction in 11.5 [class.protected] apply to upcasts across protected inheritance, too? For instance,
struct B {
int i;
};
struct I: protected B { };
struct D: I {
void f(I* ip) {
B* bp = ip; // well-formed?
bp->i = 5; // aka "ip->i = 5;"
}
};
I think the rationale for the 11.5 [class.protected] restriction applies equally well here — you don't know whether ip points to a D object or not, so D::f can't be trusted to treat the protected B subobject consistently with the policies of its actual complete object type.
The current treatment of “accessible base class” in 11.2 [class.access.base] paragraph 4 clearly makes the conversion from I* to B* well-formed. I think that's wrong and needs to be fixed. The rationale for the accessibility of a base class is whether “an invented public member” of the base would be accessible at the point of reference, although we obscured that a bit in the reformulation; it seems to me that the invented member ought to be considered a non-static member for this purpose and thus subject to 11.5 [class.protected].
(See also issues 385 and 471.).Notes from October 2004 meeting:
The CWG tentatively agreed that casting across protective inheritance should be subject to the additional restriction in 11.5 [class.protected].
According to 12.1 [class.ctor] paragraph 1, only function-specifiers are permitted in the declaration of a constructor, and constexpr is not a function-specifier. (See also issue 263, in which the resolution of a similar concern regarding the friend specifier did not change 12.1 [class.ctor] paragraph 1 but perhaps should have done so.)
Mark Mitchell raised a number of issues related to the resolution of issue 244 and of destructor lookup in general.
Issue 244 says:
... 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.
But if the reference is "p->X::~X()", the first class-name is looked up in two places (normal lookup and a lookup in the class of p). Does the new wording mean:
This is a test case that illustrates the issue:
struct A {
typedef A C;
};
typedef A B;
void f(B* bp) {
bp->B::~B(); // okay B found by normal lookup
bp->C::~C(); // okay C found by class lookup
bp->B::~C(); // B found by normal lookup C by class -- okay?
bp->C::~B(); // C found by class lookup B by normal -- okay?
}
A second issue concerns destructor references when the class involved is a template class.
namespace N {
template <typename T> struct S {
~S();
};
}
void f(N::S<int>* s) {
s->N::S<int>::~S();
}
The issue here is that the grammar uses "~class-name" for destructor names, but in this case S is a template name when looked up in N.
Finally, what about cases like:
template <typename T> void f () {
typename T::B x;
x.template A<T>::template B<T>::~B();
}
When parsing the template definition, what checks can be done on "~B"?
Sandor Mathe adds :
The standard correction for issue 244 (now in DR status) is still incomplete.
Paragraph 5 of 3.4.3 [basic.lookup.qual] is not applicable for p->T::~T since there is no nested-name-specifier. Section 3.4.5 [basic.lookup.classref] describes the lookup of p->~T but p->T::~T is still not described. There are examples (which are non-normative) that illustrate this sort of lookup but they still leave questions unanswered. The examples imply that the name after ~ should be looked up in the same scope as the name before the :: but it is not stated. The problem is that the name to the left of the :: can be found in two different scopes. Consider the following:
struct S {
struct C { ~C() { } };
};
typedef S::C D;
int main() {
D* p;
p->C::~D(); // valid?
}
Should the destructor call be valid? If there were a nested name specifier, then D should be looked for in the same scope as C. But here, C is looked for in 2 different ways. First, it is searched for in the type of the left hand side of -> and it is also looked for in the lexical context. It is found in one or if both, they must match. So, C is found in the scope of what p points at. Do you only look for D there? If so, this is invalid. If not, you would then look for D in the context of the expression and find it. They refer to the same underlying destructor so this is valid. The intended resolution of the original defect report of the standard was that the name before the :: did not imply a scope and you did not look for D inside of C. However, it was not made clear whether this was to be resolved by using the same lookup mechanism or by introducing a new form of lookup which is to look in the left hand side if that is where C was found, or in the context of the expression if that is where C was found. Of course, this begs the question of what should happen when it is found in both? Consider the modification to the above case when C is also found in the context of the expression. If you only look where you found C, is this now valid because it is in 1 of the two scopes or is it invalid because C was in both and D is only in 1?
struct S {
struct C { ~C() { } };
};
typedef S::C D;
typedef S::C C;
int main() {
D* p;
p->C::~D(); // valid?
}
I agree that the intention of the committee is that the original test case in this defect is broken. The standard committee clearly thinks that the last name before the last :: does not induce a new scope which is our current interpretation. However, how this is supposed to work is not defined. This needs clarification of the standard.
Martin Sebor adds this example (September 2003), along with errors produced by the EDG front end:
namespace N {
struct A { typedef A NA; };
template <class T> struct B { typedef B NB; typedef T BT; };
template <template <class> class T> struct C { typedef C NC; typedef T<A> CA; };
}
void foo (N::A *p)
{
p->~NA ();
p->NA::~NA ();
}
template <class T>
void foo (N::B<T> *p)
{
p->~NB ();
p->NB::~NB ();
}
template <class T>
void foo (typename N::B<T>::BT *p)
{
p->~BT ();
p->BT::~BT ();
}
template <template <class> class T>
void foo (N::C<T> *p)
{
p->~NC ();
p->NC::~NC ();
}
template <template <class> class T>
void foo (typename N::C<T>::CA *p)
{
p->~CA ();
p->CA::~CA ();
}
Edison Design Group C/C++ Front End, version 3.3 (Sep 3 2003 11:54:55)
Copyright 1988-2003 Edison Design Group, Inc.
"t.cpp", line 16: error: invalid destructor name for type "N::B<T>"
p->~NB ();
^
"t.cpp", line 17: error: qualifier of destructor name "N::B<T>::NB" does not
match type "N::B<T>"
p->NB::~NB ();
^
"t.cpp", line 30: error: invalid destructor name for type "N::C<T>"
p->~NC ();
^
"t.cpp", line 31: error: qualifier of destructor name "N::C<T>::NC" does not
match type "N::C<T>"
p->NC::~NC ();
^
4 errors detected in the compilation of "t.cpp".
John Spicer: The issue here is that we're unhappy with the destructor names when doing semantic analysis of the template definitions (not during an instantiation).
My personal feeling is that this is reasonable. After all, why would you call p->~NB for a class that you just named as N::B<T> and you could just say p->~B?
Additional note (September, 2004)
The resolution for issue 244 removed the discussion of p->N::~S, where N is a namespace-name. However, the resolution did not make this construct ill-formed; it simply left the semantics undefined. The meaning should either be defined or the construct made ill-formed.
The changes for delegating constructors overlooked the need to change 12.6.2 [class.base.init] paragraph 3:
The expression-list in a mem-initializer is used to initialize the base class or non-static data member subobject denoted by the mem-initializer-id. The semantics of a mem-initializer are as follows:
if the expression-list of the mem-initializer is omitted, the base class or member subobject is value-initialized (see 8.5 [dcl.init]);
otherwise, the subobject indicated by mem-initializer-id is direct-initialized using expression-list as the initializer (see 8.5 [dcl.init]).
The initialization of each base and member constitutes a full-expression. Any expression in a mem-initializer is evaluated as part of the full-expression that performs the initialization.
This paragraph deals only with subobjects; it needs to be made more general to apply to the complete object as well when the mem-initializer-id designates the constructor's class.
Proposed resolution (June, 2008):
Change 12.6.2 [class.base.init] paragraph 3 as follows:
The expression-list in a mem-initializer is used to initialize the base class or non-static data member subobject denoted by the mem-initializer-id. The semantics of a mem-initializer are A mem-initializer in which the mem-initializer-id names the constructor's class initializes the object by invoking the selected target constructor with the mem-initializer's expression-list. A mem-initializer in which the mem-initializer-id names a base class or non-static data member initializes the designated subobject as follows:
if the expression-list of the mem-initializer is omitted, the base class or member subobject is value-initialized (see 8.5 [dcl.init]);
otherwise, the subobject indicated by mem-initializer-id is direct-initialized using expression-list as the initializer (see 8.5 [dcl.init]).
...
The initialization of each base and member performed by each mem-initializer constitutes a full-expression. Any expression...
Notes from the September, 2008 meeting:
This text was significantly modified by N2756 (nonstatic data member initializers) and needs to be reworked in light of those changes.
According to 13.3.3.1.4 [over.ics.ref] paragraphs 3-4,
A standard conversion sequence cannot be formed if it requires binding an lvalue reference to non-const to an rvalue (except when binding an implicit object parameter; see the special rules for that case in 13.3.1 [over.match.funcs]). [Note: this means, for example, that a candidate function cannot be a viable function if it has a non-const lvalue reference parameter (other than the implicit object parameter) and the corresponding argument is a temporary or would require one to be created to initialize the lvalue reference (see 8.5.3 [dcl.init.ref]). —end note]
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.
Because this section does not mention attempting to bind an rvalue reference to an lvalue, such a “conversion sequence” might be selected as best and result in an ill-formed program. It should, instead, be treated like trying to bind an lvalue reference to non-const to an rvalue, making the function non-viable.
The list of overloads for user-defined literal operators given in 13.5.8 [over.literal] paragraph 3 should include signatures for char, wchar_t, char16_t, and char32_t.
According to the Standard (although not implemented this way in most implementations), the following code exhibits non-intuitive behavior:
struct T {
operator short() const;
operator int() const;
};
short s;
void f(const T& t) {
s = t; // surprisingly calls T::operator int() const
}
The reason for this choice is 13.6 [over.built] paragraph 18:
For every triple (L, VQ, R), where L is an arithmetic type, VQ is either volatile or empty, and R is a promoted arithmetic type, there exist candidate operator functions of the form
VQ L& operator=(VQ L&, R);
Because R is a "promoted arithmetic type," the second argument to the built-in assignment operator is int, causing the unexpected choice of conversion function.
Suggested resolution: Provide built-in assignment operators for the unpromoted arithmetic types.
Related to the preceding, but not resolved by the suggested resolution, is the following problem. Given:
struct T {
operator int() const;
operator double() const;
};
I believe the standard requires the following assignment to be ambiguous (even though I expect that would surprise the user):
double x;
void f(const T& t) { x = t; }
The problem is that both of these built-in operator=()s exist (13.6 [over.built] paragraph 18):
double& operator=(double&, int);
double& operator=(double&, double);
Both are an exact match on the first argument and a user conversion on the second. There is no rule that says one is a better match than the other.
The compilers that I have tried (even in their strictest setting) do not give a peep. I think they are not following the standard. They pick double& operator=(double&, double) and use T::operator double() const.
I hesitate to suggest changes to overload resolution, but a possible resolution might be to introduce a rule that, for built-in operator= only, also considers the conversion sequence from the second to the first type. This would also resolve the earlier question.
It would still leave x += t etc. ambiguous -- which might be the desired behavior and is the current behavior of some compilers.
Notes from the 04/01 meeting:
The difference between initialization and assignment is disturbing. On the other hand, promotion is ubiquitous in the language, and this is the beginning of a very slippery slope (as the second report above demonstrates).
Static data members of template classes and of nested classes of template classes are not themselves templates but receive much the same treatment as template. For instance, 14 [temp] paragraph 1 says that templates are only "classes or functions" but implies that "a static data member of a class template or of a class nested within a class template" is defined using the template-declaration syntax.
There are many places in the clause, however, where static data members of one sort or another are overlooked. For instance, 14 [temp] paragraph 6 allows static data members of class templates to be declared with the export keyword. I would expect that static data members of (non-template) classes nested within class templates could also be exported, but they are not mentioned here.
Paragraph 8, however, overlooks static data members altogether and deals only with "templates" in defining the effect of the export keyword; there is no description of the semantics of defining a static data member of a template to be exported.
These are just two instances of a systematic problem. The entire clause needs to be examined to determine which statements about "templates" apply to static data members, and which statements about "static data members of class templates" also apply to static data members of non-template classes nested within class templates.
(The question also applies to member functions of template classes; see issue 217, where the phrase "non-template function" in 8.3.6 [dcl.fct.default] paragraph 4 is apparently intended not to include non-template member functions of template classes. See also issue 108, which would benefit from understanding nested classes of class templates as templates. Also, see issue 249, in which the usage of the phrase "member function template" is questioned.)
Notes from the 4/02 meeting:
Daveed Vandevoorde will propose appropriate terminology.
14.2 [temp.param] paragraph 11 currently says,
If a template-parameter of a class template is a template parameter pack, it shall be the last template-parameter. [Note: These are not requirements for function templates because template arguments might be deduced (14.9.2 [temp.deduct])...
This restriction was only meant to apply to primary class templates, not partial specializations.
Suggested resolution:
If a template-parameter of a primary class template is a template parameter pack, it shall be the last template-parameter. [Note: These are not requirements for function templates or class template partial specializations because template arguments might be deduced (14.9.2 [temp.deduct])...
Consider an example like:
template <typename T, T Value> struct bar { };
template <typename... T, T ...Value> void foo(bar<T, Value>);
The current wording in 14.2 [temp.param] is unclear as to whether this is permitted or not. For comparison, 8.3.5 [dcl.fct] paragraph 13 says,
A declarator-id or abstract-declarator containing an ellipsis shall only be used in a parameter-declaration. Such a parameter-declaration is a parameter pack (14.6.3 [temp.variadic]). When it is part of a parameter-declaration-clause, the parameter pack is a function parameter pack (14.6.3 [temp.variadic]). [Note: Otherwise, the parameter-declaration is part of a template-parameter-list and the parameter pack is a template parameter pack; see 14.2 [temp.param]. —end note] A function parameter pack, if present, shall occur at the end of the parameter-declaration-list. The type T of the declarator-id of the function parameter pack shall contain a template parameter pack; each template parameter pack in T is expanded by the function parameter pack.
The requirement here that the type of a function parameter pack must contain a template parameter pack is not repeated for template non-type parameters in 14.2 [temp.param], nor is the statement that it expands the template parameter pack.
A related issue is that neither function nor template parameter packs are listed in 14.6.3 [temp.variadic] paragraph 4 among the contexts in which a pack expansion can appear.
The following is the wording from 14.3 [temp.names] paragraphs 4 and 5 that discusses the use of the "template" keyword following . or -> and in qualified names.
class X {
public:
template<std::size_t> X* alloc();
template<std::size_t> static X* adjust();
};
template<class T> void f(T* p) {
T* p1 = p->alloc<200>();
// ill-formed: < means less than
T* p2 = p->template alloc<200>();
// OK: < starts template argument list
T::adjust<100>();
// ill-formed: < means less than
T::template adjust<100>();
// OK: < starts explicit qualification
}
—end example]
If a name prefixed by the keyword template is not the name of a member template, the program is ill-formed. [Note: the keyword template may not be applied to non-template members of class templates. ]
The whole point of this feature is to say that the "template" keyword is needed to indicate that a "<" begins a template parameter list in certain contexts. The constraints in paragraph 5 leave open to debate certain cases.First, I think it should be made more clear that the template name must be followed by a template argument list when the "template" keyword is used in these contexts. If we don't make this clear, we would have to add several semantic clarifications instead. For example, if you say "p->template f()", and "f" is an overload set containing both templates and nontemplates: a) is this valid? b) are the nontemplates in the overload set ignored? If the user is forced to write "p->template f<>()" it is clear that this is valid, and it is equally clear that nontemplates in the overload set are ignored. As this feature was added purely to provide syntactic guidance, I think it is important that it otherwise have no semantic implications.
I propose that paragraph 5 be modified to:
(See also issue 30 and document J16/00-0008 = WG21 N1231.)
Notes from 04/00 meeting:
The discussion of this issue revived interest in issues 11 and 109.
Notes from the October 2003 meeting:
We reviewed John Spicer's paper N1528 and agreed with his recommendations therein.
The EDG front-end accepts:
template <typename T>
struct A {
template <typename U>
struct B {};
};
template <typename T>
struct C : public A<T>::template B<T> {
};
It rejects this code if the base-specifier is spelled A<T>::B<T>.
However, the grammar for a base-specifier does not allow the template keyword.
Suggested resolution:
It seems to me that a consistent approach to the solution that looks like it will be adopted for issue 180 (which deals with the typename keyword in similar contexts) would be to assume that B is a template if it is followed by a "<". After all, an expression cannot appear in this context.Notes from the 4/02 meeting:
We agreed that template must be allowed in this context. The syntax needs to be changed. We also opened the related issue 343.
Consider this example:
class Foo {
public:
template< typename T > T *get();
};
template< typename U >
U *testFoo( Foo &foo ) {
return foo.get< U >(); //#1
}
I am under the impression that this should compile without requiring the insertion of the template keyword before get in the expression at //#1. This notion is supported by this note excerpted from 14.3 [temp.names]/5:
[Note: just as is the case with the typename prefix, the template prefix is allowed in cases where it is not strictly necessary; i.e., when the expression on the left of the -> or ., or the nested-name-specifier is not dependent on a template parameter.]
But 14.3 [temp.names]/4 contains this text:
When the name of a member template specialization appears after . or -> in a postfix-expression, or after nested-name-specifier in a qualified-id, and the postfix-expression or qualified-id explicitly depends on a template-parameter (14.6.2), the member template name must be prefixed by the keyword template. Otherwise the name is assumed to name a non-template.
The only way that I can read this to support my assumption above is if I assume that the phrase postfix-expression is used twice above with different meaning. That is I read the first use as referring to the full expression while the second use refers to the subexpression preceding the operator. Is this the correct determination of intent? I find this text confusing. Would it be an improvement if the second occurrence of "postfix-expression" should be replaced by "the subexpression preceding the operator". Of course that begs the question "where is subexpression actually defined in the standard?"
John Spicer: I agree that the code should work, and that we should tweak the wording.
According to 14.4.2 [temp.arg.nontype] paragraph 1, bullet 3, one of the acceptable forms of a non-type, non-template template argument is:
the address of an object or function... expressed as & id-expression where the & is optional if the name refers to a function or array, or if the corresponding template-parameter is a reference
It is not clear from this whether a template argument like (&i) satisfies the requirement or not.
Notes from the March, 2009 meeting:
The consensus of the CWG was that the parentheses should be allowed.
According to
A template-argument matches a template template-parameter (call it P) when each of the template parameters in the template-parameter-list of the template-argument's corresponding class template or template alias (call it A) matches the corresponding template parameter in the template-parameter-list of P. When P's template-parameter-list contains a template parameter pack (14.6.3 [temp.variadic]), the template parameter pack will match zero or more template parameters or template parameter packs in the template-parameter-list of A with the same type and form as the template parameter pack in P (ignoring whether those template parameters are template parameter packs).
The immediately-preceding example, however, assumes that a parameter pack in the parameter will match only a parameter pack in the argument:
template<class T> class A { /* ... */ };
template<class T, class U = T> class B { /* ... */ };
template<class ... Types> class C { /* ... */ };
template<template<class ...> class Q> class Y { /* ... */ };
Y<A> ya; // ill-formed: a template parameter pack does not match a template parameter
Y<B> yb; // ill-formed: a template parameter pack does not match a template parameter
Y<C> yc; // OK
Is this allowed?
template<typename T>
struct X
{
static int s[];
int c;
};
template<typename T>
int X<T>::s[sizeof(X<T>)];
int* p = X<char>::s;
I have a compiler claiming that, for the purpose of sizeof(), X<T> is an incomplete type, when it tries to instantiate X<T>::s. It seems to me that X<char> should be considered complete enough for sizeof even though the size of s isn't known yet.
John Spicer: This is a problematic construct that is currently allowed but which I think should be disallowed.
I tried this with a number of compilers. None of which did the right thing. The EDG front end accepts it, but gives X<...>::s the wrong size.
It appears that most compilers evaluate the "declaration" part of the static data member definition only once when the definition is processed. The initializer (if any) is evaluated for each instantiation.
This problem is solvable, and if it were the only issue with incomplete arrays as template static data members, then it would make sense to solve it, but there are other problems.
The first problem is that the size of the static data member is only known if a template definition of the static data member is present. This is weird to start with, but it also means that sizes would not be available in general for exported templates.
The second problem concerns the rules for specialization. An explicit specialization for a template instance can be provided up until the point that a use is made that would cause an implicit instantiation. A reference like "sizeof(X<char>::s)" is not currently a reference that would cause an implicit instantiation of X<char>::s. This means you could use such a sizeof and later specialize the static data member with a different size, meaning the earlier sizeof gave the wrong result. We could, of course, change the "use" rules, but I'd rather see us require that static data members that are arrays have a size specified in the class or have a size based on their initializer.
Notes from the October 2003 meeting:
The example provided is valid according to the current standard. A static data member must be instantiated (including the processing of its initializer, if any) if there is any reference to it. The compiler need not, however, put out a definition in that translation unit. The standard doesn't really have a concept of a "partial instantiation" for a static data member, and although we considered adding that, we decided that to get all the size information that seems to be available one needs a full instantiation in any case, so there's no need for the concept of a partial instantiation.
Note (June, 2006):
Mark Mitchell suggested the following example:
template <int> void g();
template <typename T>
struct S {
static int i[];
void f();
};
template <typename T>
int S<T>::i[] = { 1 };
template <typename T>
void S<T>::f() {
g<sizeof (i) / sizeof (int)>();
}
template <typename T>
int S<int>::i[] = { 1, 2 };
Which g is called from S<int>::f()?
If the program is valid, then surely one would expect g<2> to be called.
If the program is valid, does S<T>::i have a non-dependent type in S<T>::f? If so, is it incomplete, or is it int[1]? (Here, int[1] would be surprising, since S<int>::i actually has type int[2].)
If the program is invalid, why?
For a simpler example, consider:
template <typename T>
struct S {
static int i[];
const int N = sizeof (i);
};
This is only valid if the type of i is dependent, meaning that the sizeof expression isn't evaluated until the class is instantiated.
14.6.4 [temp.friend] paragraph 1 bullet 3 says:
if the name of the friend is a qualified-id and a matching specialization of a function template is found in the specified class or namespace, the friend declaration refers to that function template specialization, otherwise,
I'm not sure this says what it's supposed to say. For example:
namespace N {
template<class T> int f(T);
}
class A {
friend int N::f(int);
int m;
A();
};
namespace N {
template< class T > int f(T) {
A a; // ok for T=int?
return a.m; // ok for T=int?
}
}
int m = N::f(42); // ok?
char c = N::f('a'); // Clearly ill-formed.
The key is that the wording talks about a “matching specialization,” which to me means that N::f<int> is befriended only if that specialization existed in N before the friend declaration. So it's ill-formed as written, but if we move the call to N::f<int> up to a point before the definition of A, it's well-formed.
That seems surprising, especially given that the first bullet does not require a pre-existing specialization. So I suggest replacing bullet 3 with something like:
if the name of the friend is a qualified-id and a matching function template is found in the specified class or namespace, the friend declaration refers to the deduced specialization of that function template, otherwise,
In the following example, the template parameter in the partial specialization is non-deducible:
template <class T> struct A { typedef T U; };
template <class T> struct C { };
template <class T> struct C<typename A<T>::U> { };
Several compilers issue errors for this case, but there appears to be nothing in the Standard that would make this ill-formed; it simply seems that the partial specialization will never be matched, so the primary template will be used for all specializations. Should it be ill-formed?
Notes from the April, 2006 meeting:
It was noted that there are similar issues for constructors and conversion operators with non-deducible parameters, and that they should probably be dealt with similarly.
The Standard does not specify how member and nonmember function templates are to be ordered. This question arises with an example like the following:
struct A {
template<class T> void operator<<(T&);
};
template<class T> struct B { };
template<class T> void operator<<(A&, B<T>&);
int main() {
A a;
B<A> b;
a << b;
}
The two candidates for “a << b” are:
How should we treat the implicit this parameter of #1 and the explicit first parameter of #2?
Option 0: Make them unordered.
Option 1: If either function is a non-static member function, ignore any this parameter and ignore the first parameter of any non-member function. This option will select #2, as “B<T>&” is more specialized than “T&”.
Option 2: Treat the this parameter as if it were of reference to object type, and then perform comparison to the first parameter of the other function. The other function's first parameter will either be another this parameter, or it will be a by-value or by-reference object parameter. In the example above, this option will also select #2.
The difference between option 1 and option 2 can be seen in the following example:
struct A { };
template<class T> struct B {
template<typename R> int operator*(R&); // #1
};
template <typename T> int operator*(T&, A&); // #2
int main() {
A a;
B<A> b;
b * a;
}
Should this select #1, select #2, or be ambiguous? Option 1 will select #2, because “A&” is more specialized than “T&”. Option 2 will make this example ambiguous, because “B<A>&” is more specialized than “T&”.
If one were considering two non-member templates,
template <typename T> int operator*(T&, A&); // #2
template <typename T, typename R> int operator*(B<A>&, R&); // #3
the current rules would make these unordered. Option 2 thus seems more consistent with this existing behavior.
Notes from the April, 2006 meeting:
The group favored option 2.
Consider the following example:
template <class T> struct Outer {
struct Inner {
Inner* self();
};
};
template <class T> Outer<T>::Inner*
Outer<T>::Inner::self() { return this; }
According to 14.7 [temp.res] paragraph 3 (before the salient wording was inadvertently removed, see issue 559),
A qualified-id that refers to a type and in which the nested-name-specifier depends on a template-parameter (14.7.2 [temp.dep]) but does not refer to a member of the current instantiation (14.7.2.1 [temp.dep.type]) shall be prefixed by the keyword typename to indicate that the qualified-id denotes a type, forming a typename-specifier.
Because Outer<T>::Inner is a member of the current instantiation, the Standard does not currently require that it be prefixed with typename when it is used in the return type of the definition of the self() member function. However, it is difficult to parse this definition correctly without knowing that the return type is, in fact, a type, which is what the typename keyword is for. Should the Standard be changed to require typename in such contexts?
Is this program valid?
template <typename T> int g(int);
class h{};
template <typename T> int l(){h j; return g<T>(j);}
template <typename T> int g(const h&);
class j{};
int jj(){return l<j>();}
The key issue is when "g" is looked up, i.e., whether both overloaded template "g" functions are available at the call site or only the first. Clearly, the entire postfix-expression "g<T>(j)" is dependent, but when is the set of available template functions determined?
For consistency with the rules about when the set of available overloads is determined when calling a function given by an unqualified-id, I would think that we should postpone determining the set of template functions if (and only if) any of the explicit template arguments are dependent.
John Spicer: I agree that there should be a core issue for this. The definition of "dependent name" (14.7.2 [temp.dep] paragraph 1) should probably be modified to cover this case. It currently only handles cases where the function name is a simple identifier.
Notes from the March 2004 meeting:
A related issue is a call with a qualified name and dependent arguments, e.g., x::y(depa, depb).
The list of cases in 14.7.1 [temp.local] about when a template parameter is hidden seems to be incomplete.
Consider
// example-1
struct S {
int C;
template<class> void f();
};
template<class C>
void S::f()
{
C c; // #1
}
Someone asked whether line #1 is well-formed and I responded "no" based on my understanding of the rules in 14.6.1. After a second looking, I've realized that the above case is currently missing from the list.
The list in 14.6.1 covers cases like
// example-2
template<class T>
struct S {
int C;
void f();
};
template<class C>
void S<C>::f()
{
C c; // ERROR: 'C' is 'S::C' not the template parameter
}
or
// example-3
struct A { int C; }
template<class C>
struct S : A {
C c; // ERROR: 'C' is 'A::C', not the template parameter
};
But the case of a 'member template' is missing. I believe it should
follow the same rule as above. The reason is this.
In the case listed in 14.6.1 (having to do with members of classes), the "algorithm" seems to be this:
I believe that any rule, coherent with 14.6.1/5 and 14.6.1/7, for covering the cases of member templates (example-1) will be described by the above "algorithm".
Am I missing something?
[1] of course, the standard text does not formally speak of "template parameter scope", but we all know that the template parameters "live" somewhere. I'm using that terminology to designate the declarative region of the template parameters.
Mike Miller: I have a somewhat different perspective on this question. I think your example-1 is fundamentally different from your example-2 and example-3. Looking, for instance, at your example-2, I see four nested scopes:
namespace scope
template scope (where the parameter is)
class S scope
S::f() block scope
Naturally, S::C hides the template parameter C. The same is true of your example-3, with three scopes:
namespace scope
template scope
class S scope (includes 10.2 base class lookup)
Again, it's clear that the C inherited from A hides the template parameter in the containing scope.
The scopes I see in your example-1, however, are different:
namespace scope
struct S scope
template scope (where the parameter is)
S::f() block scope
Here it seems clear to me that the template parameter hides the class member.
It might help to look at the case where the function template is defined inline in the class:
struct S {
int C;
template<class C> int f() {
C c; // #1
}
};
It would be pretty strange, I think, if the #1 C were the member and not the template parameter. It would also be odd if the name lookup were different between an inline definition and an out-of-line definition.
See also issue 459.
Notes from the March 2004 meeting:
Basically, the standard is okay. We think Gaby's desired cases like #1 should be ill-formed.
There is a wording problem in 14.7.1 [temp.local] paragraph 7. It says:
In the definition of a member of a class template that appears outside of the class template definition, the name of a member of this template hides the name of a template-parameter.
It should say "hides the name of a template-parameter of the class template (but not a template-parameter of the member, if the member is itself a template)" or words to that effect.
The Standard is currently silent on the dependency status of enumerations and enumerators that are members of class templates. There are three questions that must be answered in this regard:
Are enumeration members of class templates dependent types?
It seems clear that nested enumerations must be dependent. For example:
void f(int);
template<typename T> struct S {
enum E { e0 };
void g() {
f(e0);
}
};
void f(S<int>::E);
void x() {
S<int> si;
si->g(); // Should call f(S<int>::E)
}
Is sizeof applied to a nested enumeration a value-dependent expression (14.7.2.3 [temp.dep.constexpr])?
There are three distinct cases that might have different answers to this question:
template<typename T> struct S {
enum E { e0 };
};
Here, the size of E is, in principle, known at the time the template is defined.
template<short I> struct S {
enum E { e0 = I };
};
In this case, the minimum size required for E cannot be determined until instantiation, but it is clear that the underlying type need be no larger than short.
template<typename T> struct S {
enum E { e0 = T::e0; };
}
Here, nothing can be known about the size of E at the time the template is defined.
14.7.2.3 [temp.dep.constexpr] paragraph 2 says that a sizeof expression is value-dependent if the type of the operand is type-dependent. Unless enumerations are given special treatment, all three of these examples will have value-dependent sizes. This could be surprising for the first case, at least, if not the second as well.
Are nested enumerators value-dependent expressions?
Again the question of dependent initializers comes into play. As an example, consider:
template<short I> struct S {
enum E { e0, e1 = I, e2 };
};
There seem to be three possible approaches as to whether the enumerators of E are value-dependent:
The enumerators of a nested enumeration are all value-dependent, regardless of whether they have a value-dependent initializer or not. This is the current position of 14.7.2.3 [temp.dep.constexpr] paragraph 2, which says that an identifier is value-dependent if it is a name declared with a dependent type.
The enumerators of a nested enumeration are all value-dependent if any of the enumeration's enumerators has a value-dependent initializer. In this approach, e0 would be value-dependent, even though it is clear that it has the value 0.
An enumerator of a nested enumeration is value-dependent only if it has a value-dependent initializer (explict or implicit). This approach would make e1 and e2 value-dependent, but not e0.
An example that bears on the third approach is the following:
template<typename T> struct S {
enum E { N = UINT_MAX, O = T::O };
int a[N + 2];
};
With the normal treatment of enumerations, the type of a might be either int[UINT_MAX+2] or int[1], depending on whether the value of T::O was such that the underlying type of E is unsigned int or long.
One possibility for addressing this problem under the third
approach would be to treat a given enumerator as having the type of
its initializer in such cases, rather than the enumeration type. This
would be similar to the way enumerators are treated within the
enumerator list, before the enumeration declaration is complete
(
Notes from the April, 2005 meeting:
The CWG agreed on the following positions:
Nested enumerations are dependent types.
The result of the sizeof operator applied to a nested enumeration is value-dependent unless there are no dependent initializers in its definition; the first case above is not dependent, while the second and third are dependent.
The approach described in 3.C above is correct. This is similar to the treatment of static const integral data members, which are dependent only if their initializer is dependent.
Notes from the October, 2005 meeting:
There was no consensus among the CWG regarding question #3 (which enumerators should be considered value-dependent). The argument in favor of 3.C is principally that the values of enumerators with non-dependent initializers are known at definition time, so there is no need to treat them as dependent.
One objection to 3.C is that, according to the consensus of the CWG, the enumeration type is dependent and thus even the known values of the enumeration would have a dependent type, which could affect the results when such enumerations are used in expressions. A possible response to this concern would be to treat non-dependent initializers as having the type of the initializer rather than the enumeration type, similar to the treatment of enumerators within the enumerator-list (7.2 [dcl.enum] paragraph 5). However, this approach would be inconsistent with the treatment of other enumeration types. It would also interfere with overload resolution (e.g., the call in the example under question #1 above would resolve to f(int) with this approach rather than f(S<int>::E)).
Those in favor of option 3.A also suggested that it would be simpler and require less drafting: if all the enumerators have the (dependent) type of the enumeration, 14.7.2.3 [temp.dep.constexpr] paragraph 2 already says that a name with a dependent type is value-dependent, so nothing further would need to be said. Option 3.C would require additional caveats to exempt some enumerators.
The proponents of 3.A also pointed out that there are many other cases where a known value with a dependent type is treated as dependent:
static const T t = 0;
... A<t> ...
or
template <int I> void f() {
g(I-I);
}
With regard to current practice, g++ and MSVC++ implement 3.A, while EDG implements 3.C.
Notes from the July, 2009 meeting:
The consensus of the CWG was that all the types and values are dependent.
template <class T> class Foo {
public:
typedef int Bar;
Bar f();
};
template <class T> typename Foo<T>::Bar Foo<T>::f() { return 1;}
--------------------
In the class template definition, the declaration of the member function
is interpreted as:
int Foo<T>::f();In the definition of the member function that appears outside of the class template, the return type is not known until the member function is instantiated. Must the return type of the member function be known when this out-of-line definition is seen (in which case the definition above is ill-formed)? Or is it OK to wait until the member function is instantiated to see if the type of the return type matches the return type in the class template definition (in which case the definition above is well-formed)?
Suggested resolution: (John Spicer)
My opinion (which I think matches several posted on the reflector recently) is that the out-of-class definition must match the declaration in the template. In your example they do match, so it is well formed.
I've added some additional cases that illustrate cases that I think either are allowed or should be allowed, and some cases that I don't think are allowed.
template <class T> class A { typedef int X; };
template <class T> class Foo {
public:
typedef int Bar;
typedef typename A<T>::X X;
Bar f();
Bar g1();
int g2();
X h();
X i();
int j();
};
// Declarations that are okay
template <class T> typename Foo<T>::Bar Foo<T>::f()
{ return 1;}
template <class T> typename Foo<T>::Bar Foo<T>::g1()
{ return 1;}
template <class T> int Foo<T>::g2() { return 1;}
template <class T> typename Foo<T>::X Foo<T>::h() { return 1;}
// Declarations that are not okay
template <class T> int Foo<T>::i() { return 1;}
template <class T> typename Foo<T>::X Foo<T>::j() { return 1;}
In general, if you can match the declarations up using only information
from the template, then the declaration is valid.
Declarations like Foo::i and Foo::j are invalid because for a given instance of A<T>, A<T>::X may not actually be int if the class is specialized.
This is not a problem for Foo::g1 and Foo::g2 because for any instance of Foo<T> that is generated from the template you know that Bar will always be int. If an instance of Foo is specialized, the template member definitions are not used so it doesn't matter whether a specialization defines Bar as int or not.
Implementations differ in their treatment of the following code:
template <class T>
struct A {
typename T::X x;
};
template <class T>
struct B {
typedef T* X;
A<B> a;
};
int main ()
{
B<int> b;
}
Some implementations accept it. At least one rejects it because the instantiation of A<B<int> > requires that B<int> be complete, and it is not at the point at which A<B<int> > is being instantiated.
Erwin Unruh:
In my view the programm is ill-formed. My reasoning:
So each class needs the other to be complete.
The problem can be seen much easier if you replace the typedef with
typedef T (*X) [sizeof(B::a)];
Now you have a true recursion. The compiler cannot easily distinguish between a true recursion and a potential recursion.
John Spicer:
Using a class to form a qualified name does not require the class to be complete, it only requires that the named member already have been declared. In other words, this kind of usage is permitted:
class A {
typedef int B;
A::B ab;
};
In the same way, once B has been declared in A, it is also visible to any template that uses A through a template parameter.
The standard could be more clear in this regard, but there are two notes that make this point. Both 3.4.3.1 [class.qual] and 5.1.1 [expr.prim.general] paragraph 7 contain a note that says "a class member can be referred to using a qualified-id at any point in its potential scope (3.3.7 [basic.scope.class])." A member's potential scope begins at its point of declaration.
In other words, a class has three states: incomplete, being completed, and complete. The standard permits a qualified name to be used once a name has been declared. The quotation of the notes about the potential scope was intended to support that.
So, in the original example, class A does not require the type of T to be complete, only that it have already declared a member X.
Bill Gibbons:
The template and non-template cases are different. In the non-template case the order in which the members become declared is clear. In the template case the members of the instantiation are conceptually all created at the same time. The standard does not say anything about trying to mimic the non-template case during the instantiation of a class template.
Mike Miller:
I think the relevant specification is 14.7.4.1 [temp.point] paragraph 3, dealing with the point of instantiation:
For a class template specialization... if the specialization is implicitly instantiated because it is referenced from within another template specialization, if the context from which the specialization is referenced depends on a template parameter, and if the specialization is not instantiated previous to the instantiation of the enclosing template, the point of instantiation is immediately before the point of instantiation of the enclosing template. Otherwise, the point of instantiation for such a specialization immediately precedes the namespace scope declaration or definition that refers to the specialization.
That means that the point of instantiation of A<B<int> > is before that of B<int>, not in the middle of B<int> after the declaration of B::X, and consequently a reference to B<int>::X from A<B<int> > is ill-formed.
To put it another way, I believe John's approach requires that there be an instantiation stack, with the results of partially-instantiated templates on the stack being available to instantiations above them. I don't think the Standard mandates that approach; as far as I can see, simply determining the implicit instantiations that need to be done, rewriting the definitions at their respective points of instantiation with parameters substituted (with appropriate "forward declarations" to allow for non-instantiating references), and compiling the result normally should be an acceptable implementation technique as well. That is, the implicit instantiation of the example (using, e.g., B_int to represent the generated name of the B<int> specialization) could be something like
struct B_int;
struct A_B_int {
B_int::X x; // error, incomplete type
};
struct B_int {
typedef int* X;
A_B_int a;
};
Notes from 10/01 meeting:
This was discussed at length. The consensus was that the template case should be treated the same as the non-template class case it terms of the order in which members get declared/defined and classes get completed.
Proposed resolution:
In 14.7.4.1 [temp.point] paragraph 3 change:
the point of instantiation is immediately before the point of instantiation of the enclosing template. Otherwise, the point of instantiation for such a specialization immediately precedes the namespace scope declaration or definition that refers to the specialization.
To:
the point of instantiation is the same as the point of instantiation of the enclosing template. Otherwise, the point of instantiation for such a specialization immediately precedes the nearest enclosing declaration. [Note: The point of instantiation is still at namespace scope but any declarations preceding the point of instantiation, even if not at namespace scope, are considered to have been seen.]
Add following paragraph 3:
If an implicitly instantiated class template specialization, class member specialization, or specialization of a class template references a class, class template specialization, class member specialization, or specialization of a class template containing a specialization reference that directly or indirectly caused the instantiation, the requirements of completeness and ordering of the class reference are applied in the context of the specialization reference.
and the following example
template <class T> struct A {
typename T::X x;
};
struct B {
typedef int X;
A<B> a;
};
template <class T> struct C {
typedef T* X;
A<C> a;
};
int main ()
{
C<int> c;
}
Notes from the October 2002 meeting:
This needs work. Moved back to drafting status.
Three points have been raised where the wording in 14.8.1 [temp.inst] may not be sufficiently clear.
A class template specialization is implicitly instantiated... if the completeness of the class type affects the semantics of the program...
It is not clear what it means for the "completeness... [to affect] the semantics." Consider the following example:
template<class T> struct A;
extern A<int> a;
void *foo() { return &a; }
template<class T> struct A
{
#ifdef OPTION
void *operator &() { return 0; }
#endif
};
The question here is whether it is necessary for template class A to declare an operator & for the semantics of the program to be affected. If it does not do so, the meaning of &a will be the same whether the class is complete or not and thus arguably the semantics of the program are not affected.
Presumably what was intended is whether the presence or absence of certain member declarations in the template class might be relevant in determining the meaning of the program. A clearer statement may be desirable.
If the overload resolution process can determine the correct function to call without instantiating a class template definition, it is unspecified whether that instantiation actually takes place.
The intent of this wording, as illustrated in the example in that paragraph, is to allow a "smart" implementation not to instantiate class templates if it can determine that such an instantiation will not affect the result of overload resolution, even though the algorithm described in clause 13 [over] requires that all the viable functions be enumerated, including functions that might be found as members of specializations.
Unfortunately, the looseness of the wording allowing this latitude for implementations makes it unclear what "the overload resolution process" is — is it the algorithm in 13 [over] or something else? — and what "the correct function" is.
If an implicit instantiation of a class template specialization is required and the template is declared but not defined, the program is ill-formed.
Here, it is not clear what conditions "require" an implicit instantiation. From the context, it would appear that the intent is to refer to the conditions in paragraph 4 that cause a specialization to be instantiated.
This interpretation, however, leads to different treatment of template and non-template incomplete classes. For example, by this interpretation,
class A;
template <class T> struct TA;
extern A a;
extern TA<int> ta;
void f(A*);
void f(TA<int>*);
int main()
{
f(&a); // well-formed; undefined if A
// has operator &() member
f(&ta); // ill-formed: cannot instantiate
}
A different approach would be to understand "required" in paragraph 6 to mean that a complete type is required in the expression. In this interpretation, if an incomplete type is acceptable in the context and the class template definition is not visible, the instantiation is not attempted and the program is well-formed.
The meaning of "required" in paragraph 6 must be clarified.
Notes on 10/01 meeting:
It was felt that item 1 is solved by addition of the word "might" in the resolution for issue 63; item 2 is not much of a problem; and item 3 could be solved by changing "required" to "required to be complete".
Issue 470 specified the explicit instantiation of members of explicitly-instantiated class templates. In restricting the affected members to those “whose definition is visible at the point of instantiation,” however, this resolution introduced an incompatibility between explicitly instantiating a member function or static data member and explicitly instantiating the class template of which it is a member (14.8.2 [temp.explicit] paragraph 3 requires only that the class template definition, not that of the member function or static data member, be visible at the point of the explicit instantiation). It would be better to treat the member instantiations the same, regardless of whether they are directly or indirectly explicitly instantiated.
Notes from the April, 2006 meeting:
In forwarding document J16/06-0057 = WG21 N1987 to be approved by the full Committee, the CWG reaffirmed its position that explicitly instantiating a class template only explicitly instantiates those of its members that have been defined before the point of the explicit instantiation. The effect of the position advocated above would be to require all non-exported member functions to be defined in the translation unit in which the class template is explicitly instantiated (cf paragraph 4), and we did not want to require that. We did agree that the “visible” terminology should be replaced by wording along the lines of “has been defined.”
14.8.2 [temp.explicit] paragraph 1 says,
An explicit instantiation of a function template shall not use the inline or constexpr specifiers.
This wording should be revised to apply to member functions of class templates as well.
Paragraph 17 of 14.8.3 [temp.expl.spec] says,
A member or a member template may be nested within many enclosing class templates. In an explicit specialization for such a member, the member declaration shall be preceded by a template<> for each enclosing class template that is explicitly specialized.
This is curious, because paragraph 3 only allows explicit specialization of members of implicitly-instantiated class specializations, not explicit specializations. Furthermore, paragraph 4 says,
Definitions of members of an explicitly specialized class are defined in the same manner as members of normal classes, and not using the explicit specialization syntax.
Paragraph 18 provides a clue for resolving the apparent contradiction:
In an explicit specialization declaration for a member of a class template or a member template that appears in namespace scope, the member template and some of its enclosing class templates may remain unspecialized, except that the declaration shall not explicitly specialize a class member template if its enclosing class templates are not explicitly specialized as well. In such explicit specialization declaration, the keyword template followed by a template-parameter-list shall be provided instead of the template<> preceding the explicit specialization declaration of the member.
It appears from this and the following example that the phrase “explicitly specialized” in paragraphs 17 and 18, when referring to enclosing class templates, does not mean that explicit specializations have been declared for them but that their names in the qualified-id are followed by template argument lists. This terminology is confusing and should be changed.
Proposed resolution (October, 2005):
Change 14.8.3 [temp.expl.spec] paragraph 17 as indicated:
A member or a member template may be nested within many enclosing class templates. In an explicit specialization for such a member, the member declaration shall be preceded by a template<> for each enclosing class template that is explicitly specialized specialization. [Example:...
Change 14.8.3 [temp.expl.spec] paragraph 18 as indicated:
In an explicit specialization declaration for a member of a class template or a member template that appears in namespace scope, the member template and some of its enclosing class templates may remain unspecialized, except that the declaration shall not explicitly specialize a class member template if its enclosing class templates are not explicitly specialized as well that is, the template-id naming the template may be composed of template parameter names rather than template-arguments. In For each unspecialized template in such an explicit specialization declaration, the keyword template followed by a template-parameter-list shall be provided instead of the template<> preceding the explicit specialization declaration of the member. The types of the template-parameters in the template-parameter-list shall be the same as those specified in the primary template definition. In such declarations, an unspecialized template-id shall not precede the name of a template specialization in the qualified-id naming the member. [Example:...
Notes from the April, 2006 meeting:
The revised wording describing “unspecialized” templates needs more work to ensure that the parameter names in the template-id are in the correct order; the distinction between template argyments and parameters is also probably not clear enough. It might be better to replace this paragraph completely and avoid the “unspecialized” wording altogether.
The Standard does not fully describe the syntax to be used when a member of an explicitly-specialized member class or member class template is defined in namespace scope. 14.8.3 [temp.expl.spec] paragraph 4 says that the “explicit specialization syntax” (presumably referring to “template<>”) is not used in defining a member of an explicit specialization when a class template is explicitly specialized as a class. However, nothing is said anywhere about how to define a member of a specialization when:
the entity being specialized is a class (member of a template class) rather than a class template.
the result of the specialization is a class template rather than a class (cf 14.8.3 [temp.expl.spec] paragraph 18, which describes this case as a “member template that... remain[s] unspecialized”).
(See paper J16/05-0148 = WG21 N1888 for further details, including a survey of existing implementation practice.)
Notes from the October, 2005 meeting:
The CWG felt that the best approach, balancing consistency with implementation issues and existing practice, would be to require that template<> be used when defining members of all explicit specializations, including those currently covered by 14.8.3 [temp.expl.spec] paragraph 4.
Given
template <class T> static T f(T t) { ... }
template <> int f(int t) { ... }
what is the linkage of f(int)?
Section 14 [temp] paragraph 4 says,
Entities generated from a template with internal linkage are distinct from all entities generated in other translation units.
But is the explicit specialization “generated from” the primary template? Does it inherit the local linkage? If so, where do I find a reference saying so explicitly?
James Widman: Data points: EDG 3.8 inherits, GCC 4.0 does not.
Mike Miller: There's a pretty strong presumption that the linkage of an explicit specialization cannot be different from that of its primary template, given that storage class specifiers cannot appear in an explicit specialization (7.1.1 [dcl.stc] paragraph 1).
Notes from the April, 2007 meeting:
The CWG agreed that the linkage of an explicit specialization must be that of the template. Gabriel dos Reis will investigate the reason for the different behavior of g++.
It does not appear that the following example is well-formed, although most compilers accept it:
template <typename T> T foo();
template <> int foo();
The reason is that 14.8.3 [temp.expl.spec] paragraph 11 only allows trailing template-arguments to be omitted if they “can be deduced from the function argument type,” and there are no function arguments in this example.
14.8.3 [temp.expl.spec] should probably say “function type” instead of “function argument type.” Also, a subsection should probably be added to 14.9.2 [temp.deduct] to cover “Deducing template arguments from declarative contexts” or some such. It would be essentially the same as 14.9.2.2 [temp.deduct.funcaddr] except that the function type from the declaration would be used as the type of P.
Proposed resolution (March, 2008):
Insert the following as a new subsection after 14.9.2.5 [temp.deduct.type]:
14.9.2.6 Deducing template arguments in a declaration that names a specialization of a function template [temp.deduct.funcdecl] Template arguments can be deduced from the function type specified when declaring a specialization of a function template. [Note: this can occur in the context of an explicit specialization, an explicit instantiation, or a friend declaration. —end note] The function template's function type and the declared type are used as the types of P and A, and the deduction is done as described in 14.9.2.5 [temp.deduct.type].
Change 14.8.3 [temp.expl.spec] paragraph 11 as follows:
A trailing template-argument can be left unspecified in the template-id naming an explicit function template specialization provided it can be deduced from the function argument type (14.9.2.6 [temp.deduct.funcdecl])...
Notes from the September, 2008 meeting:
The proposed resolution is probably more than is needed. Instead of a complete new section, the material could become a paragraph in 14.6.6 [temp.fct].
According to 14.8.3 [temp.expl.spec] paragraph 14,
An explicit specialization of a function template is inline only if it is explicitly declared to be...
This could be read to require that the inline keyword must appear in the declaration. However, 8.4 [dcl.fct.def] paragraph 10 says that a deleted function is implicitly inline, so it should be made clear that defining an explicit specialization as deleted makes it inline.
According to 14.8.3 [temp.expl.spec] paragraph 1, only non-deleted function templates may be explicitly specialized. There doesn't appear to be a compelling need for this restriction, however, and it could be useful to forbid use of implicitly-instantiated specializations while still allowing use of explicitly-specialized versions.
The last two sentences of 14.9.2 [temp.deduct] paragraph 5 read:
When all template arguments have been deduced or obtained from default template arguments, all uses of template parameters in non-deduced 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.
Shouldn't the substitution occur for all uses of the parameters, so that any of them could result in deduction failure?
Proposed resolution (October, 2006):
Change 14.9.2 [temp.deduct] paragraph 5 as follows:
...When all template arguments have been deduced or obtained from default template arguments, all uses of template parameters in non-deduced contexts the function type are replaced with the corresponding deduced or default argument values. If the substitution results in an invalid type, as described above, type deduction fails.
Notes from the September, 2008 meeting:
This issue was returned to "drafting" status in order to coordinate the wording with the concepts proposal.
The current rules in 14.9.2 [temp.deduct] say that type deduction fails as a result of attempting to use a type that is not a class type in a qualified name. However, it is now possible to use enumeration names as nested-name-specifiers, so this rule needs to be updated accordingly.
An expression used in an if statement is implicitly converted to type bool (6.4 [stmt.select]). According to the rules of template argument deduction for conversion functions given in 14.9.2.3 [temp.deduct.conv], the following example is ill-formed:
struct X {
template<class T> operator const T&() const;
};
int main()
{
if( X() ) {}
}
Following the logic in 14.9.2.3 [temp.deduct.conv], A is bool and P is const T (because cv-qualification is dropped from P before the reference is removed), and deduction fails.
It's not clear whether this is the intended outcome or not.
Notes from the April, 2005 meeting:
The CWG observed that there is nothing special about either bool or the context in the example above; instead, it will be a problem wherever a copy occurs, because cv-qualification is always dropped in a copy operation. This appears to be a case where the conversion deduction rules are not properly symmetrical with the rules for arguments. The example should be accepted.
The rules for deducing function template arguments from a conversion function template include provisions in 14.9.2.3 [temp.deduct.conv] paragraph 2 for array and function return types, even though such types are prohibited and cannot occur in the conversion-type-id of a conversion function template. They should be removed.
Consider this program:
struct F {
template<class T>
operator const T&() { static T t; return t; }
};
int main() {
F f;
int i = f; // ill-formed
}
It's ill-formed, because according to 14.9.2.3 [temp.deduct.conv], we try to match const T with int.
(The reference got removed from P because of paragraph 3, but the const isn't removed, because paragraph 2 bullet 3 comes before paragraph 3 and thus isn't applied any more.)
Changing the declaration of the conversion operator to
operator T&() { ... }
makes the program compile, which is counter-intuitive to me: I'm in an rvalue (read-only) context, and I can use a conversion to T&, but I can't use a conversion to const T&?
Issue 226 removed the original prohibition on default template-arguments for function templates. However, the note in 14.9.2.5 [temp.deduct.type] paragraph 19 still reflects that prohibition. It should be revised or removed.
14.9.2.5 [temp.deduct.type] paragraph 22 describes how we cope with partial ordering between two function templates that differ because one has a function parameter pack while the other has a normal function parameter. However, this paragraph was meant to apply to template parameter packs as well, e.g., to help with partial ordering of class template partial specializations:
template <class T1, class ...Z> class S; // #1
template <class T1, class ...Z> class S<T1, const Z&...> {}; // #2
template <class T1, class T2> class S<T1, const T2&> {};; // #3
S<int, const int&> s; // both #2 and #3 match; #3 is more specialized
(See also issue 818.)
Proposed resolution (March, 2009):
Change 14.9.2.5 [temp.deduct.type] paragraphs 9-10 as follows (and add the example above to paragraph 9):
If P has a form that contains <T> or <i>, then each argument Pi of the respective template argument list of P is compared with the corresponding argument Ai of the corresponding template argument list of A. If the template argument list of P contains a pack expansion that is not the last template argument, the entire template argument list is a non-deduced context. If Pi is a pack expansion, then the pattern of Pi is compared with each remaining argument in the template argument list of A. Each comparison deduces template arguments for subsequent positions in the template parameter packs expanded by Pi. During partial ordering (14.9.2.4 [temp.deduct.partial]), if Ai was originally a pack expansion and Pi is not a pack expansion, or if P does not contain a template argument corresponding to Ai, argument deduction fails.
Similarly, if P has a form that contains (T), then each parameter type Pi of the respective parameter-type-list of P is compared with the corresponding parameter type Ai of the corresponding parameter-type-list of A. If the parameter-declaration corresponding to Pi is a function parameter pack, then the type of its declarator-id is compared with each remaining parameter type in the parameter-type-list of A. Each comparison deduces template arguments for subsequent positions in the template parameter packs expanded by the function parameter pack. During partial ordering (14.9.2.4 [temp.deduct.partial]), if Ai was originally a function parameter pack and Pi is not a function parameter pack, or if P does not contain a function parameter type corresponding to Ai, argument deduction fails. [Note: A function parameter pack can only occur at the end of a parameter-declaration-list (8.3.5 [dcl.fct]). —end note]
14.9.2.1 [temp.deduct.call] paragraph 3 gives the deduction of rvalue references special treatment in the context of a function call:
If P is of the form T&&, where T is a template parameter, and the argument is an lvalue, the type A& is used in place of A for type deduction.
A similar provision is needed, but is not present, in declarative contexts. For example:
template<typename T> void f(T&&);
template<> void f(int&) { } // #1
template<> void f(int&&) { } // #2
void g(int i) {
f(i); // calls f<int&>(int&), i.e., #1
f(0); // calls f<int>(int(&&), i.e., #2
}
There need to be rules that deduce the template arguments for the specializations in the same way that the arguments are deduced in the calls.
I have a question about exception handling with respect to derived to base conversions of pointers caught by reference.
What should the result of this program be?
struct S {};
struct SS : public S {};
int main()
{
SS ss;
int result = 0;
try
{
throw &ss; // throw object has type SS*
// (pointer to derived class)
}
catch (S*& rs) // (reference to pointer to base class)
{
result = 1;
}
catch (...)
{
result = 2;
}
return result;
}
The wording of 15.3 [except.handle] paragraph 3 would seem to say that the catch of S*& does not match and so the catch ... would be taken.
All of the compilers I tried (EDG, g++, Sun, and Microsoft) used the catch of S*& though.
What do we think is the desired behavior for such cases?
My initial reaction is that this is a bug in all of these compilers, but the fact that they all do the same thing gives me pause.
On a related front, if the handler changes the parameter using the reference, what is caught by a subsequent handler?
extern "C" int printf(const char *, ...);
struct S {};
struct SS : public S {};
SS ss;
int f()
{
try
{
throw &ss;
}
catch (S*& rs) // (reference to pointer to base class)
{
rs = 0;
throw;
}
catch (...)
{
}
return 0;
}
int main()
{
try { f(); }
catch (S*& rs) {
printf("rs=%p, &ss=%p\n", rs, &ss);
}
}
EDG, g++, and Sun all catch the original (unmodified) value. Microsoft catches the modified value. In some sense the EDG/g++/Sun behavior makes sense because the later catch could catch the derived class instead of the base class, which would be difficult to do if you let the catch clause update the value to be used by a subsequent catch.
But on this non-pointer case, all of the compilers later catch the modified value:
extern "C" int printf(const char *, ...);
int f()
{
try
{
throw 1;
}
catch (int& i)
{
i = 0;
throw;
}
catch (...)
{
}
return 0;
}
int main()
{
try { f(); }
catch (int& i) {
printf("i=%p\n", i);
}
}
To summarize:
(See also issue 729.)
Notes from the October, 2009 meeting:
The consensus of the CWG was that it should not be possible to catch a pointer to a derived class using a reference to a base class pointer, and that a handler that takes a reference to non-const pointer should allow the pointer to be modified by the handler.
Given the following example:
int f() {
try { /* ... */ }
catch(const int*&) {
return 1;
}
catch(int*&) {
return 2;
}
return 3;
}
can f() return 2? That is, does an int* exception object match a const int*& handler?
According to 15.3 [except.handle] paragraph 3, it does not:
A handler is a match for an exception object of type E if
The handler is of type cv T or cv T& and E and T are the same type (ignoring the top-level cv-qualifiers), or
the handler is of type cv T or cv T& and T is an unambiguous public base class of E, or
the handler is of type cv1 T* cv2 and E is a pointer type that can be converted to the type of the handler by either or both of
a standard pointer conversion (4.10 [conv.ptr]) not involving conversions to pointers to private or protected or ambiguous classes
a qualification conversion
the handler is a pointer or pointer to member type and E is std::nullptr_t.
Only the third bullet allows qualification conversions, but only the first bullet applies to a handler of reference-to-pointer type. This is consistent with how other reference bindings work; for example, the following is ill-formed:
int* p;
const int*& r = p;
(The consistency is not complete; the reference binding would be permitted if r had type const int* const &, but a handler of that type would still not match an int* exception object.)
However, implementation practice seems to be in the other direction; both EDG and g++ do match an int* with a const int*&, and the Microsoft compiler issues an error for the presumed hidden handler in the code above. Should the Standard be changed to reflect existing practice?
(See also issue 388.)
Notes from the October, 2009 meeting:
The CWG agreed that matching the exception object with a handler should, to the extent possible, mimic ordinary reference binding in cases like this.
There is no prohibition against specifying a function type in an
exception-specification, and the normal conversion of a
function type to a pointer-to-function type occurs in both
throw-expressions (
See also issue 37.
Given this piece of code and S having a user-defined ctor, at precisely which point must std::uncaught_exception() return true and where false?
try { S s0; throw s0; } catch (S s2) { }
My understanding of the semantics of the code is as follows:
Is my understanding correct?
15.1 [except.throw] paragraph 3 talks about “the exception object” when describing the semantics of the throw-expression:
a throw-expression initializes a temporary object, called the exception object...
However, 15.5.1 [except.terminate] paragraph 1 talks about “the expression to be thrown” when enumerating the conditions under which terminate() is called:
when the exception handling mechanism, after completing evaluation of the expression to be thrown but before the exception is caught (15.1 [except.throw]), calls a user function that exits via an uncaught exception...
And, 15.5.3 [except.uncaught] paragraph 1 refers to “the object to be thrown” in the description of uncaught_exception():
The function std::uncaught_exception() returns true after completing evaluation of the object to be thrown...
Are all these objects one and the same? I believe the answer is important in case the construction of the temporary exception object throws another exception.
Suppose they are the same. Then uncaught_exception() invoked from the copy ctor for s1 (from the example [above]) must return false and a new exception (e.g., bad_alloc) may be thrown and caught by a matching handler (i.e., without calling terminate()).
But if they are not the same, then uncaught_exception() invoked from the copy ctor for s1 must return true and throwing another exception would end up calling terminate(). This would, IMO, have pretty severe consequences on writing exception safe exception classes.
As in the first case, different compilers behave differently, with most compilers not calling terminate() when the ctor for the temporary exception object throws. Unfortunately, the two compilers that I trust the most do call terminate().
FWIW, my feeling is that it should be possible for the copy ctor invoked to initialize the temporary exception object to safely exit by throwing another exception, and that the new exception should be allowed to be caught without calling terminate.
Mike Miller: The way I see this, a throw-expression has an assignment-expression as an operand. This expression is “the expression to be thrown.” Evaluation of this expression yields an object; this object is “the object to be thrown.” This object is then copied to the exception object.
Martin Sebor: Here's a survey of the return value from uncaught_exception() in the various stages of exception handling, implemented by current compilers:
| expr | temp | unwind | handlr | 2nd ex | |
|---|---|---|---|---|---|
| HP aCC 6 | 0 | 0 | 1 | 0 | OK |
| Compaq C++ 6.5 | 0 | 0 | 1 | 1 | ABRT |
| EDG eccp 3.4 | 0 | 1 | 1 | 1 | ABRT |
| g++ 3.4.2 | 0 | 0 | 1 | 0 | OK |
| Intel C++ 7.0 | 0 | 0 | 1 | 0 | OK |
| MIPSpro 7.4.1 | 0 | 0 | 1 | 1 | ABRT |
| MSVC 7.0 | 0 | 0 | 1 | 0 | OK |
| SunPro 5.5 | 1 | 1 | 1 | 0 | OK |
| VisualAge 6.0 | 0 | 1 | 1 | 1 | OK |
In the table above:
| expr | is the evaluation of the assignment-expression in the throw-expression |
| temp | is the invocation of the copy ctor for the unnamed temporary exception object created by the runtime. |
| unwind | is stack unwinding. |
| handlr | is the invocation of the copy ctor in the exception-declaration in the catch handler. |
| 2nd ex | describes the behavior of the implementation when the invocation of the copy ctor for the unnamed temporary exception object [temp] throws another exception. |
Proposed resolution (October, 2004):
Change 15.1 [except.throw] paragraph 3 as follows:
A throw-expression initializes a temporary object, called the exception object, the by copying the thrown object (i.e., the result of evaluating its assignment-expression operand) to it. The type of which the exception object is determined by removing any top-level cv-qualifiers from the static type of the operand of throw and adjusting the type from “array of T” or “function returning T” to “pointer to T” or “pointer to function returning T,” respectively. [Note: the temporary object created for by a throw-expression that whose operand is a string literal is never of type char* or wchar_t*; that is, the special conversions for string literals from the types “array of const char” and “array of const wchar_t” to the types “pointer to char” and “pointer to wchar_t,” respectively (4.2 [conv.array]), are never applied to the operand of a throw-expression. —end note] The temporary is an lvalue and is used to initialize the variable named in the matching handler (15.3 [except.handle]). The type of the operand of a throw-expression shall not be an incomplete type, or a pointer to an incomplete type other than (possibly cv-qualified) void. [...]
Change the note in 15.3 [except.handle] paragraph 3 as follows:
[Note: a throw-expression operand that which is an integral constant expression of integer type that evaluates to zero does not match a handler of pointer type; that is, the null pointer constant conversions (4.10 [conv.ptr], 4.11 [conv.mem]) do not apply. —end note]
Change 15.5.1 [except.terminate] paragraph 1 bullet 1 as follows:
when the exception handling mechanism, after completing evaluation of the expression to be thrown operand of throw but before the exception is caught (15.1 [except.throw]), calls a user function that exits via an uncaught exception,
Change 15.5.3 [except.uncaught] paragraph 1 as follows:
The function std::uncaught_exception() returns true after completing evaluation of the object to be thrown operand of throw until completing the initialization of the exception-declaration in the matching handler (18.8.4 [uncaught]).
Change 18.8.4 [uncaught] paragraph 1 by adding the indicated words:
Returns: true after completing evaluation of the operand of a throw-expression until either completing initialization of the exception-declaration in the matching handler or entering unexpected() due to the throw; or after entering terminate() for any reason other than an explicit call to terminate(). [Note: This includes stack unwinding (15.2 [except.ctor]). —end note]
Notes from the April, 2005 meeting:
The CWG discussed this resolution both within the group and with other interested parties. Among the points that were made:
Martin Sebor pointed to a posting in which he argues that writing copy constructors is more difficult if an exception during the copy to the exception object will result in a call to std::terminate().
In response to a question about why the copy to the exception object is different from the copy from the exception object to the object in the exception-declaration, it was observed that the writer of the handler can avoid the second copy (by using a reference declaration), but the first copy is unavoidable.
John Spicer observed that not exiting via exception should be a design constraint for copy constructors in exception objects, regardless of whether std::terminate() is called or not.
Adopting the position that uncaught_exception() returns false during the copy to the exception object would reduce the differences between the case where that copy is elided and the case where it is performed.
Jason Merrill observed that making uncaught_exception() return false during the copy to the exception object would simplify the code generated by g++; as it currently stands, the compiler must generate code to catch exceptions during that copy so std::terminate() can be called.
Bjarne Stroustrup worried that allowing the copy constructor to throw an exception during the copy to the exception object could result in a serious and specific exception being silently transformed into a more trivial and generic one (although the CWG later noted that this risk already exists if something in the expression being thrown throws an exception before the expression completes).
The CWG felt that more input from a wider audience was necessary before a decision could be made on the appropriate resolution.
Notes from the April, 2006 meeting:
The CWG agreed with the position that std::uncaught_exception() should return false during the copy to the exception object and that std::terminate() should not be called if that constructor exits with an exception. The issue was returned to “drafting” status for rewording to reflect this position.
Additional notes (September, 2007):
Although this issue deals primarily with when std::uncaught_exception() begins to return true, the specification of when it begins to return false is also problematic. There are two parallel sections that define the meaning of std::uncaught_exception() and each has a different problem. 15.5.3 [except.uncaught] reads,
The function std::uncaught_exception() returns true after completing evaluation of the object to be thrown until completing the initialization of the exception-declaration in the matching handler (18.8.4 [uncaught]).
The problem here is that whether an exception is considered caught (the underlying condition tested by the function) is here presented in terms of having initialized the exception-declaration, while in other places it is specified by having an active handler for the exception, e.g., 15.1 [except.throw] paragraph 6:
An exception is considered caught when a handler for that exception becomes active (15.3 [except.handle]).
This distinction is important because of 15.3 [except.handle] paragraph 3:
A handler is considered active when initialization is complete for the formal parameter (if any) of the catch clause. [Note: the stack will have been unwound at that point. —end note] Also, an implicit handler is considered active when std::terminate() or std::unexpected() is entered due to a throw.
Note that there is no exception-declaration to be initialized for the std::terminate() and std::unexpected() cases; nevertheless, according to 18.8.4 [uncaught], std::uncaught_exception() is supposed to return false when one of those two functions is entered.
The specification in 18.8.4 [uncaught] is not well phrased, however, and is open to misinterpretation. It reads,
Returns: true after completing evaluation of a throw-expression until either completing initialization of the exception-declaration in the matching handler or entering unexpected() due to the throw; or after entering terminate() for any reason other than an explicit call to terminate().
The problem here is lack of parallelism: does “after entering terminate” refer to the condition for returning true or false? This would be better phrased along the lines of
Returns: true after completing evaluation of a throw-expression until a handler for the exception becomes active (15.3 [except.handle]).
The list of identifier characters specified in the C++ standard annex _N2691_.E [extendid] and the C99 standard annex D are different. The C99 standard includes more characters.
The C++ standard says that the characters are from "ISO/IEC PDTR 10176" while the C99 standard says "ISO/IEC TR 10176". I'm guessing that the PDTR is an earlier draft of the TR.
Should the list in the C++ standard be updated?
Tom Plum: In my opinion, the "identifier character" issue has not been resolved with certainty within SC22.
One critical difference in C99 was the decision to allow a compiler to accept more characters than are given in the annex. This allows for future expansion.
The broader issue concerns the venue in which the "identifier character" issue will receive ongoing resolution.
Notes from 10/00 meeting:
The core language working group expressed a strong preference (13/0/5 in favor/opposed/abstaining) that the list of identifier characters should be extensible, as is the case in C99. However, the fact that this topic is under active discussion by other bodies was deemed sufficient reason to defer any changes to the C++ specification until the situation is more stable.
Notes from October, 2005 meeting:
The working group expressed interest in the kind of approach taken by XML 1.1, in which the definition of an identifier character is done by excluding large ranges of the Unicode character set and accepting any character outside those ranges, rather than by affirmatively designating each identifier character in each language. As noted above, consideration of this issue was previously deferred pending other related standardization efforts. Clark Nelson will investigate whether these have reached a point at which progress on this issue in C++ is now possible.
Additional note (May, 2008):
Issue 663 also deals with this appendix, and the proposed resolution there is to update the table to reflect the newest available technical report, ISO/IEC TR 10176:2003. That resolution might be seen as sufficient for this issue, as well. However, that approach does not address several of the concerns mentioned in the discussion above: coordination with WG14, the extensibility of the list of identifiers, the alternative approach used in the XML specification, etc.
Is the following well-formed?
auto concept HasDestructor<typename T> {
T::~T();
}
concept_map HasDestructor<int&> { }
According to _N2914_.14.10.2.1 [concept.map.fct] paragraph 4, the destructor requirement in the concept map results in an expression x.~X(), where X is the type int&. According to 5.2.4 [expr.pseudo], this expression is ill-formed because the object type and the type-name must be the same type, but the object type cannot be a reference type (references are dropped from types used in expressions, 5 [expr] paragraph 5).
It is not clear whether this should be addressed by changing 5.2.4 [expr.pseudo] or _N2914_.14.10.2.1 [concept.map.fct].
The C++ Standard uses the phrase “indeterminate value” without defining it. C99 defines it as “either an unspecified value or a trap representation.” Should C++ follow suit?
In addition, 4.1 [conv.lval] paragraph 1 says that applying the lvalue-to-rvalue conversion to an “object [that] is uninitialized” results in undefined behavior; this should be rephrased in terms of an object with an indeterminate value.
The definition of an argument does not seem to cover many assumed use cases, and we believe that is not intentional. There should be answers to questions such as: Are lambda-captures arguments? Are type names in a throw-spec arguments? “Argument” to casts, typeid, alignof, alignas, decltype and sizeof? why in x[arg] arg is not an argument, but the value forwarded to operator[]() is? Does not apply to operators as call-points not bounded by parentheses? Similar for copy initialization and conversion? What are deduced template “arguments?” what are “default arguments?” can attributes have arguments? What about concepts, requires clauses and concept_map instantiations? What about user-defined literals where parens are not used?
According to 1.4 [intro.compliance] paragraph 7,
A freestanding implementation is one in which execution may take place without the benefit of an operating system, and has an implementation-defined set of libraries that includes certain language-support libraries (17.6.1.3 [compliance]).
This definition links two relatively separate topics: the lack of an operating system and the minimal set of libraries. Furthermore, 3.6.1 [basic.start.main] paragraph 1 says:
[Note: in a freestanding environment, start-up and termination is implementation-defined; start-up contains the execution of constructors for objects of namespace scope with static storage duration; termination contains the execution of destructors for objects with static storage duration. —end note]
It would be helpful if the two characteristics (lack of an operating system and restricted set of libraries) were named separately and if these statements were clarified to identify exactly what is implementation-defined.
Notes from the October, 2009 meeting:
The CWG felt that it needed a specific proposal in a paper before attempting to resolve this issue.
There should be a list of incompatibilities between the current and previous Standards, as in ISO/IEC TR 10176 4.1.1 paragraph 9.
(See document N2733 for an initial list of this information.)
Does the Standard require that an uninitialized auto variable have a stable (albeit indeterminate) value? That is, does the Standard require that the following function return true?
bool f() {
unsigned char i; // not initialized
unsigned char j = i;
unsigned char k = i;
return j == k; // true iff "i" is stable
}
3.9.1 [basic.fundamental]
paragraph 1
requires that uninitialized unsigned char variables have a
valid value, so the initializations of j and k are
well-formed and required not to trap. The question here is whether
the value of i is allowed to change between those
initializations.
Mike Miller: 1.9 [intro.execution] paragraph 10 says,
An instance of each object with automatic storage duration (3.7.3 [basic.stc.auto] ) is associated with each entry into its block. Such an object exists and retains its last-stored value during the execution of the block and while the block is suspended...I think that the most reasonable way to read this is that the only thing that is allowed to change the value of an automatic (non-volatile?) value is a "store" operation in the abstract machine. There are no "store" operations to i between the initializations of j and k, so it must retain its original (indeterminate but valid) value, and the result of the program is well-defined.
The quibble, of course, is whether the wording "last-stored value" should be applied to a "never-stored" value. I think so, but others might differ.
Tom Plum: 7.1.6.1 [dcl.type.cv] paragraph 8 says,
[Note: volatile is a hint to the implementation to avoid aggressive optimization involving the object because the value of the object might be changed by means undetectable by an implementation. See 1.9 [intro.execution] for detailed semantics. In general, the semantics of volatile are intended to be the same in C++ as they are in C. ]>From this I would infer that non-volatile means "shall not be changed by means undetectable by an implementation"; that the compiler is entitled to safely cache accesses to non-volatile objects if it can prove that no "detectable" means can modify them; and that therefore i shall maintain the same value during the example above.
Nathan Myers: This also has practical code-generation consequences. If the uninitialized auto variable lives in a register, and its value is really unspecified, then until it is initialized that register can be used as a temporary. Each time it's "looked at" the variable has the value that last washed up in that register. After it's initialized it's "live" and cannot be used as a temporary any more, and your register pressure goes up a notch. Fixing the uninit'd value would make it "live" the first time it is (or might be) looked at, instead.
Mike Ball: I agree with this. I also believe that it was certainly never my intent that an uninitialized variable be stable, and I would have strongly argued against such a provision. Nathan has well stated the case. And I am quite certain that it would be disastrous for optimizers. To ensure it, the frontend would have to generate an initializer, because optimizers track not only the lifetimes of variables, but the lifetimes of values assigned to those variables. This would put C++ at a significant performance disadvantage compared to other languages. Not even Java went this route. Guaranteeing defined behavior for a very special case of a generally undefined operation seems unnecessary.
According to 1.9 [intro.execution] paragraph 14, “sequenced before” is a relation between “evaluations.” However, 3.6.3 [basic.start.term] paragraph 3 says,
If the completion of the initialization of a non-local object with static storage duration is sequenced before a call to std::atexit (see <cstdlib>, 18.5 [support.start.term]), the call to the function passed to std::atexit is sequenced before the call to the destructor for the object. If a call to std::atexit is sequenced before the completion of the initialization of a non-local object with static storage duration, the call to the destructor for the object is sequenced before the call to the function passed to std::atexit. If a call to std::atexit is sequenced before another call to std::atexit, the call to the function passed to the second std::atexit call is sequenced before the call to the function passed to the first std::atexit call.
Except for the calls to std::atexit, these events do not correspond to “evaluation” of expressions that appear in the program. If the “sequenced before” relation is to be applied to them, a more comprehensive definition is needed.
According to 2.2 [lex.phases] paragraph 1, in translation phase 1,
Any source file character not in the basic source character set (2.3 [lex.charset]) is replaced by the universal-character-name that designates that character.
If a character that is not in the basic character set is preceded by a backslash character, for example
"\á"
the result is equivalent to
"\\u00e1"
that is, a backslash character followed by the spelling of the universal-character-name. This is different from the result in C99, which accepts characters from the extended source character set without replacing them with universal-character-names.
2.12 [lex.key] paragraph 2 says,
Furthermore, the alternative representations shown in Table 4 for certain operators and punctuators (2.6 [lex.digraph]) are reserved and shall not be used otherwise:
Also, 2.6 [lex.digraph] paragraph 2 says,
In all respects of the language, each alternative token behaves the same, respectively, as its primary token, except for its spelling.
It is not clear whether the following example is well-formed:
#define STR2(x) #x
#define STR(x) STR2(x)
int main() {
return sizeof STR('\0'bitor 0) - sizeof STR('\0'bitor 0);
}
In this example, bitor is not the | operator but the identifier in a user-defined-character-literal. Does this violate the restrictions of 2.12 [lex.key] and 2.6 [lex.digraph]?
According to 2.14.3 [lex.ccon] paragraph 1,
A character literal that does not begin with u, U, or L is an ordinary character literal, also referred to as a narrow-character literal. An ordinary character literal that contains a single c-char has type char, with value equal to the numerical value of the encoding of the c-char in the execution character set.
However, the definition of c-char includes as one possibility a universal-character-name. The value of a universal-character-name cannot, in general, be represented as a char, so this specification is impossible to satisfy.
(See also issue 411 for related questions.)
2.14.5 [lex.string] paragraph 5 reads
Escape sequences and universal-character-names in string literals have the same meaning as in character literals, except that the single quote ' is representable either by itself or by the escape sequence \', and the double quote " shall be preceded by a \. In a narrow string literal, a universal-character-name may map to more than one char element due to multibyte encoding.
The first sentence refers us to 2.14.3 [lex.ccon], where we read in the first paragraph that "An ordinary character literal that contains a single c-char has type char [...]." Since the grammar shows that a universal-character-name is a c-char, something like '\u1234' must have type char (and thus be a single char element); in paragraph 5, we read that "A universal-character-name is translated to the encoding, in the execution character set, of the character named. If there is no such encoding, the universal-character-name is translated to an implemenation-defined encoding."
This is in obvious contradiction with the second sentence. In addition, I'm not really clear what is supposed to happen in the case where the execution (narrow-)character set is UTF-8. Consider the character \u0153 (the oe in the French word oeuvre). Should '\u0153' be a char, with an "error" value, say '?' (in conformance with the requirement that it be a single char), or an int, with the two char values 0xC5, 0x93, in an implementation defined order (in conformance with the requirement that a character representable in the execution character set be represented). Supposing the former, should "\u0153" be the equivalent of "?" (in conformance with the first sentence), or "\xC5\x93" (in conformance with the second).
Notes from October 2003 meeting:
We decided we should forward this to the C committee and let them resolve it. Sent via e-mail to John Benito on November 14, 2003.
Reply from John Benito:
I talked this over with the C project editor, we believe this was handled by the C committee before publication of the current standard.
WG14 decided there needed to be a more restrictive rule for one-to-one mappings: rather than saying "a single c-char" as C++ does, the C standard says "a single character that maps to a single-byte execution character"; WG14 fully expect some (if not many or even most) UCNs to map to multiple characters.
Because of the fundamental differences between C and C++ character types, I am not sure the C committee is qualified to answer this satisfactorily for WG21. WG14 is willing to review any decision reached for compatibility.
I hope this helps.
(See also issue 912 for a related question.)
The list of declarations that are not definitions given in 3.1 [basic.def] paragraph 2 does not mention several plausible candidates: parameter declarations in non-defining function declarations, non-static data members, and template parameters. It might be argued that none of these are declarations (paragraph 1 does not use the syntactic non-terminal declaration but does explicitly refer to clause 7 [dcl.dcl], where that non-terminal is defined). However, the list in paragraph 2 does mention static member declarations, which also are not declarations, so the intent is not clear.
The current description of unqualified name lookup in 3.4.1 [basic.lookup.unqual] paragraph 8 does not correctly handle complex cases of nesting. The Standard currently reads,
A name used in the definition of a function that is a member function (9.3) of a class X shall be declared in one of the following ways:In particular, this formulation does not handle the following example:
- before its use in the block in which it is used or in an enclosing block (6.3), or
- shall be a member of class X or be a member of a base class of X (10.2), or
- if X is a nested class of class Y (9.7), shall be a member of Y, or shall be a member of a base class of Y (this lookup applies in turn to Y's enclosing classes, starting with the innermost enclosing class), or
- if X is a local class (9.8) or is a nested class of a local class, before the definition of class X in a block enclosing the definition of class X, or
- if X is a member of namespace N, or is a nested class of a class that is a member of N, or is a local class or nested class within a local class of a function that is a member of N, before the member function definition, in namespace N or in one of N's enclosing namespaces.
struct outer {
static int i;
struct inner {
void f() {
struct local {
void g() {
i = 5;
}
};
}
};
};
Here the reference to i is from a member function of a local
class of a member function of a nested class. Nothing in the rules
allows outer::i to be found, although intuitively it should
be found.
A more comprehensive formulation is needed that allows traversal of any combination of blocks, local classes, and nested classes. Similarly, the final bullet needs to be augmented so that a function need not be a (direct) member of a namespace to allow searching that namespace when the reference is from a member function of a class local to that function. That is, the current rules do not allow the following example:
int j; // global namespace
struct S {
void f() {
struct local2 {
void g() {
j = 5;
}
};
}
};
The description of name lookup in the parameter-declaration-clause of member functions in 3.4.1 [basic.lookup.unqual] paragraphs 7-8 is flawed in at least two regards.
First, both paragraphs 7 and 8 apply to the parameter-declaration-clause of a member function definition and give different rules for the lookup. Paragraph 7 applies to names "used in the definition of a class X outside of a member function body...," which includes the parameter-declaration-clause of a member function definition, while paragraph 8 applies to names following the function's declarator-id (see the proposed resolution of issue 41), including the parameter-declaration-clause.
Second, paragraph 8 appears to apply to the type names used in the parameter-declaration-clause of a member function defined inside the class definition. That is, it appears to allow the following code, which was not the intent of the Committee:
struct S {
void f(I i) { }
typedef int I;
};
There seems to be some confusion in the Standard regarding the relationship between 3.4.1 [basic.lookup.unqual] (Unqualified name lookup) and 3.4.2 [basic.lookup.argdep] (Argument-dependent lookup). For example, 3.4.1 [basic.lookup.unqual] paragraph 3 says,
The lookup for an unqualified name used as the postfix-expression of a function call is described in 3.4.2 [basic.lookup.argdep].
In other words, nothing in 3.4.1 [basic.lookup.unqual] applies to function names; the entire lookup is described in 3.4.2 [basic.lookup.argdep].
3.4.2 [basic.lookup.argdep] does not appear to share this view of its responsibility. The closest it comes is in 3.4.2 [basic.lookup.argdep] paragraph 2a:
...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.
Presumably, "ordinary unqualified lookup" is a reference to the processing described in 3.4.1 [basic.lookup.unqual], but, as noted above, 3.4.1 [basic.lookup.unqual] explicitly precludes applying that processing to function names. The details of "ordinary unqualified lookup" of function names are not described anywhere.
The other clauses that reference 3.4.2 [basic.lookup.argdep], clauses 13 [over] and 14 [temp], are split over the question of the relationship between 3.4.1 [basic.lookup.unqual] and 3.4.2 [basic.lookup.argdep]. 13.3.1.1.1 [over.call.func] paragraph 3, for instance, says
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]).
I.e., this reference assumes that 3.4.2 [basic.lookup.argdep] is self-contained. The same is true of 13.3.1.2 [over.match.oper] paragraph 3, second bullet:
The set of non-member candidates is the result of the unqualified lookup of operator@ in the context of the expression according to the usual rules for name lookup in unqualified function calls (3.4.2 [basic.lookup.argdep]), except that all member functions are ignored.
On the other hand, however, 14.7.4.2 [temp.dep.candidate] paragraph 1 explicitly assumes that 3.4.1 [basic.lookup.unqual] and 3.4.2 [basic.lookup.argdep] are both involved in function name lookup and do different things:
For a function call that depends on a template parameter, if the function name is an unqualified-id but not a template-id, the candidate functions are found using the usual lookup rules (3.4.1 [basic.lookup.unqual], 3.4.2 [basic.lookup.argdep]) except that:
- For the part of the lookup using unqualified name lookup (3.4.1 [basic.lookup.unqual]), only function declarations with external linkage from the template definition context are found.
- For the part of the lookup using associated namespaces (3.4.2 [basic.lookup.argdep]), only function declarations with external linkage found in either the template definition context or the template instantiation context are found.
Suggested resolution:
Change 3.4.1 [basic.lookup.unqual] paragraph 1 from
...name lookup ends as soon as a declaration is found for the name.
to
...name lookup ends with the first scope containing one or more declarations of the name.
Change the first sentence of 3.4.1 [basic.lookup.unqual] paragraph 3 from
The lookup for an unqualified name used as the postfix-expression of a function call is described in 3.4.2 [basic.lookup.argdep].
to
An unqualified name used as the postfix-expression of a function call is looked up as described below. In addition, argument-dependent lookup (3.4.2 [basic.lookup.argdep]) is performed on this name to complete the resulting set of declarations.
The last bullet of the second paragraph of section 3.4.2 [basic.lookup.argdep] says that:
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.
The first problem with this wording is that it is misleading, since one cannot get such a function argument whose type would be a template-id. The bullet should be speaking about template specializations instead.
The second problem is owing to the use of the word "defined" in the phrases "are the namespace in which the template is defined", "in which any template template arguments are defined", and "as template template arguments are defined". The bullet should use the word "declared" instead, since scenarios like the one below are possible:
namespace A {
template<class T>
struct test {
template<class U>
struct mem_templ { };
};
// declaration in namespace 'A'
template<> template<>
struct test<int>::mem_templ<int>;
void foo(test<int>::mem_templ<int>&)
{ }
}
// definition in the global namespace
template<> template<>
struct A::test<int>::mem_templ<int> {
};
int main()
{
A::test<int>::mem_templ<int> inst;
// According to the current definition of 3.4.2
// foo is not found.
foo(inst);
}
In addition, the bullet doesn't make it clear whether a T which is a class template specialization must also be treated as a class type, i.e. if the contents of the second bullet of the second paragraph of section 3.4.2 [basic.lookup.argdep].
must apply to it or not. The same stands for a T which is a function template specialization. This detail can make a difference in an example such as the one below:
- If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a member, if any; and its direct and indirect base classes. Its associated namespaces are the namespaces in which its associated classes are defined. [This wording is as updated by core issue 90.]
template<class T>
struct slist_iterator {
friend bool operator==(const slist_iterator& x, const slist_iterator& y)
{ return true; }
};
template<class T>
struct slist {
typedef slist_iterator<T> iterator;
iterator begin()
{ return iterator(); }
iterator end()
{ return iterator(); }
};
int main()
{
slist<int> my_list;
slist<int>::iterator mi1 = my_list.begin(), mi2 = my_list.end();
// Must the the friend function declaration
// bool operator==(const slist_iterator<int>&, const slist_iterator<int>&);
// be found through argument dependent lookup? I.e. is the specialization
// 'slist<int>' the associated class of the arguments 'mi1' and 'mi2'. If we
// apply only the contents of the last bullet of 3.4.2/2, then the type
// 'slist_iterator<int>' has no associated classes and the friend declaration
// is not found.
mi1 == mi2;
}
Suggested resolution:
Replace the last bullet of the second paragraph of section 3.4.2 [basic.lookup.argdep]
with
- 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.
- If T is a class template specialization, its associated namespaces and classes are those associated with T when T is regarded as a class type; 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 the primary templates making template template arguments are declared; and the classes in which any primary member templates used as template template arguments are declared.
- If T is a function template specialization, its associated namespaces and classes are those associated with T when T is regarded as a function type; 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 the primary templates making template template arguments are declared; and the classes in which any primary member templates used as template template arguments are declared.
Replace the second bullet of the second paragraph of section 3.4.2 [basic.lookup.argdep]
with
- If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a member, if any; and its direct and indirect base classes. Its associated namespaces are the namespaces in which its associated classes are defined.
- If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a member, if any; and its direct and indirect base classes. Its associated namespaces are the namespaces in which its associated classes are declared [Note: in case of any of the associated classes being a class template specialization, its associated namespace is acually the namespace containing the declaration of the primary class template of the class template specialization].
3.4.2 [basic.lookup.argdep] paragraph 2 excludes dependent parameter types and return types from consideration in determining the associated classes and namespaces of a function template. Presumably this means that an example like
namespace N {
template<class T> struct A { };
void f(void (*)());
}
template <class T>
void g(T, N::A<T>);
void g();
int main() {
f(g);
}
is ill-formed because the second parameter of the function template g does not add namespace N to the list of associated namespaces. This was probably unintentional.
Both 3.4.3.1 [class.qual] and 3.4.3.2 [namespace.qual] specify that some lookups are to be performed “in the context of the entire postfix-expression,” ignoring the fact that qualified-ids can appear outside of expressions.
It was suggested in document J16/05-0156 = WG21 N1896 that these uses be changed to “the context in which the qualified-id occurs,” but it isn't clear that this formulation adequately covers all the places a qualified-id can occur.
It is unclear to what extent entities without names match across translation units. For example,
struct S {
int :2;
enum { a, b, c } x;
static class {} *p;
};
If this declaration appears in multiple translation units, are all these members "the same" in each declaration?
A similar question can be asked about non-member declarations:
// Translation unit 1:
extern enum { d, e, f } y;
// Translation unit 2:
extern enum { d, e, f } y;
// Translation unit 3:
enum { d, e, f } y;
Is this valid C++? Is it valid C?
James Kanze: S::p cannot be defined, because to do so requires a type specifier and the type cannot be named. ::y is valid C because C only requires compatible, not identical, types. In C++, it appears that there is a new type in each declaration, so it would not be valid. This differs from S::x because the unnamed type is part of a named type — but I don't know where or if the Standard says that.
John Max Skaller: It's not valid C++, because the type is a synthesised, unique name for the enumeration type which differs across translation units, as if:
extern enum _synth1 { d,e,f} y;
..
extern enum _synth2 { d,e,f} y;
had been written.
However, within a class, the ODR implies the types are the same:
class X { enum { d } y; };
in two translation units ensures that the type of member y is the same: the two X's obey the ODR and so denote the same class, and it follows that there's only one member y and one type that it has.
(See also issues 132 and 216.)
The standard says that an unnamed class or enum definition can be given a "name for linkage purposes" through a typedef. E.g.,
typedef enum {} E;
extern E *p;
can appear in multiple translation units.
How about the following combination?
// Translation unit 1:
struct S;
extern S *q;
// Translation unit 2:
typedef struct {} S;
extern S *q;
Is this valid C++?
Also, if the answer is "yes", consider the following slight variant:
// Translation unit 1:
struct S {}; // <<-- class has definition
extern S *q;
// Translation unit 2:
typedef struct {} S;
extern S *q;
Is this a violation of the ODR because two definitions of type S consist of differing token sequences?
The following declarations are allowed within a translation unit:
struct S;
enum { S };
However, 3.5 [basic.link] paragraph 9 seems to say these two declarations cannot appear in two different translation units. That also would mean that the inclusion of a header containing the above in two different translation units is not valid C++.
I suspect this is an oversight and that users should be allowed to have the declarations above appear in different translation units. (It is a fairly common thing to do, I think.)
Mike Miller: I think you meant "enum E { S };" -- enumerators only have external linkage if the enumeration does (3.5 [basic.link] paragraph 4), and 3.5 [basic.link] paragraph 9 only applies to entities with external linkage.
I don't remember why enumerators were given linkage; I don't think it's necessary for mangling non-type template arguments. In any event, I can't think why cross-TU name collisions between enumerators and other entities would cause a problem, so I guess a change here would be okay. I can think of three changes that would have that effect:
Daveed Vandevoorde: I don't think any of these are sufficient in the sense that the problem isn't limited to enumerators. E.g.:
struct X; extern void X();shouldn't create cross-TU collisions either.
Mike Miller: So you're saying that cross-TU collisions should only be prohibited if both names denote entities of the same kind (both functions, both objects, both types, etc.), or if they are both references (regardless of what they refer to, presumably)?
Daveed Vandevoorde: Not exactly. Instead, I'm saying that if two entities (with external linkage) can coexist when they're both declared in the same translation unit (TU), then they should also be allowed to coexist when they're declared in two different translation units.
For example:
int i; void i(); // ErrorThis is an error within a TU, so I don't see a reason to make it valid across TUs.
However, "tag names" (class/struct/union/enum) can sometimes coexist with identically named entities (variables, functions & enumerators, but not namespaces, templates or type names).
Is a compiler allowed to interleave constructor calls when performing dynamic initialization of nonlocal objects? What I mean by interleaving is: beginning to execute a particular constructor, then going off and doing something else, then going back to the original constructor. I can't find anything explicit about this in clause 3.6.2 [basic.start.init].
I'll present a few different examples, some of which get a bit wild. But a lot of what this comes down to is exactly what the standard means when it talks about the order of initialization. If it says that some object x must be initialized before a particular event takes place, does that mean that x's constructor must be entered before that event, or does it mean that it must be exited before that event? If object x must be initialized before object y, does that mean that x's constructor must exit before y's constructor is entered?
(The answer to that question might just be common sense, but I couldn't find an answer in clause 3.6.2 [basic.start.init]. Actually, when I read 3.6.2 [basic.start.init] carefully, I find there are a lot of things I took for granted that aren't there.)
OK, so a few specific scenerios.
<runtime gunk>
<Enter A's constructor>
<Enter f>
<runtime gunk>
<Enter B's constructor>
<Enter f>
<Leave f>
<Leave B's constructor>
<Leave f>
<Leave A's constructor>
The implication of a 'yes' answer for users is that any function
called by a constructor, directly or indirectly, must be reentrant.At this point, you might be thinking we could avoid all of this nonsense by removing compilers' freedom to defer initialization until after the beginning of main(). I'd resist that, for two reasons. First, it would be a huge change to make after the standard has been out. Second, that freedom is necessary if we want to have support for dynamic libraries. I realize we don't yet say anything about dynamic libraries, but I'd hate to make decisions that would make such support even harder.
3.6.3 [basic.start.term] paragraph 2 says,
If a function contains a local object of static storage duration that has been destroyed and the function is called during the destruction of an object with static storage duration, the program has undefined behavior if the flow of control passes through the definition of the previously destroyed local object.
I would like to turn this behavior from undefined to well-defined behavior for the purpose of achieving a graceful shutdown, especially in a multi-threaded world.
Background: Alexandrescu describes the “phoenix singleton” in Modern C++ Design. This is a class used as a function local static, that will reconstruct itself, and reapply itself to the atexit chain, if the program attempts to use it after it is destructed in the atexit chain. It achieves this by setting a “destructed flag” in its own state in its destructor. If the object is later accessed (and a member function is called on it), the member function notes the state of the “destructed flag” and does the reconstruction dance. The phoenix singleton pattern was designed to address issues only in single-threaded code where accesses among static objects can have a non-scoped pattern. When we throw in multi-threading, and the possibility that threads can be running after main returns, the chances of accessing a destroyed static significantly increase.
The very least that I would like to see happen is to standardize what I believe is existing practice: When an object is destroyed in the atexit chain, the memory the object occupied is left in whatever state the destructor put it in. If this can only be reliably done for objects with standard layout, that would be an acceptable compromise. This would allow objects to set “I'm destructed” flags in their state and then do something well-defined if accessed, such as throw an exception.
A possible refinement of this idea is to have the compiler set up a 3-state flag around function-local statics instead of the current 2-state flag:
We have the first two states today. We might choose to add the third state, and if execution passes over a function-local static with “destroyed” state, an exception could be thrown. This would mean that we would not have to guarantee memory stability in destroyed objects of static duration.
This refinement would break phoenix singletons, and is not required for the ~mutex()/~condition() I've described and prototyped. But it might make it easier for Joe Coder to apply this kind of guarantee to his own types.
There are several problems with 3.7 [basic.stc]:
3.7 [basic.stc] paragraph 2 says that "Static and automatic storage durations are associated with objects introduced by declarations (3.1 [basic.def]) and implicitly created by the implementation (12.2 [class.temporary])."
In fact, objects "implicitly created by the implementation" are the temporaries described in (12.2 [class.temporary]), and have neither static nor automatic storage duration, but a totally different duration, described in 12.2 [class.temporary].
3.7 [basic.stc] uses the expression "local object" in several places, without ever defining it. Presumably, what is meant is "an object declared at block scope", but this should be said explicitly.
In a recent discussion in comp.lang.c++.moderated, on poster interpreted "local objects" as including temporaries. This would require them to live until the end of the block in which they are created, which contradicts 12.2 [class.temporary]. If temporaries are covered by this section, and the statement in 3.7 [basic.stc] seems to suggest, and they aren't local objects, then they must have static storage duration, which isn't right either.
I propose adding a fourth storage duration to the list after 3.7 [basic.stc] paragraph 1:
And rewriting the second paragraph of this section as follows:
Temporary storage duration is associated with objects implicitly created by the implementation, and is described in 12.2 [class.temporary]. Static and automatic storage durations are associated with objects defined by declarations; in the following, an object defined by a declaration with block scope is a local object. The dynamic storage duration is associated with objects created by the operator new.
Steve Adamczyk: There may well be an issue here, but one should bear in mind the difference between storage duration and object lifetime. As far as I can see, there is no particular problem with temporaries having automatic or static storage duration, as appropriate. The point of 12.2 [class.temporary] is that they have an unusual object lifetime.
Notes from Ocrober 2002 meeting:
It might be desirable to shorten the storage duration of temporaries to allow reuse of them. The as-if rule allows some reuse, but such reuse requires analysis, including noting whether the addresses of such temporaries have been taken.
The global allocation functions are implicitly declared in every translation unit with exception-specifications (3.7.4 [basic.stc.dynamic] paragraph 2). It is not clear what should happen if a replacement allocation function is declared without an exception-specification. Is that a conflict with the implicitly-declared function (as it would be with explicitly-declared functions, and presumably is if the <new> header is included)? Or does the new declaration replace the implicit one, including the lack of an exception-specification? Or does the implicit declaration prevail? (Regardless of the exception-specification or lack thereof, it is presumably undefined behavior for an allocation function to exit with an exception that cannot be caught by a handler of type std::bad_alloc (3.7.4.1 [basic.stc.dynamic.allocation] paragraph 3).)
3.7.4.2 [basic.stc.dynamic.deallocation] paragraph 4 mentions that the effect of using an invalid pointer value is undefined. However, the standard never says what it means to 'use' a value.
There are a number of possible interpretations, but it appears that each of them leads to undesired conclusions:
int *x = new int(0); delete x; x = 0;into undefined behaviour. As this is a common idiom, this is clearly undesirable.
int *x = new int(0); delete x; x->~int();into undefined behaviour; according to 5.2.4 [expr.pseudo], the variable x is 'evaluated' as part of evaluating the pseudo destructor call. This, in turn, would mean that all containers (23 [containers]) of pointers show undefined behaviour, e.g. 23.3.4.3 [list.modifiers] requires to invoke the destructor as part of the clear() method of the container.
If any other meaning was intended for 'using an expression', that meaning should be stated explicitly.
(See also issue 623.)
When an object is deleted, 3.7.4.2 [basic.stc.dynamic.deallocation] says that the deallocation “[renders] invalid all pointers referring to any part of the deallocated storage.” According to 3.9.2 [basic.compound] paragraph 3, a pointer whose address is one past the end of an array is considered to point to an unrelated object that happens to reside at that address. Does this need to be clarified to specify that the one-past-the-end pointer of an array is not invalidated by deleting the following object? (See also 5.3.5 [expr.delete] paragraph 4, which also mentions that the system deallocation function renders a pointer invalid.)
Consider
extern "C" int printf (const char *,...);
struct Base { Base();};
struct Derived: virtual public Base {
Derived() {;}
};
Derived d;
extern Derived& obj = d;
int i;
Base::Base() {
if ((Base *) &obj) i = 4;
printf ("i=%d\n", i);
}
int main() { return 0; }
12.7 [class.cdtor] paragraph 2 makes this valid, but 3.8 [basic.life] paragraph 5 implies that it isn't valid.
Steve Adamczyk: A second issue:
extern "C" int printf(const char *,...);
struct A { virtual ~A(); int x; };
struct B : public virtual A { };
struct C : public B { C(int); };
struct D : public C { D(); };
int main() { D t; printf("passed\n");return 0; }
A::~A() {}
C::C(int) {}
D::D() : C(this->x) {}
Core issue 52 almost, but not quite, says that in evaluating "this->x" you do a cast to the virtual base class A, which would be an error according to 12.7 [class.cdtor] paragraph 2 because the base class B constructor hasn't started yet. 5.2.5 [expr.ref] should be clarified to say that the cast does need to get done.
James Kanze submitted the same issue via comp.std.c++ on 11 July 2003:
Richard Smith: Nonsense. You can use "this" perfectly happily in a constructor, just be careful that (a) you're not using any members that are not fully initialised, and (b) if you're calling virtual functions you know exactly what you're doing.
In practice, and I think in intent, you are right. However, the standard makes some pretty stringent restrictions in 3.8 [basic.life]. To start with, it says (in paragraph 1):
The lifetime of an object is a runtime property of the object. The lifetime of an object of type T begins when:(Emphasis added.) Then when we get down to paragraph 5, it says:The lifetime of an object of type T ends 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, the constructor calls has COMPLETED.
- if T is a class type with a non-trivial destructor, the destructor call STARTS, or
- the storage which the object occupies is reused or released.
Before the lifetime of an object has started but after the storage which the object will occupy has been allocated [which sounds to me like it would include in the constructor, given the text above] or, after the lifetime of an object has ended and before the storage which the object occupied is reused or released, any pointer that refers to the storage location where the object will be or was located may be used but only in limited ways. [...] If the object will be or was of a non-POD class type, the program has undefined behavior if:
[...]
- the pointer is implicitly converted to a pointer to a base class type, or [...]
I can't find any exceptions for the this pointer.
Note that calling a non-static function in the base class, or even constructing the base class in initializer list, involves an implicit conversion of this to a pointer to the base class. Thus undefined behavior. I'm sure that this wasn't the intent, but it would seem to be what this paragraph is saying.
In 3.9 [basic.types] paragraph 10, the standard makes it quite clear that volatile qualified types are PODs:
Arithmetic types (3.9.1 [basic.fundamental]), enumeration types, pointer types, and pointer to member types (3.9.2 [basic.compound]), and cv-qualified versions of these types (3.9.3 [basic.type.qualifier]) are collectively called scalar types. Scalar types, POD-struct types, POD-union types (clause 9 [class]), arrays of such types and cv-qualified versions of these types (3.9.3 [basic.type.qualifier]) are collectively called POD types.
However in 3.9 [basic.types] paragraph 3, the standard makes it clear that PODs can be copied “as if” they were a collection of bytes by memcpy:
For any POD type T, if two pointers to T point to distinct T objects obj1 and obj2, where neither obj1 nor obj2 is a base-class subobject, if the value of obj1 is copied into obj2, using the std::memcpy library function, obj2 shall subsequently hold the same value as obj1.
The problem with this is that a volatile qualified type may need to be copied in a specific way (by copying using only atomic operations on multithreaded platforms, for example) in order to avoid the “memory tearing” that may occur with a byte-by-byte copy.
I realise that the standard says very little about volatile qualified types, and nothing at all (yet) about multithreaded platforms, but nonetheless this is a real issue, for the following reason:
The forthcoming TR1 will define a series of traits that provide information about the properties of a type, including whether a type is a POD and/or has trivial construct/copy/assign operations. Libraries can use this information to optimise their code as appropriate, for example an array of type T might be copied with a memcpy rather than an element-by-element copy if T is a POD. This was one of the main motivations behind the type traits chapter of the TR1. However it's not clear how volatile types (or POD's which have a volatile type as a member) should be handled in these cases.
Notes from the April, 2005 meeting:
It is not clear whether the volatile qualifier actually guarantees atomicity in this way. Also, the work on the memory model for multithreading being done by the Evolution Working Group seems at this point likely to specify additional semantics for volatile data, and that work would need to be considered before resolving this issue.
3.9 [basic.types] paragraph 10 requires that a class have at least one constexpr constructor other than the copy constructor in order to be considered a literal type. However, a constexpr constructor template might be instantiated in such a way that the constexpr specifier is ignored (7.1.5 [dcl.constexpr] paragraph 5). It is therefore not known whether a class with a constexpr constructor template is a literal type or not until the constructor template is specialized, which could mean that an example like
struct IntValue {
template<typename T>
constexpr IntValue(T t) : val(t) { }
constexpr intmax_t get_value() { return val; }
private:
intmax_t val;
};
is ill-formed, because it is an error to declare a member function (like get_value()) of a non-literal class to be constexpr (7.1.5 [dcl.constexpr] paragraph 6).
3.9 [basic.types] paragraph 10 should be revised so that either a constexpr constructor or constexpr constructor template allows a class to be a literal type.
3.9.1 [basic.fundamental] does not impose a requirement on the floating point types that there be an exact representation of the value zero. This omission is significant in 4.12 [conv.bool] paragraph 1, in which any non-zero value converts to the bool value true.
Suggested resolution: require that all floating point types have an exact representation of the value zero.
3.9.1 [basic.fundamental] paragraph 2 says that
There are four signed integer types: "signed char", "short int", "int", and "long int."
This would indicate that const int is not a signed integer type.
There is no normative requirement on the ranges of the integral types, although the footnote in 3.9.1 [basic.fundamental] paragraph 2 indicates the intent (for int, at least) that they match the values given in the <climits> header. Should there be an explicit requirement of some sort?
(See also paper N1693.)
The relationship between the values representable by corresponding signed and unsigned integer types is not completely described, but 3.9 [basic.types] paragraph 4 says,
The value representation of an object is the set of bits that hold the value of type T.
and 3.9.1 [basic.fundamental] paragraph 3 says,
The range of nonnegative values of a signed integer type is a subrange of the corresponding unsigned integer type, and the value representation of each corresponding signed/unsigned type shall be the same.
I.e., the maximum value of each unsigned type must be larger than the maximum value of the corresponding signed type.
C90 doesn't have this restriction, and C99 explicitly says (6.2.6.2, paragraph 2),
For signed integer types, the bits of the object representation shall be divided into three groups: value bits, padding bits, and the sign bit. There need not be any padding bits; there shall be exactly one sign bit. Each bit that is a value bit shall have the same value as the same bit in the object representation of the corresponding unsigned type (if there are M value bits in the signed type and N in the unsigned type, then M <= N).
Unlike C++, the sign bit is not part of the value, and on an architecture that does not have native support of unsigned types, an implementation can emulate unsigned integers by simply ignoring what would be the sign bit in the signed type and be conforming.
The question is whether we intend to make a conforming implementation on such an architecture impossible. More generally, what range of architectures do we intend to support? And to what degree do we want to follow C99 in its evolution since C89?
(See paper J16/08-0141 = WG21 N2631.)
The status of rvalue references to functions is not clear in the current wording. For example, 3.10 [basic.lval] paragraph 2 says,
An lvalue refers to an object or function. Some rvalue expressions—those of (possibly cv-qualified) class or array type—also refer to objects. [Footnote: Expressions such as invocations of constructors and of functions that return a class type refer to objects, and the implementation can invoke a member function upon such objects, but the expressions are not lvalues. —end footnote]
This would tend to indicate that there are no rvalues of function type. However, 5 [expr] paragraph 6 says,
If an expression initially has the type “rvalue reference to T” (8.3.2 [dcl.ref], 8.5.3 [dcl.init.ref]), the type is adjusted to “T” prior to any further analysis, and the expression designates the object or function denoted by the rvalue reference. If the expression is the result of calling a function, whether implicitly or explicitly, it is an rvalue; otherwise, it is an lvalue.
This explicitly indicates that rvalue references to functions are possible and that, in some cases, they yield function-typed rvalues. Furthermore, _N2914_.20.2.4 [concept.operator] paragraph 20 describes the concept Callable as:
auto concept Callable<typename F, typename... Args> {
typename result_type;
result_type operator()(F&, Args...);
result_type operator()(F&&, Args...);
}
It would be strange if Callable were satisfied for a function object type but not for a function type.
However, assuming that rvalue references to functions are
intended to be supported, it is not clear how an rvalue of function
type is supposed to behave. For instance,
For an ordinary function call, the postfix expression shall be either an lvalue that refers to a function (in which case the function-to-pointer standard conversion (4.3 [conv.func]) is suppressed on the postfix expression), or it shall have pointer to function type.
From this, it appears that an rvalue of function type cannot be used in a function call. It can't be converted to a pointer to function, either, as 4.3 [conv.func] paragraph 1 says,
An lvalue of function type T can be converted to an rvalue of type “pointer to T.” The result is a pointer to the function.
(See also issues 664 and especially 690. The approach described in the latter issue, viewing rvalue references as essentially lvalues rather than as essentially rvalues, could resolve the specification problems described above by eliminating the concept of an rvalue of function type.)
Section 4.4 [conv.qual] covers the case of multi-level pointers, but does not appear to cover the case of pointers to arrays of pointers. The effect is that arrays are treated differently from simple scalar values.
Consider for example the following code: (from the thread "Pointer to array conversion question" begun in comp.lang.c++.moderated)
int main()
{
double *array2D[2][3];
double * (*array2DPtr1)[3] = array2D; // Legal
double * const (*array2DPtr2)[3] = array2DPtr1; // Legal
double const * const (*array2DPtr3)[3] = array2DPtr2; // Illegal
}
and compare this code with:-
int main()
{
double *array[2];
double * *ppd1 = array; // legal
double * const *ppd2 = ppd1; // legal
double const * const *ppd3 = ppd2; // certainly legal (4.4/4)
}
The problem appears to be that the pointed to types in example 1 are unrelated since nothing in the relevant section of the standard covers it - 4.4 [conv.qual] does not mention conversions of the form "cv array of N pointer to T" into "cv array of N pointer to cv T"
It appears that reinterpret_cast is the only way to perform the conversion.
Suggested resolution:
Artem Livshits proposed a resolution :-
"I suppose if the definition of "similar" pointer types in 4.4 [conv.qual] paragraph 4 was rewritten like this:
it would address the problem.T1 is cv1,0 P0 cv1,1 P1 ... cv1,n-1 Pn-1 cv1,n T
and
T2 is cv1,0 P0 cv1,1 P1 ... cv1,n-1 Pn-1 cv1,n T
where Pi is either a "pointer to" or a "pointer to an array of Ni"; besides P0 may be also a "reference to" or a "reference to an array of N0" (in the case of P0 of T2 being a reference, P0 of T1 may be nothing).
In fact I guess Pi in this notation may be also a "pointer to member", so 4.4 [conv.qual]/{4,5,6,7} would be nicely wrapped in one paragraph."
It is not clear what constraints are placed on a floating point implementation by the wording of the Standard. For instance, is an implementation permitted to generate a "fused multiply-add" instruction if the result would be different from what would be obtained by performing the operations separately? To what extent does the "as-if" rule allow the kinds of optimizations (e.g., loop unrolling) performed by FORTRAN compilers?
(Moved from issue 760.)
Although it was considered and rejected as part of issue 643, more recent developments may argue in favor of allowing the use of this in a late-specified return type. In particular, declaring the return type for a forwarding function in a derived class template that invokes a member function of a dependent base class is difficult without this facility. For example:
template <typename T> struct derived: base<T> {
auto invoke() -> decltype(this->base_func()) {
return this->base_func();
}
};
A lambda with an empty capture list has identical semantics to a regular function type. By requiring this mapping we get an efficient lambda type with a known API that is also compatible with existing operating system and C library functions.
Notes from the July, 2009 meeting:
This functionality is part of the “unified function syntax” proposal and will be considered in that context.
There does not appear to be any technical difficulty that would require the restriction in 5.1.2 [expr.prim.lambda] paragraph 5 against default arguments in lambda-expressions.
There does not appear to be any technical difficulty that would require the current restriction that the return type of a lambda can be deduced only if the body of the lambda consists of a single return statement. In particular, multiple return statements could be permitted if they all return the same type.
A cast to an rvalue reference type produces an rvalue, and rvalues must have complete types (3.10 [basic.lval] paragraph 9). However, none of the sections dealing with cast operators in 5.2 [expr.post] require that the referred-to type must be complete in an rvalue reference cast.
(Note that the approach described for issue 690, in which an rvalue reference type would be essentially an lvalue instead of an rvalue, would address this issue as well, since lvalues can have incomplete types.)5.2.1 [expr.sub] paragraph 2 deals with one particular aspect of the overloaded operator[], which seems out of place. Either 5.2.1 [expr.sub] should be augmented to discuss the overloaded operator[] in general or the information in paragraph 2 should be moved into 13.5.5 [over.sub].
According to 5.2.3 [expr.type.conv] paragraphs 1 and 3 (stated directly or by reference to another section of the Standard), all the following expressions create temporaries:
T(1)
T(1, 2)
T{1}
T{}
However, paragraph 2 says,
The expression T(), where T is a simple-type-specifier or typename-specifier for a non-array complete effective object type or the (possibly cv-qualified) void type, creates an rvalue of the specified type, which is value-initialized (8.5 [dcl.init]; no initialization is done for the void() case).
This does not say that the result is a temporary, which means that the lifetime of the result is not specified by 12.2 [class.temporary]. Presumably this is just an oversight.
Notes from the October, 2009 meeting:
The specification in 5.2.3 [expr.type.conv] is in error, not because it fails to state that T() is a temporary but because it requires a temporary for the other cases with fewer than two operands. The case where T is a class type is covered by 12.2 [class.temporary] paragraph 1 (“a conversion that creates an rvalue”), and a temporary should not be created when T is not a class type.
Given the following declarations:
struct S {
signed long long sll: 3;
};
S s = { -1 };
the expressions s.sll-- < 0u and s.sll < 0u have different results. The reason for this is that s.sll-- is an rvalue of type signed long long (5.2.6 [expr.post.incr]), which means that the usual arithmetic conversions (5 [expr] paragraph 10) convert 0u to signed long long and the result is true. s.sll, on the other hand, is a bit-field lvalue, which is promoted (4.5 [conv.prom] paragraph 3) to int; both operands of < have the same rank, so s.sll is converted to unsigned int to match the type of 0u and the result is false. This disparity seems undesirable.
The original proposed resolution for issue 160 included changing extended_type_info (5.2.8 [expr.typeid] paragraph 1, footnote 61) to std::extended_type_info. There was no consensus on whether this name ought to be part of namespace std or in a vendor-specific namespace, so the question was moved into a separate issue.
5.2.8 [expr.typeid] paragraph 4 says,
When typeid is applied to a type-id, the result refers to a std::type_info object representing the type of the type-id. If the type of the type-id is a reference type, the result of the typeid expression refers to a std::type_info object representing the referenced type. If the type of the type-id is a class type or a reference to a class type, the class shall be completely-defined.
I'm wondering whether this is not overly restrictive. I can't think of a reason to require that T be completely-defined in typeid(T) when T is a class type. In fact, several popular compilers enforce that restriction for typeid(T), but not for typeid(T&). Can anyone explain this?
Nathan Sidwell: I think this restriction is so that whenever the compiler has to emit a typeid object of a class type, it knows what the base classes are, and can therefore emit an array of pointers-to-base-class typeids. Such a tree is necessary to implement dynamic_cast and exception catching (in a commonly implemented and obvious manner). If the class could be incomplete, the compiler might have to emit a typeid for incomplete Foo in one object file and a typeid for complete Foo in another object file. The compilation system will then have to make sure that (a) those compare equal and (b) the complete Foo gets priority, if that is applicable.
Unfortunately, there is a problem with exceptions that means there still can be a need to emit typeids for incomplete class. Namely one can throw a pointer-to-pointer-to-incomplete. To implement the matching of pointer-to-derived being caught by pointer-to-base, it is necessary for the typeid of a pointer type to contain a pointer to the typeid of the pointed-to type. In order to do the qualification matching on a multi-level pointer type, one has a chain of pointer typeids that can terminate in the typeid of an incomplete type. You cannot simply NULL-terminate the chain, because one must distinguish between different incomplete types.
Dave Abrahams: So if implementations are still required to be able to do it, for all practical purposes, why aren't we letting the user have the benefits?
Notes from the April, 2006 meeting:
There was some concern expressed that this might be difficult under the IA64 ABI. It was also observed that while it is necessary to handle exceptions involving incomplete types, there is no requirement that the RTTI data structures be used for exception handling.
During the discussion of issue 799, which specified the result of using reinterpret_cast to convert an operand to its own type, it was observed that it is probably reasonable to allow reinterpret_cast between any two types that have the same size and alignment.
5.3.1 [expr.unary.op] paragraph 2 indicates that the type of an address-of-member expression reflects the class in which the member was declared rather than the class identified in the nested-name-specifier of the qualified-id. This treatment is unintuitive and can lead to strange code and unexpected results. For instance, in
struct B { int i; };
struct D1: B { };
struct D2: B { };
int (D1::* pmD1) = &D2::i; // NOT an error
More seriously, template argument deduction can give surprising
results:
struct A {
int i;
virtual void f() = 0;
};
struct B : A {
int j;
B() : j(5) {}
virtual void f();
};
struct C : B {
C() { j = 10; }
};
template <class T>
int DefaultValue( int (T::*m) ) {
return T().*m;
}
... DefaultValue( &B::i ) // Error: A is abstract
... DefaultValue( &C::j ) // returns 5, not 10.
Suggested resolution: 5.3.1 [expr.unary.op] should be changed to read,
If the member is a nonstatic member (perhaps by inheritance) of the class nominated by the nested-name-specifier of the qualified-id having type T, the type of the result is "pointer to member of class nested-name-specifier of type T."and the comment in the example should be changed to read,
// has type int B::*
Notes from 04/00 meeting:
The rationale for the current treatment is to permit the widest possible use to be made of a given address-of-member expression. Since a pointer-to-base-member can be implicitly converted to a pointer-to-derived-member, making the type of the expression a pointer-to-base-member allows the result to initialize or be assigned to either a pointer-to-base-member or a pointer-to-derived-member. Accepting this proposal would allow only the latter use.
Additional notes:
Another problematic example has been mentioned:
class Base {
public:
int func() const;
};
class Derived : public Base {
};
template<class T>
class Templ {
public:
template<class S>
Templ(S (T::*ptmf)() const);
};
void foo()
{
Templ<Derived> x(&Derived::func); // ill-formed
}
In this example, even though the conversion of &Derived::func to int (Derived::*)() const is permitted, the initialization of x cannot be done because template argument deduction for the constructor fails.
If the suggested resolution were adopted, the amount of code broken by the change might be reduced by adding an implicit conversion from pointer-to-derived-member to pointer-to-base-member for appropriate address-of-member expressions (not for arbitrary pointers to members, of course).
(See also issue 247.)
Requirements for the alignment of pointers returned by new-expressions are given in 5.3.4 [expr.new] paragraph 10:
For arrays of char and unsigned char, the difference between the result of the new-expression and the address returned by the allocation function shall be an integral multiple of the most stringent alignment requirement (3.9 [basic.types]) of any object type whose size is no greater than the size of the array being created.
The intent of this wording is that the pointer returned by the new-expression will be suitably aligned for any data type that might be placed into the allocated storage (since the allocation function is constrained to return a pointer to maximally-aligned storage). However, there is an implicit assumption that each alignment requirement is an integral multiple of all smaller alignment requirements. While this is probably a valid assumption for all real architectures, there's no reason that the Standard should require it.
For example, assume that int has an alignment requirement of 3 bytes and double has an alignment requirement of 4 bytes. The current wording only requires that a buffer that is big enough for an int or a double be aligned on a 4-byte boundary (the more stringent requirement), but that would allow the buffer to be allocated on an 8-byte boundary — which might not be an acceptable location for an int.
Suggested resolution: Change "of any object type" to "of every object type."
A similar assumption can be found in 5.2.10 [expr.reinterpret.cast] paragraph 7:
...converting an rvalue of type "pointer to T1" to the type "pointer to T2" (where ... the alignment requirements of T2 are no stricter than those of T1) and back to its original type yields the original pointer value...
Suggested resolution: Change the wording to
...converting an rvalue of type "pointer to T1" to the type "pointer to T2" (where ... the alignment requirements of T1 are an integer multiple of those of T2) and back to its original type yields the original pointer value...
The same change would also be needed in paragraph 9.
Looking up operator new in a new-expression uses a different mechanism from ordinary lookup. According to 5.3.4 [expr.new] paragraph 9,
If the new-expression begins with a unary :: operator, the allocation function's name is looked up in the global scope. Otherwise, if the allocated type is a class type T or array thereof, the allocation function's name is looked up in the scope of T. If this lookup fails to find the name, or if the allocated type is not a class type, the allocation function's name is looked up in the global scope.
Note in particular that the scope in which the new-expression occurs is not considered. For example,
void f() {
void* operator new(std::size_t, void*);
int* i = new int; // okay?
}
In this example, the implicit reference to operator new(std::size_t) finds the global declaration, even though the block-scope declaration of operator new with a different signature would hide it from an ordinary reference.
This seems strange; either the block-scope declaration should be ill-formed or it should be found by the lookup.
Notes from October 2004 meeting:
The CWG agreed that the block-scope declaration should not be found by the lookup in a new-expression. It would, however, be found by ordinary lookup if the allocation function were invoked explicitly.
(See also issue 256.)
An implementation may have an unspecified amount of array allocation overhead (5.3.4 [expr.new] paragraph 10), so that evaluation of a new-expression in which the new-type-id is T[n] involves a request for more than n * sizeof(T) bytes of storage through the relevant operator new[] function.
The placement operator new[] function does not and cannot check whether the requested size is less than or equal to the size of the provided region of memory (18.6.1.3 [new.delete.placement] paragraphs 5-6). A program using placement array new must calculate what the requested size will be in advance.
Therefore any program using placement array new must take into account the implementation's array allocation overhead, which cannot be obtained or calculated by any portable means.
Notes from the April, 2005 meeting:
While the CWG agreed that there is no portable means to accomplish this task in the current language, they felt that a paper is needed to analyze the numerous mechanisms that might address the problem and advance a specific proposal. There is no volunteer to write such a paper at this time.
5.3.4 [expr.new] paragraph 10 says that the result of an array allocation function and the value of the array new-expression from which it was invoked may be different, allowing for space preceding the array to be used for implementation purposes such as saving the number of elements in the array. However, there is no corresponding description of the relationship between the operand of an array delete-expression and the argument passed to its deallocation function.
3.7.4.2 [basic.stc.dynamic.deallocation] paragraph 3 does state that
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.
This statement might be read as requiring an implementation, when processing an array delete-expression and calling the deallocation function, to perform the inverse of the calculation applied to the result of the allocation function to produce the value of the new-expression. (5.3.5 [expr.delete] paragraph 2 requires that the operand of an array delete-expression "be the pointer value which resulted from a previous array new-expression.") However, it is not completely clear whether the "shall" expresses an implementation requirement or a program requirement (or both). Furthermore, there is no direct statement about user-defined deallocation functions.
Suggested resolution: A note should be added to 5.3.5 [expr.delete] to clarify that any offset added in an array new-expression must be subtracted in the array delete-expression.
The meaning of an old-style cast is described in terms of const_cast, static_cast, and reinterpret_cast in 5.4 [expr.cast] paragraph 5. Ignoring const_cast for the moment, it basically says that if the conversion performed by a given old-style cast is one of those performed by static_cast, the conversion is interpreted as if it were a static_cast; otherwise, it's interpreted as if it were a reinterpret_cast, if possible. The following example is given in illustration:
struct A {};
struct I1 : A {};
struct I2 : A {};
struct D : I1, I2 {};
A *foo( D *p ) {
return (A*)( p ); // ill-formed static_cast interpretation
}
The obvious intent here is that a derived-to-base pointer conversion is one of the conversions that can be performed using static_cast, so (A*)(p) is equivalent to static_cast<A*>(p), which is ill-formed because of the ambiguity.
Unfortunately, the description of static_cast in 5.2.9 [expr.static.cast] does NOT support this interpretation. The problem is in the way 5.2.9 [expr.static.cast] lists the kinds of casts that can be performed using static_cast. Rather than saying something like "All standard conversions can be performed using static_cast," it says
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.
Given the declarations above, the hypothetical declaration
A* t(p);
is NOT well-formed, because of the ambiguity. Therefore the old-style cast (A*)(p) is NOT one of the conversions that can be performed using static_cast, and (A*)(p) is equivalent to reinterpret_cast<A*>(p), which is well-formed under 5.2.10 [expr.reinterpret.cast] paragraph 7.
Other situations besides ambiguity which might raise similar questions include access violations, casting from virtual base to derived, and casting pointers-to-members when virtual inheritance is involved.
In C, this is ill-formed (cf C99 6.5.8):
void f(char* s) {
if (s < 0) { }
}
...but in C++, it's not. Why? Who would ever need to write (s > 0) when they could just as well write (s != 0)?
This has been in the language since the ARM (and possibly earlier); apparently it's because the pointer conversions (4.10 [conv.ptr]) need to be performed on both operands whenever one of the operands is of pointer type. So it looks like the "null-ptr-to-real-pointer-type" conversion is hitching a ride with the other pointer conversions.
6.4.1 [stmt.if] is silent about whether the else clause of an if statement is executed if the condition is not evaluated. (This could occur via a goto or a longjmp.) C99 covers the goto case with the following provision:
If the first substatement is reached via a label, the second substatement is not executed.
It should probably also be stated that the condition is not evaluated when the “then” clause is entered directly.
7 [dcl.dcl] paragraph 3 reads,
In a simple-declaration, the optional
init-declarator-list can be omitted only when... the
decl-specifier-seq contains either a class-specifier,
an elaborated-type-specifier with a class-key
(9.1 [class.name]
), or an
enum-specifier. In these cases and whenever a
class-specifier or enum-specifier is present in the
decl-specifier-seq, the identifiers in those specifiers are
among the names being declared by the declaration... In such cases,
and except for the declaration of an unnamed bit-field
(9.6 [class.bit]
), the
decl-specifier-seq shall introduce one or more names into the
program, or shall redeclare a name introduced by a previous
declaration. [Example:
enum { }; // ill-formed
typedef class { }; // ill-formed
—end example]
In the absence of any explicit restrictions in
7.1.3 [dcl.typedef]
, this paragraph appears
to allow declarations like the following:
typedef struct S { }; // no declarator
typedef enum { e1 }; // no declarator
In fact, the final example in
7 [dcl.dcl]
paragraph 3 would seem to
indicate that this is intentional: since it is illustrating the
requirement that the decl-specifier-seq must introduce a name
in declarations in which the init-declarator-list is omitted,
presumably the addition of a class name would have made the example
well-formed.
On the other hand, there is no good reason to allow such declarations; the only reasonable scenario in which they might occur is a mistake on the programmer's part, and it would be a service to the programmer to require that such errors be diagnosed.
Suppose we've got this class definition:
struct X {
void f();
static int n;
};
I think I can deduce from the existing standard that the following member definitions are ill-formed:
static void X::f() { }
static int X::n;
To come to that conclusion, however, I have to put together several things in different parts of the standard. I would have expected to find an explicit statement of this somewhere; in particular, I would have expected to find it in 7.1.1 [dcl.stc]. I don't see it there, or anywhere.
Gabriel Dos Reis: Or in 3.5 [basic.link] which is about linkage. I would have expected that paragraph to say that that members of class types have external linkage when the enclosing class has an external linkage. Otherwise 3.5 [basic.link] paragraph 8:
Names not covered by these rules have no linkage.
might imply that such members do not have linkage.
Notes from the April, 2005 meeting:
The question about the linkage of class members is already covered by 3.5 [basic.link] paragraph 5.
The phrase “top-level cv-qualifier” is used numerous times in the Standard, but it is not defined. The phrase could be misunderstood to indicate that the const in something like const T& is at the “top level,” because where it appears is the highest level at which it is permitted: T& const is ill-formed.
7.1.6.3 [dcl.type.elab] paragraph 1 seems to impose an ordering constraint on the elements of friend class declarations. However, the general rule is that declaration specifiers can appear in any order. Should
class C friend;
be well-formed?
Although in most contexts “= expression” can be replaced by “{ expression }”, enumerator-definitions accept only the “=” form. This could be surprising.
Additional note (October, 2009):
The Committee may wish to consider default arguments in this like as well.
There is disagreement among implementations as to when an enumeration type is complete. For example,
enum E { e = E() };
is rejected by some and accepted by another. The Standard does not appear to resolve this question definitively.
According to 7.3 [basic.namespace] paragraph 1,
The name of a namespace can be used to access entities declared in that namespace; that is, the members of the namespace.
implying that all declarations in a namespace, including definitions of members of nested namespaces, explicit instantiations, and explicit specializations, introduce members of the containing namespace. 7.3.1.2 [namespace.memdef] paragraph 3 clarifies the intent somewhat:
Every name first declared in a namespace is a member of that namespace.
However, current changes to clarify the behavior of deleted functions (which must be deleted on their “first declaration”) state that an explicit specialization of a function template is its first declaration.
Section 7.3.3 [namespace.udecl] paragraph 8 says:
A using-declaration is a declaration and can therefore be used repeatedly where (and only where) multiple declarations are allowed.It contains the following example:
namespace A {
int i;
}
namespace A1 {
using A::i;
using A::i; // OK: double declaration
}
void f()
{
using A::i;
using A::i; // error: double declaration
}
However, if "using A::i;" is really a declaration, and not a definition, it is far from clear that repeating it should be an error in either context.
Consider:
namespace A {
int i;
void g();
}
void f() {
using A::g;
using A::g;
}
Surely the definition of f should be analogous to
void f() {
void g();
void g();
}
which is well-formed because "void g();" is a declaration and
not a definition.
Indeed, if the double using-declaration for A::i is prohibited in f, why should it be allowed in namespace A1?
Proposed Resolution (04/99): Change the comment "// error: double declaration" to "// OK: double declaration". (This should be reviewed against existing practice.)
Notes from 04/00 meeting:
The core language working group was unable to come to consensus over what kind of declaration a using-declaration should emulate. In a straw poll, 7 members favored allowing using-declarations wherever a non-definition declaration could appear, while 4 preferred to allow multiple using-declarations only in namespace scope (the rationale being that the permission for multiple using-declarations is primarily to support its use in multiple header files, which are seldom included anywhere other than namespace scope). John Spicer pointed out that friend declarations can appear multiple times in class scope and asked if using-declarations would have the same property under the "like a declaration" resolution.
As a result of the lack of agreement, the issue was returned to "open" status.
See also issues 56, 85, and 138..
Additional notes (January, 2005):
Some related issues have been raised concerning the following example (modified from a C++ validation suite test):
struct A
{
int i;
static int j;
};
struct B : A { };
struct C : A { };
struct D : virtual B, virtual C
{
using B::i;
using C::i;
using B::j;
using C::j;
};
Currently, it appears that the using-declarations of i are ill-formed, on the basis of 7.3.3 [namespace.udecl] paragraph 10:
Since a using-declaration is a declaration, the restrictions on declarations of the same name in the same declarative region (3.3 [basic.scope]) also apply to using-declarations.
Because the using-declarations of i refer to different objects, declaring them in the same scope is not permitted under 3.3 [basic.scope]. It might, however, be preferable to treat this case as many other ambiguities are: allow the declaration but make the program ill-formed if a name reference resolves to the ambiguous declarations.
The status of the using-declarations of j, however, is less clear. They both declare the same entity and thus do not violate the rules of 3.3 [basic.scope]. This might (or might not) violate the restrictions of 9.2 [class.mem] paragraph 1:
Except when used to declare friends (11.4 [class.friend]) or to introduce the name of a member of a base class into a derived class (7.3.3 [namespace.udecl], 11.3 [class.access.dcl]), member-declarations declare members of the class, and each such member-declaration shall declare at least one member name of the class. A member shall not be declared twice in the member-specification, except that a nested class or member class template can be declared and then later defined.
Do the using-declarations of j repeatedly declare the same member? Or is the preceding sentence an indication that a using-declaration is not a declaration of a member?
The following came up recently on comp.lang.c++.moderated (edited for brevity):
namespace N1 {
template<typename T> void f( T* x ) {
// ... other stuff ...
delete x;
}
}
namespace N2 {
using N1::f;
template<> void f<int>( int* ); // A: ill-formed
class Test {
~Test() { }
friend void f<>( Test* x ); // B: ill-formed?
};
}
I strongly suspect, but don't have standardese to prove, that the friend declaration in line B is ill-formed. Can someone show me the text that allows or disallows line B?
Here's my reasoning: Writing "using" to pull the name into namespace N2 merely allows code in N2 to use the name in a call without qualification (per 7.3.3 [namespace.udecl]). But just as declaring a specialization must be done in the namespace where the template really lives (hence line A is ill-formed), I suspect that declaring a specialization as a friend must likewise be done using the original namespace name, not obliquely through a "using". I see nothing in 7.3.3 [namespace.udecl] that would permit this use. Is there?
Andrey Tarasevich: 14.6.4 [temp.friend] paragraph 2 seems to get pretty close: "A friend 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".
Herb Sutter: OK, thanks. Then the question in this is the word "declared" -- in particular, we already know we cannot declare a specialization of a template in any other namespace but the original one.
John Spicer: This seems like a simple question, but it isn't.
First of all, I don't think the standard comments on this usage one way or the other.
A similar example using a namespace qualified name is ill-formed based on 8.3 [dcl.meaning] paragraph 1:
namespace N1 {
void f();
}
namespace N2 {
using N1::f;
class A {
friend void N2::f();
};
}
Core issue 138 deals with this example:
void foo();
namespace A{
using ::foo;
class X{
friend void foo();
};
}
The proposed resolution (not yet approved) for issue 138 is that the friend declares a new foo that conflicts with the using-declaration and results in an error.
Your example is different than this though because the presence of the explicit argument list means that this is not declaring a new f but is instead using a previously declared f.
One reservation I have about allowing the example is the desire to have consistent rules for all of the "declaration like" uses of template functions. Issue 275 (in DR status) addresses the issue of unqualified names in explicit instantiation and explicit specialization declarations. It requires that such declarations refer to templates from the namespace containing the explicit instantiation or explicit specialization. I believe this rule is necessary for those directives but is not really required for friend declarations -- but there is the consistency issue.
Notes from April 2003 meeting:
This is related to issue 138. John Spicer is supposed to update his paper on this topic. This is a new case not covered in that paper. We agreed that the B line should be allowed.
The Standard does not appear to specify what happens for code like the following:
namespace one {
template<typename T> void fun(T);
}
using one::fun;
template<typename T> void fun(T);
7.3.3 [namespace.udecl] paragraph 13 does not appear to apply because it deals only with functions, not function templates:
If a function declaration in namespace scope or block scope has the same name and the same parameter types as a function introduced by a using-declaration, and the declarations do not declare the same function, the program is ill-formed.
John Spicer: For function templates I believe the rule should be that if they have the same function type (parameter types and return type) and have identical template parameter lists, the program is ill-formed.
7.3.3 [namespace.udecl] paragraph 20 says,
If a using-declaration uses the keyword typename and specifies a dependent name (14.7.2 [temp.dep]), the name introduced by the using-declaration is treated as a typedef-name (7.1.3 [dcl.typedef]).
This wording does not address use of typename in a using-declaration with a non-dependent name; the primary specification of the typename keyword in 14.7 [temp.res] does not appear to describe this case, either.
The status of an example like the following is unclear in the current Standard:
struct B {
void f();
};
template<typename T> struct S: T {
using B::f;
};
7.3.3 [namespace.udecl] does not deal explicitly with dependent base classes, but does say in paragraph 3,
In a using-declaration used as a member-declaration, the nested-name-specifier shall name a base class of the class being defined. If such a using-declaration names a constructor, the nested-name-specier shall name a direct base class of the class being defined; otherwise it introduces the set of declarations found by member name lookup (10.2 [class.member.lookup], 3.4.3.1 [class.qual]).
In the definition of S, B::f is not a dependent name but resolves to an apparently unrelated class. However, because S could be instantiated as S<B>, presumably 14.7 [temp.res] paragraph 8 would apply:
No diagnostic shall be issued for a template definition for which a valid specialization can be generated.
Note also the resolution of issue 515, which permitted a similar use of a dependent base class named with a non-dependent name.
It is not clear whether some of the wording in 7.5 [dcl.link] that applies only to function types and names ought also to apply to object names. In particular, paragraph 3 says,
Every implementation shall provide for linkage to functions written in the C programming language, "C", and linkage to C++ functions, "C++".
Nothing is said about variable names, apparently meaning that implementations need not provide C (or even C++!) linkage for variable names. Also, paragraph 5 says,
Except for functions with C++ linkage, a function declaration without a linkage specification shall not precede the first linkage specification for that function. A function can be declared without a linkage specification after an explicit linkage specification has been seen; the linkage explicitly specified in the earlier declaration is not affected by such a function declaration.
There doesn't seem to be a good reason for these provisions not to apply to variable names, as well.
The current syntax requires that multiple attributes that appertain to the same entity be grouped into a single attribute-specifier. The migration from existing vendor-specific attributes would be easier if the syntax allowed multiple attribute-specifiers at each location where an attribute-specifier currently appears.
In function, pointer, and pointer-to-member declarators, the attribute-specifier appertains to the type being declared, but the syntax has the attribute-specifieropt appearing before the full type is seen — i.e., before the cv-qualifier-seqopt and, for the function case, before the ref-qualifieropt. GNU attributes appear after these elements (and, for the function case, after the exception-specificationopt as well). It would be better, both logically and for consistency with existing practice, to move the attribute-specifieropt accordingly.
Split off from issue 453.
It is in general not possible to determine at compile time whether a reference is used before it is initialized. Nevertheless, there is some sentiment to require a diagnostic in the obvious cases that can be detected at compile time, such as the name of a reference appearing in its own initializer. The resolution of issue 453 originally made such uses ill-formed, but the CWG decided that this question should be a separate issue.
Rationale (October, 2005):
The CWG felt that this error was not likely to arise very often in practice. Implementations can warn about such constructs, and the resolution for issue 453 makes executing such code undefined behavior; that seemed to address the situation adequately.
Note (February, 2006):
Recent discussions have suggested that undefined behavior be reduced. One possibility (broadening the scope of this issue to include object declarations as well as references) was to require a diagnostic if the initializer uses the value, but not just the address, of the object or reference being declared:
int i = i; // Ill-formed, diagnostic required
void* p = &p; // Okay
8.3.5 [dcl.fct]/2 restricts the use of void as parameter type, but does not mention CV qualified versions. Since void f(volatile void) isn't a callable function anyway, 8.3.5 [dcl.fct] should also ban cv-qualified versions. (BTW, this follows C)
Suggested resolution:
A possible resolution would be to add (cv-qualified) before void in
The parameter list (void) is equivalent to the empty parameter list. Except for this special case, (cv-qualified) void shall not be a parameter type (though types derived from void, such as void*, can).
The current wording of 8.3.5 [dcl.fct] paragraph 6 encompasses more than it should:
If the type of a parameter includes a type of the form “pointer to array of unknown bound of T” or “reference to array of unknown bound of T,” the program is ill-formed. [Footnote: This excludes parameters of type “ptr-arr-seq T2” where T2 is “pointer to array of unknown bound of T” and where ptr-arr-seq means any sequence of “pointer to” and “array of” derived declarator types. This exclusion applies to the parameters of the function, and if a parameter is a pointer to function or pointer to member function then to its parameters also, etc. —end footnote]
The normative wording (contrary to the intention expressed in the footnote) excludes declarations like
template<class T> struct S {};
void f(S<int (*)[]>);
and
struct S {};
void f(int(*S::*)[]);
but not
struct S {};
void f(int(S::*)[]);
8.3.5 [dcl.fct] paragraph 2 says,
The parameter list (void) is equivalent to the empty parameter list.
This special case is intended for C compatibility, but C99 describes it differently (6.7.5.3 paragraph 10):
The special case of an unnamed parameter of type void as the only item in the list specifies that the function has no parameters.
The C99 formulation allows typedefs for void, while C++ (and C90) accept only the keyword itself in this role. Should the C99 approach be adopted?
Notes from the October, 2006 meeting:
The CWG did not take a formal position on this issue; however, there was some concern expressed over the treatment of function templates and member functions of class templates if the C++ rule were changed: for a template parameter T, would a function taking a single parameter of type T become a no-parameter function if it were instantiated with T = void?
It seems strange that it is possible to call a function with an explict argument of {} but that it is not possible to specify that same argument as a default in a function declaration.
8.3.5 [dcl.fct] paragraph 5 specifies that cv-qualifiers are deleted from parameter types. However, it's not clear what this should mean for function templates. For example,
template<class T> struct A {
typedef A arr[3];
};
template<class T> void f(const typename A<T>::arr) { }
template void f<int>(const A<int>::arr); // #1
template <class T> struct B {
void g(T);
};
template <class T> void B<T>::g(const T) { } // #2
If cv-qualifiers are dropped, then the explicit instantiation in #1 will fail to match; if cv-qualifiers are retained, then the definition in #2 does not match the declaration.
The standard is not precise enough about when the default arguments of member functions are parsed. This leads to confusion over whether certain constructs are legal or not, and the validity of certain compiler implementation algorithms.
8.3.6 [dcl.fct.default] paragraph 5 says "names in the expression are bound, and the semantic constraints are checked, at the point where the default argument expression appears"
However, further on at paragraph 9 in the same section there is an example, where the salient parts are
int b;
class X {
int mem2 (int i = b); // OK use X::b
static int b;
};
which appears to contradict the former constraint. At the point the default
argument expression appears in the definition of X, X::b has not been
declared, so one would expect ::b to be bound. This of course appears to
violate 3.3.7 [basic.scope.class] paragraph 1(2) "A name N used in
a class S shall
refer to the same declaration in its context and when reevaluated in the
complete scope of S. No diagnostic is required."
Furthermore 3.3.7 [basic.scope.class] paragraph 1(1) gives the scope of names declared in class to "consist not only of the declarative region following the name's declarator, but also of .. default arguments ...". Thus implying that X::b is in scope in the default argument of X::mem2 previously.
That previous paragraph hints at an implementation technique of saving the token stream of a default argument expression and parsing it at the end of the class definition (much like the bodies of functions defined in the class). This is a technique employed by GCC and, from its behaviour, in the EDG front end. The standard leaves two things unspecified. Firstly, is a default argument expression permitted to call a static member function declared later in the class in such a way as to require evaluation of that function's default arguments? I.e. is the following well formed?
class A {
static int Foo (int i = Baz ());
static int Baz (int i = Bar ());
static int Bar (int i = 5);
};
If that is well formed, at what point does the non-sensicalness of
class B {
static int Foo (int i = Baz ());
static int Baz (int i = Foo());
};
become detected? Is it when B is complete? Is it when B::Foo or B::Baz is
called in such a way to require default argument expansion? Or is no
diagnostic required?
The other problem is with collecting the tokens that form the default argument expression. Default arguments which contain template-ids with more than one parameter present a difficulty in determining when the default argument finishes. Consider,
template <int A, typename B> struct T { static int i;};
class C {
int Foo (int i = T<1, int>::i);
};
The default argument contains a non-parenthesized comma. Is it required
that this comma is seen as part of the default argument expression and not
the beginning of another of argument declaration? To accept this as
part of the default argument would require name lookup of T (to determine
that the '<' was part of a template argument list and not a less-than
operator) before C is complete. Furthermore, the more pathological
class D {
int Foo (int i = T<1, int>::i);
template <int A, typename B> struct T {static int i;};
};
would be very hard to accept. Even though T is declared after Foo, T is
in scope within Foo's default argument expression.
Suggested resolution:
Append the following text to 8.3.6 [dcl.fct.default] paragraph 8.
The default argument expression of a member function declared in the class definition consists of the sequence of tokens up until the next non-parenthesized, non-bracketed comma or close parenthesis. Furthermore such default argument expressions shall not require evaluation of a default argument of a function declared later in the class.
This would make the above A, B, C and D ill formed and is in line with the existing compiler practice that I am aware of.
Notes from the October, 2005 meeting:
The CWG agreed that the first example (A) is currently well-formed and that it is not unreasonable to expect implementations to handle it by processing default arguments recursively.
Additional notes, May, 2009:
Presumably the following is ill-formed:
int f(int = f());
However, it is not clear what in the Standard makes it so. Perhaps there needs to be a statement to the effect that a default argument only becomes usable after the complete declarator of which it is a part.
Is this program well-formed?
struct S {
static int f2(int = f1()); // OK?
static int f1(int = 2);
};
int main()
{
return S::f2();
}
A class member function can in general refer to class members that are declared lexically later. But what about referring to default arguments of member functions that haven't yet been declared?
It seems to me that if f2 can refer to f1, it can also refer to the default argument of f1, but at least one compiler disagrees.
Paragraph 9 of 8.5 [dcl.init] says:
If no initializer is specified for an object, and the object is of (possibly cv-qualified) non-POD class type (or array thereof), the object shall be default-initialized; if the object is of const-qualified type, the underlying class type shall have a user-declared default constructor. Otherwise, if no initializer is specified for an object, the object and its subobjects, if any, have an indeterminate initial value; if the object or any of its subobjects are of const-qualified type, the program is ill-formed.
What if a const POD object has no non-static data members? This wording requires an empty initializer for such cases:
struct Z {
// no data members
operator int() const { return 0; }
};
void f() {
const Z z1; // ill-formed: no initializer
const Z z2 = { }; // well-formed
}
Similar comments apply to a non-POD const object, all of whose non-static data members and base class subobjects have default constructors. Why should the class of such an object be required to have a user-declared default constructor?
(See also issue 78.)
According to 8.5 [dcl.init] paragraph 5,
To zero-initialize an object of type T means:
...
if T is a reference type, no initialization is performed.
However, a reference is not an object, so this makes no sense.
In this example:
struct A {};
struct B: A {
B(int);
B(B&);
B(A);
};
void foo(B);
void bar() {
foo(0);
}
we are copy-initializing a B from 0. So by 13.3.1.4 [over.match.copy] we consider all the converting constructors of B, and choose B(int) to create a B. Then, by 8.5 [dcl.init] paragraph 15, we direct-initialize the parameter from that temporary B. By 13.3.1.3 [over.match.ctor] we consider all constructors. The copy constructor cannot be called with a temporary, but B(A) is callable.
As far as I can tell, the Standard says that this example is well-formed, and calls B(A). EDG and G++ have rejected this example with a message about the copy constructor not being callable, but I have been unsuccessful in finding anything in the Standard that says that we only consider the copy constructor in the second step of copy-initialization. I wouldn't mind such a rule, but it doesn't seem to be there. And implementing issue 391 causes G++ to start accepting the example.
This question came up before in a GCC bug report; in the discussion of that bug Nathan Sidwell said that some EDG folks explained to him why the testcase is ill-formed, but unfortunately didn't provide that explanation in the bug report.
I think the resolution of issue 391 makes this example well-formed; it was previously ill-formed because in order to bind the temporary B(0) to the argument of A(const A&) we needed to make another temporary B, and that's what made the example ill-formed. If we want this example to stay ill-formed, we need to change something else.
Steve Adamczyk:
I tracked down my response to Nathan at the time, and it related to my paper N1232 (on the auto_ptr problem). The change that came out of that paper is in 13.3.3.1 [over.best.ics] paragraph 4:
However, when considering the argument of a user-defined conversion function that is a candidate by 13.3.1.3 [over.match.ctor] when invoked for the copying of the temporary in the second step of a class copy-initialization, or by 13.3.1.4 [over.match.copy], 13.3.1.5 [over.match.conv], or 13.3.1.6 [over.match.ref] in all cases, only standard conversion sequences and ellipsis conversion sequences are allowed.
This is intended to prevent use of more than one implicit user- defined conversion in an initialization.
I told Nathan B(A) can't be called because its argument would require yet another user-defined conversion, but I was wrong. I saw the conversion from B to A and immediately thought “user-defined,” but in fact because B is a derived class of A the conversion according to 13.3.3.1 [over.best.ics] paragraph 6 is a derived-to-base Conversion (even though it will be implemented by calling a copy constructor).
So I agree with you: with the analysis above and the change for issue 391 this example is well-formed. We should discuss whether we want to make a change to keep it ill-formed.
8.5.1 [dcl.init.aggr] paragraph 4 says,
An initializer-list is ill-formed if the number of initializer-clauses exceeds the number of members or elements to initialize.
However, in a new-expression, the number of elements to be initialized is potentially unknown at compile time. How should an overly-long initializer-list in a new-expression be treated?
A POD-struct is not permitted to have a user-declared copy assignment operator (9 [class] paragraph 4). However, a template assignment operator is not considered a copy assignment operator, even though its specializations can be selected by overload resolution for performing copy operations (12.8 [class.copy] paragraph 9 and especially footnote 114). Consequently, X in the following code is a POD, notwithstanding the fact that copy assignment (for a non-const operand) is a member function call rather than a bitwise copy:
struct X {
template<typename T> const X& operator=(T&);
};
void f() {
X x1, x2;
x1 = x2; // calls X::operator=<X>(X&)
}
Is this intentional?
There doesn't seem to be a prohibition in 9.5 [class.union] against a declaration like
union { int : 0; } x;
Should that be valid? If so, 8.5 [dcl.init]
paragraph 5 third bullet, which deals with
default-initialization of unions, should say that no initialization is
done if there are no data members.
What about:
union { } x;
static union { };
If the first example is well-formed, should either or both of these cases
be well-formed as well?
(See also the resolution for issue 151.)
Notes from 10/00 meeting: The resolution to issue 178, which was accepted as a DR, addresses the first point above (default initialization). The other questions have not yet been decided, however.
Is the signedness of x in the following example implementation-defined?
template <typename T> struct A {
T x : 7;
};
template struct A<long>;
A similar example could be created with a typedef.
Lawrence Crowl: According to 9.6 [class.bit] paragraph 3,
It is implementation-defined whether a plain (neither explicitly signed nor unsigned) char, short, int or long bit-field is signed or unsigned.
This clause is conspicuously silent on typedefs and template parameters.
Clark Nelson: At least in C, the intention is that the presence or absence of this redundant keyword is supposed to be remembered through typedef declarations. I don't remember discussing it in C++, but I would certainly hope that we don't want to do something different. And presumably, we would want template type parameters to work the same way.
So going back to the original example, in an instantiation of A<long>, the signedness of the bit-field is implementation-defined, but in an instantiation of A<signed long>, the bit-field is definitely signed.
Peter Dimov: How can this work? Aren't A<long> and A<signed long> the same type?
(See also issue 739.)9.6 [class.bit] paragraph 3 says,
It is implementation-defined whether a plain (neither explicitly signed nor unsigned) char, short, int or long bit-field is signed or unsigned.
The implications of this permission for an implementation that chooses to treat plain bit-fields as unsigned are not clear. Does this mean that the type of such a bit-field is adjusted to the unsigned variant or simply that sign-extension is not performed when the value is fetched? C99 is explicit in specifying the former (6.7.2 paragraph 5: “for bit-fields, it is implementation-defined whether the specifier int designates the same type as signed int or the same type as unsigned int”), while C90 takes the latter approach (6.5.2.1: “Whether the high-order bit position of a (possibly qualified) 'plain' int bit-field is treated as a sign bit is implementation-defined”).
(See also issue 675 and issue 741.)Additional note, May, 2009:
As an example of the implications of this question, consider the following declaration:
struct S {
int i: 2;
signed int si: 2;
unsigned int ui: 2;
} s;
Is it implementation-defined which expression, cond?s.i:s.si or cond?s.i:s.ui, is an lvalue (the lvalueness of the result depends on the second and third operands having the same type, per 5.16 [expr.cond] paragraph 4)?
The term "ambiguous base class" doesn't seem to be actually defined anywhere. 10.2 [class.member.lookup] paragraph 7 seems like the place to do it.
According to 10.4 [class.abstract] paragraph 6,
Member functions can be called from a constructor (or destructor) of an abstract class; the effect of making a virtual call (10.3 [class.virtual]) to a pure virtual function directly or indirectly for the object being created (or destroyed) from such a constructor (or destructor) is undefined.
This prohibition is unnecessarily restrictive. It should not apply to cases in which the pure virtual function has been defined.
Currently the "pure" specifier for a virtual member function has two meanings that need not be related:
The prohibition of virtual calls to pure virtual functions arises from the first meaning and unnecessarily penalizes those who only need the second.
For example, consider a scenario such as the following. A class B is defined containing a (non-pure) virtual function f that provides some initialization and is thus called from the base class constructor. As time passes, a number of classes are derived from B and it is noticed that each needs to override f, so it is decided to make B::f pure to enforce this convention while still leaving the original definition of B::f to perform its needed initialization. However, the act of making B::f pure means that every reference to f that might occur during the execution of one of B's constructors must be tracked down and edited to be a qualified reference to B::f. This process is tedious and error-prone: needed edits might be overlooked, and calls that actually should be virtual when the containing function is called other than during construction/destruction might be incorrectly changed.
Suggested resolution: Allow virtual calls to pure virtual functions if the function has been defined.
Referring to a private member of a class, 11 [class.access] paragraph 1 says,
its name can be used only by members and friends of the class in which it is declared.
That wording does not appear to reflect the intent of access control, however. Consider the following:
struct S {
void f(int);
private:
void f(double);
};
void g(S* sp) {
sp->f(2); // Ill-formed?
}
The statement from 11 [class.access] paragraph 1 says that the name f can be used only by members and friends of S. Function g is neither, and it clearly contains a use of the name f. That appears to make it ill-formed, in spite of the fact that overload resolution will select the public member.
A related question is whether the use of the term “name” in the description of the effect of access control means that it does not apply to constructors and destructors, which do not have names.
Mike Miller: The phrase “its name can be used” should be understood as “it can be referred to by name.” Paragraph 4, among other places, makes it clear that access control is applied after overload resolution. The “name” phrasing is there to indicate that access control does not apply where the name is not used (in a call via a pointer, for example).
I have heard a claim that the following code is valid, but I don't see why.
struct A {
int foo ();
};
struct B: A {
private:
using A::foo;
};
int main ()
{
return B ().foo ();
}
It seems to me that the using declaration in B should hide the public foo in A. Then the call to B::foo should fail because B::foo is not accessible in main.
Am I missing something?
Steve Adamczyk: This is similar to the last example in 11.2 [class.access.base]. In prose, the rule is that if you have access to cast to a base class and you have access to the member in the base class, you are given access in the derived class. In this case, A is a public base class of B and foo is public in A, so you can access foo through a B object. The actual permission for this is in the fourth bullet in 11.2 [class.access.base] paragraph 4.
The wording changes for issue 9 make this clearer, but I believe even without them this example could be discerned to be valid.
See my paper J16/96-0034, WG21/N0852 on this topic.
Steve Clamage: But a using-declaration is a declaration (7.3.3 [namespace.udecl]). Compare with
struct B : A {
private:
int foo();
};
In this case, the call would certainly be invalid, even though your argument about casting B to an A would make it OK. Your argument basically says that an access adjustment to make something less accessible has no effect. That also doesn't sound right.
Steve Adamczyk: I agree that is strange. I do think that's what 11.2 [class.access.base] says, but perhaps that's not what we want it to say.
Consider the following example:
struct B { void f(){} };
class N : protected B { };
struct P: N { friend int main(); };
int main() {
N n;
B& b = n; // R
b.f();
}
This code is rendered well-formed by bullet 3 of 11.2 [class.access.base] paragraph 4, which says that a base class B of N is accessible at R if
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
This provision circumvents the additional restrictions on access to protected members found in 11.5 [class.protected] — main() could not call B::f() directly because the reference is not via an object of the class through which access is obtained. What is the purpose of this rule?
With the change from a scope-based to an entity-based definition of friendship (see issues 372 and 580), it could well make sense to grant friendship to enumerations and variables, for example:
enum E: int;
class C {
static const int i = 5; // Private
friend E;
friend int x;
};
enum E { e = C::i; }; // OK: E is a friend
int x = C::i; // OK: x is a friend
According to the current wording of 11.4 [class.friend] paragraph 3, the friend declaration of E is well-formed but ignored, while the friend declaration of x is ill-formed.
Although it is not possible to specify a constructor's template arguments in a constructor invocation (because the constructor has no name but is invoked by use of the constructor's class's name), it is possible to “name” the constructor in declarative contexts: per 3.4.3.1 [class.qual] paragraph 2,
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... Such a constructor name shall be used only in the declarator-id of a declaration that names a constructor.
Should it therefore be possible to specify template-arguments for a templated constructor in an explicit instantiation or specialization? For example,
template <int dim> struct T {};
struct X {
template <int dim> X (T<dim> &) {};
};
template X::X<> (T<2> &);
If so, that should be clarified in the text. In particular, 12.1 [class.ctor] paragraph 1 says,
Constructors do not have names. A special declarator syntax using an optional sequence of function-specifiers (7.1.2 [dcl.fct.spec]) followed by the constructor’s class name followed by a parameter list is used to declare or define the constructor.
This certainly sounds as if the parameter list must immediately follow the class name, with no allowance for a template argument list.
It would be worthwhile in any event to revise this wording to utilize the “considered to name” approach of 3.4.3.1 [class.qual]; as it stands, this wording sounds as if the following would be acceptable:
struct S {
S();
};
S() { } // qualified-id not required?
Notes from the October, 2006 meeting:
It was observed that explicitly specifying the template arguments in a constructor declaration is never actually necessary because the arguments are, by definition, all deducible and can thus be omitted.
A posting in comp.lang.c++.moderated prompted me to try the following code:
struct S {
template<typename T, int N> (&operator T())[N];
};
The goal is to have a (deducible) conversion operator template to a reference-to-array type.
This is accepted by several front ends (g++, EDG), but I now believe that 12.3.2 [class.conv.fct] paragraph 1 actually prohibits this. The issue here is that we do in fact specify (part of) a return type.
OTOH, I think it is legitimate to expect that this is expressible in the language (preferably not using the syntax above ;-). Maybe we should extend the syntax to allow the following alternative?
struct S {
template<typename T, int N> operator (T(&)[N])();
};
Eric Niebler: If the syntax is extended to support this, similar constructs should also be considered. For instance, I can't for the life of me figure out how to write a conversion member function template to return a member function pointer. It could be useful if you were defining a null_t type. This is probably due to my own ignorance, but getting the syntax right is tricky.
Eg.
struct null_t {
// null object pointer. works.
template<typename T> operator T*() const { return 0; }
// null member pointer. works.
template<typename T,typename U> operator T U::*() const { return 0; }
// null member fn ptr. doesn't work (with Comeau online). my error?
template<typename T,typename U> operator T (U::*)()() const { return 0; }
};
Martin Sebor: Intriguing question. I have no idea how to do it in a single declaration but splitting it up into two steps seems to work:
struct null_t {
template <class T, class U>
struct ptr_mem_fun_t {
typedef T (U::*type)();
};
template <class T, class U>
operator typename ptr_mem_fun_t<T, U>::type () const {
return 0;
}
};
Note: In the April 2003 meeting, the core working group noticed that the above doesn't actually work.
Note that destructors suffer from similar problems as those of constructors dealt with in issue 194 and in 263 (constructors as friends). Also, the wording in 12.4 [class.dtor], paragraph 1 does not permit a destructor to be defined outside of the memberlist.
Change 12.4 [class.dtor], paragraph 1 from
...A special declarator syntax using an optional function-specifier (7.1.2 [dcl.fct.spec]) followed by ~ followed by the destructor's class name followed by an empty parameter list is used to declare the destructor in a class definition. In such a declaration, the ~ followed by the destructor's class name can be enclosed in optional parentheses; such parentheses are ignored....
to
...A special declarator syntax using an optional sequence of function-specifiers (7.1.2 [dcl.fct.spec]), an optional friend keyword, an optional sequence of function-specifiers (7.1.2 [dcl.fct.spec]) followed by an optional :: scope-resolution-operator followed by an optional nested-name-specifier followed by ~ followed by the destructor's class name followed by an empty parameter list is used to declare the destructor. The optional nested-name-specifier shall not be specified in the declaration of a destructor within the member-list of the class of which the destructor is a member. In such a declaration, the optional :: scope-resolution-operator followed by an optional nested-name-specifier followed by ~ followed by the destructor's class name can be enclosed in optional parentheses; such parentheses are ignored....
Paragraph 4 of 12.5 [class.free] speaks of looking up a deallocation function. While it is an error if a placement deallocation function alone is found by this lookup, there seems to be an assumption that a placement deallocation function and a usual deallocation function can both be declared in a given class scope without creating an ambiguity. The normal mechanism by which ambiguity is avoided when functions of the same name are declared in the same scope is overload resolution; however, there is no mention of overload resolution in the description of the lookup. In fact, there appears to be nothing in the current wording that handles this case. That is, the following example appears to be ill-formed, according to the current wording:
struct S {
void operator delete(void*);
void operator delete(void*, int);
};
void f(S* p) {
delete p; // ill-formed: ambiguous operator delete
}
Suggested resolution (Mike Miller, March 2002):
I think you might get the right effect by replacing the last sentence of 12.5 [class.free] paragraph 4 with something like:
After removing all placement deallocation functions, the result of the lookup shall contain an unambiguous and accessible deallocation function.
In an example like,
struct Y {};
template <typename T>
struct X : public virtual Y { };
template <typename T>
class A : public X<T> {
template <typename S>
A (S)
: S ()
{
}
};
template A<int>::A (Y);
Should S be found? (S is a dependent name, so if it resolves to a base class type in the instantiated template, it should satisfy the requirements.) All the compilers I tried allowed this example, but 12.6.2 [class.base.init] paragraph 2 says,
Names in a mem-initializer-id are looked up in the scope of the constructor’s class and, if not found in that scope, are looked up in the scope containing the constructor’s definition.
The name S is not declared in those scopes.
Mike Miller: Here's another example that is accepted by most/all compilers but not by the current wording:
namespace N {
struct B { B(int); };
typedef B typedef_B;
struct D: B {
D();
};
}
N::D::D(): typedef_B(0) { }
Except for the fact that the constructor function parameter names are ignored (see paragraph 7), what the compilers seem to be doing is essentially ordinary unqualified name lookup.
Notes from the October, 2009 meeting:
The eventual resolution of this issue should take into account the template parameter scope introduced by the resolution of issue 481.
References to non-static data members inside the body of a non-static member function (which includes the mem-initializers of a constructor definition) are implicitly transformed to member access expressions using (*this) (9.3.1 [class.mfct.non-static] paragraph 3). Although 5.1.1 [expr.prim.general] paragraph 3 permits use of this in a brace-or-equal-initializer for a non-static data member, 12.6.2 [class.base.init] does not give details about the value of this in that context, and there is no parallel to the transformation of member references into class member access expressions. This leaves use of non-static data members in this context underspecified.
[Picked up by evolution group at October 2002 meeting.]
(See also paper J16/99-0005 = WG21 N1182.)At the London meeting, 12.8 [class.copy] paragraph 15 was changed to limit the optimization described to only the following cases:
Can we find an appropriate description for the desired cases?
Rationale (04/99): The absence of this optimization does not constitute a defect in the Standard, although the proposed resolution in the paper should be considered when the Standard is revised.
Note (March, 2008):
The Evolution Working Group has accepted the intent of this issue and referred it to CWG for action (not for C++0x). See paper J16/07-0033 = WG21 N2173.
Notes from the June, 2008 meeting:
The CWG decided to take no action on this issue until an interested party produces a paper with analysis and a proposal.
A constructor of the form T::T(T&&) is a candidate function for copy construction; however, the declaration of such a constructor does not inhibit the implicit declaration and definition of a copy constructor. This can lead to surprising results. We should consider suppressing the implicit copy constructor if a move constructor is declared.
Paper N2987 suggests that an implicitly-declared copy or move constructor should be explicit if the corresponding constructor of any of its subobjects is explicit. During the discussion at the October, 2009 meeting, the CWG deemed this a separable question from the major emphasis of that paper, and this issue was opened as a placeholder for that discussion.
Consider the following example:
class B1 {};
typedef void (B1::*PB1) (); // memptr to B1
class B2 {};
typedef void (B2::*PB2) (); // memptr to B2
class D1 : public B1, public B2 {};
typedef void (D1::*PD) (); // memptr to D1
struct S {
operator PB1(); // can be converted to PD
} s;
struct T {
operator PB2(); // can be converted to PD
} t;
void foo() {
s == t; // Is this an error?
}
According to 13.6 [over.built] paragraph 16, there is an operator== for PD (“For every pointer to member type...”), so why wouldn't it be used for this comparison?
Mike Miller: The problem, as I understand it, is that 13.3.1.2 [over.match.oper] paragraph 3, bullet 3, sub-bullet 3 is broader than it was intended to be. It says that candidate built-in operators must “accept operand types to which the given operand or operands can be converted according to 13.3.3.1 [over.best.ics].” 13.3.3.1.2 [over.ics.user] describes user-defined conversions as having a second standard conversion sequence, and there is nothing to restrict that second standard conversion sequence.
My initial thought on addressing this would be to say that user-defined conversion sequences whose second standard conversion sequence contains a pointer conversion or a pointer-to-member conversion are not considered when selecting built-in candidate operator functions. They would still be applicable after the hand-off to Clause 5 (e.g., in bringing the operands to their common type, 5.10 [expr.eq], or composite pointer type, 5.9 [expr.rel]), just not in constructing the list of built-in candidate operator functions.
I started to suggest restricting the second standard conversion sequence to conversions having Promotion or Exact Match rank, but that would exclude the Boolean conversions, which are needed for !, &&, and ||. (It would have also restricted the floating-integral conversions, though, which might be a good idea. They can't be used implicitly, I think, because there would be an ambiguity among all the promoted integral types; however, none of the compilers I tested even tried those conversions because the errors I got were not ambiguities but things like “floating point operands not allowed for %”.)
Bill Gibbons: I recall seeing this problem before, though possibly not in committee discussions. As written this rule makes the set of candidate functions dependent on what classes have been defined, including classes not otherwise required to have been defined in order for "==" to be meaningful. For templates this implies that the set is dependent on what templates have been instantiated, e.g.
template<class T> class U : public T { };
U<B1> u; // changes the set of candidate functions to include
// operator==(U<B1>,U<B1>)?
There may be other places where the existence of a class definition, or worse, a template instantiation, changes the semantics of an otherwise valid program (e.g. pointer conversions?) but it seems like something to be avoided.
According to 13.3.3 [over.match.best] paragraph 4, the following program appears to be ill-formed:
void f(int, int=0);
void f(int=0, int);
void g() {
f();
}
Though I do not expect this is the intent of this paragraph in the standard.
13.3.3 [over.match.best] paragraph 4:
If the best viable function resolves to a function for which multiple
declarations were found, and if at least
two of these declarations or the declarations they refer to in the case of
using-declarations specify a
default argument that made the function viable, the program is ill-formed.
[Example:
namespace A {
extern "C" void f(int = 5);
}
namespace B {
extern "C" void f(int = 5);
}
using A::f;
using B::f;
void use() {
f(3); //OK, default argument was not used for viability
f(); //Error: found default argument twice
}
end example]
It's not clear how overloading and partial ordering handle non-deduced pairs of corresponding arguments. For example:
template<typename T>
struct A { typedef char* type; };
template<typename T> char* f1(T, typename A<T>::type); // #1
template<typename T> long* f1(T*, typename A<T>::type*); // #2
long* p1 = f1(p1, 0); // #3
I thought that #3 is ambiguous but different compilers disagree on that. Comeau C/C++ 4.3.3 (EDG 3.0.3) accepted the code, GCC 3.2 and BCC 5.5 selected #1 while VC7.1+ yields ambiguity.
I intuitively thought that the second pair should prevent overloading from triggering partial ordering since both arguments are non-deduced and has different types - (char*, char**). Just like in the following:
template<typename T> char* f2(T, char*); // #3 template<typename T> long* f2(T*, char**); // #4 long* p2 = f2(p2, 0); // #5
In this case all the compilers I checked found #5 to be ambiguous. The standard and DR 214 is not clear about how partial ordering handle such cases.
I think that overloading should not trigger partial ordering (in step 13.3.3 [over.match.best]/1/5) if some candidates have non-deduced pairs with different (specialized) types. In this stage the arguments are already adjusted so no need to mention it (i.e. array to pointer). In case that one of the arguments is non-deuced then partial ordering should only consider the type from the specialization:
template<typename T> struct B { typedef T type; };
template<typename T> char* f3(T, T); // #7
template<typename T> long* f3(T, typename B<T>::type); // #8
char* p3 = f3(p3, p3); // #9
According to my reasoning #9 should yield ambiguity since second pair is (T, long*). The second type (i.e. long*) was taken from the specialization candidate of #8. EDG and GCC accepted the code. VC and BCC found an ambiguity.
John Spicer: There may (or may not) be an issue concerning whether nondeduced contexts are handled properly in the partial ordering rules. In general, I think nondeduced contexts work, but we should walk through some examples to make sure we think they work properly.
Rani's description of the problem suggests that he believes that partial ordering is done on the specialized types. This is not correct. Partial ordering is done on the templates themselves, independent of type information from the specialization.
Notes from October 2004 meeting:
John Spicer will investigate further to see if any action is required.
(See also issue 885.)
The following example is ambiguous according to the Standard:
struct Y {
operator int();
operator double();
};
void f(Y y) {
double d;
d = y; // Ambiguous: Y::operator int() or Y::operator double()?
}
The reason for the ambiguity is that 13.6 [over.built] paragraph 18 says that there are candidate functions double& operator=(double&, int) and double& operator=(double&, double) (among others). In each case, the second argument is converted by a user-defined conversion sequence (13.3.3.1.2 [over.ics.user]) where the initial and final standard conversion sequences are the identity conversion — i.e., the conversion sequences for the second argument are indistinguishable for each of these candidate functions, and they are thus ambiguous.
Intuitively one might expect that, because it converts directly to the target type in the assignment, Y::operator double() would be selected, and in fact, most compilers do select it, but there is currently no rule to distinghish between these user-defined conversions. Should there be?
Additional note (May, 2008):
Here is another example that is somewhat similar:
enum En { ec };
struct S {
operator int();
operator En();
};
void foo () {
S() == 0; // ambiguous?
}
According to 13.6 [over.built] paragraph 12, the candidate functions are
where R is int and L is every promoted arithmetic type. Overload resolution proceeds in two steps: first, for each candidate function, determine which implicit conversion sequence is used to convert from the argument type to the parameter type; then compare the candidate functions on the basis of the relative costs of those conversion sequences.
In the case of operator==(int, int) there is a clear winner: S::operator int() is chosen because the identity conversion int -> int is better than the promotion En -> int. For all the other candidates, the conversion for the first parameter is ambiguous: both S::operator int() and S::operator En() require either an integral conversion (for integral L) or a floating-integral conversion (for floating point L) and are thus indistinguishable.
These additional candidates are not removed from the set of viable functions, however; because of 13.3.3.1 [over.best.ics] paragraph 10, they are assigned the “ambiguous conversion sequence,” which “is treated as a user-defined sequence that is indistinguishable from any other user-defined conversion sequence.” As a result, all the viable functions are indistinguishable and the call is ambiguous. Like the earlier example, one might naively think that the exact match with S::operator int() and bool operator==(int, int) would be selected, but that is not the case.
Consider the following example:
struct NullClass {
template<typename T> operator T () { return 0 ; }
};
int main() {
NullClass n;
n==5; // #1
return 0;
}
The comparison at #1 is, according to the current Standard, ambiguous. According to 13.6 [over.built] paragraph 12, the candidates for operator==(L, R) include functions “for every pair of promoted arithmetic types,” so L could be either int or long, and the conversion operator template will provide an exact match for either.
Some implementations unambiguously choose the int candidate. Perhaps the overload resolution rules could be tweaked to prefer candidates in which L and R are the same type?
According to 14 [temp] paragraph 5,
Except that a function template can be overloaded either by (non-template) functions with the same name or by other function templates with the same name (14.9.3 [temp.over] ), a template name declared in namespace scope or in class scope shall be unique in that scope.3.3.11 [basic.scope.hiding] paragraph 2 agrees that only functions, not function templates, can hide a class name declared in the same scope:
A class name (9.1 [class.name] ) or enumeration name (7.2 [dcl.enum] ) can be hidden by the name of an object, function, or enumerator declared in the same scope.However, 3.3 [basic.scope] paragraph 4 treats functions and template functions together in this regard:
Given a set of declarations in a single declarative region, each of which specifies the same unqualified name,
- they shall all refer to the same entity, or all refer to functions and function templates; or
- exactly one declaration shall declare a class name or enumeration name that is not a typedef name and the other declarations shall all refer to the same object or enumerator, or all refer to functions and function templates; in this case the class name or enumeration name is hidden
John Spicer: You should be able to take an existing program and replace an existing function with a function template without breaking unrelated parts of the program. In addition, all of the compilers I tried allow this usage (EDG, Sun, egcs, Watcom, Microsoft, Borland). I would recommend that function templates be handled exactly like functions for purposes of name hiding.
Martin O'Riordan: I don't see any justification for extending the purview of what is decidedly a hack, just for the sake of consistency. In fact, I think we should go further and in the interest of consistency, we should deprecate the hack, scheduling its eventual removal from the C++ language standard.
The hack is there to allow old C programs and especially the 'stat.h' file to compile with minimum effort (also several other Posix and X headers). People changing such older programs have ample opportunity to "do it right". Indeed, if you are adding templates to an existing program, you should probably be placing your templates in a 'namespace', so the issue disappears anyway. The lookup rules should be able to provide the behaviour you need without further hacking.
Exported templates were a great idea that is generally understood to have failed. In the decade since the standard was adopted, only one implementation has appeared. No current vendors appear interested in creating another. We tentatively suggest this makes the feature ripe for deprecation. Our main concern with deprecation is that it might turn out that exported constrained templates become an important compile-time optimization, as the constraints would be checked once in the exported definition and not in each translation unit consuming the exported declarations.
By analogy with typename, the keyword template used to indicate that a dependent name will be a template name should be optional in contexts where a type is required, e.g., base class lists. We could also consider member and parameter declarations.
This was suggested by issue 314.
The Standard does not normatively define which > and >> tokens are to be taken as closing a template-argument-list; instead, 14.3 [temp.names] paragraph 3 uses the undefined and imprecise term “non-nested:”
When parsing a template-id, the first non-nested > is taken as the end of the template-argument-list rather than a greater-than operator. Similarly, the first non-nested >> is treated as two consecutive but distinct > tokens, the first of which is taken as the end of the template-argument-list and completes the template-id.
The (non-normative) footnote clarifies that
A > that encloses the type-id of a dynamic_cast, static_cast, reinterpret_cast or const_cast, or which encloses the template-arguments of a subsequent template-id, is considered nested for the purpose of this description.
Aside from the questionable wording of this footnote (e.g., in what sense does a single terminating character “enclose” anything, and is a nested template-id “subsequent?”) and the fact that it is non-normative, it does not provide a complete definition of what “nesting” is intended to mean. For example, is the first > in this putative template-id “nested” or not?
X<a ? b > c : d>
None of my compilers accept this, which surprised me a little. Is the base-to-derived member function conversion considered to be a runtime-only thing?
template <class D>
struct B
{
template <class X> void f(X) {}
template <class X, void (D::*)(X) = &B<D>::f<X> >
struct row {};
};
struct D : B<D>
{
void g(int);
row<int,&D::g> r1;
row<char*> r2;
};
John Spicer: This is not among the permitted conversions listed in 14.3.
I'm not sure there is a terribly good reason for that. Some of the template argument rules for external entities were made conservatively because of concerns about issues of mangling template argument names.
David Abrahams: I'd really like to see that restriction loosened. It is a serious inconvenience because there appears to be no way to supply a usable default in this case. Zero would be an OK default if I could use the function pointer's equality to zero as a compile-time switch to choose an empty function implementation:
template <bool x> struct tag {};
template <class D>
struct B
{
template <class X> void f(X) {}
template <class X, void (D::*pmf)(X) = 0 >
struct row {
void h() { h(tag<(pmf == 0)>(), pmf); }
void h(tag<1>, ...) {}
void h(tag<0>, void (D::*q)(X)) { /*something*/}
};
};
struct D : B<D>
{
void g(int);
row<int,&D::g> r1;
row<char*> r2;
};
But there appears to be no way to get that effect either. The result is that you end up doing something like:
template <class X, void (D::*pmf)(X) = 0 >
struct row {
void h() { if (pmf) /*something*/ }
};
which invariably makes compilers warn that you're switching on a constant expression.
[Picked up by evolution group at October 2002 meeting.]
How are default template arguments handled with respect to template template parameters? Two separate questions have been raised:
template <class T, class U = int>
class ARG { };
template <class X, template <class Y> class PARM>
void f(PARM<X>) { } // specialization permitted?
void g() {
ARG<int> x; // actually ARG<int, int>
f(x); // does ARG (2 parms, 1 with default)
// match PARM (1 parm)?
Template template parameters are deducible
(14.9.2.5 [temp.deduct.type]
paragraph 9),
but 14.4.3 [temp.arg.template]
does not
specify how matching is done.
Jack Rouse: I implemented template template parameters assuming template signature matching is analogous to function type matching. This seems like the minimum reasonable implementation. The code in the example would not be accepted by this compiler. However, template default arguments are compile time entities so it seems reasonable to relax the matching rules to allow cases like the one in the example. But I would consider this to be an extension to the language.
Herb Sutter: An open issue in the LWG is that the standard doesn't explicitly permit or forbid implementations' adding additional template-parameters to those specified by the standard, and the LWG may be leaning toward explicitly permitting this. [Under this interpretation,] if the standard is ever modified to allow additional template-parameters, then writing "a template that takes a standard library template as a template template parameter" won't be just ugly because you have to mention the defaulted parameters; it would not be (portably) possible at all except possibly by defining entire families of overloaded templates to account for all the possible numbers of parameters vector<> (or anything else) might actually have. That seems unfortunate.
template <template <class T, class U = int> class PARM>
class C {
PARM<int> pi;
};
Jack Rouse: I decided they could not in the compiler I support. This continues the analogy with function type matching. Also, I did not see a strong need to allow default arguments in this context.
A class template used as a template template argument can have default template arguments from its declarations. How are the two sources of default arguments to be reconciled? The default arguments from the template template formal could override. But it could be cofusing if a template-id using the argument template, ARG<int>, behaves differently from a template-id using the template formal name, FORMAL<int>.
Rationale (10/99): Template template parameters are intended to be handled analogously to function function parameters. Thus the number of parameters in a template template argument must match the number of parameters in a template template parameter, regardless of whether any of those paramaters have default arguments or not. Default arguments are allowed for the parameters of a template template parameter, and those default arguments alone will be considered in a specialization of the template template parameter within a template definition; any default arguments for the parameters of a template template argument are ignored.
Note (Mark Mitchell, February, 2006):
Perhaps it is already obvious to all, but it seems worth noting that this extension would change the meaning of conforming programs:
struct Dense { static const unsigned int dim = 1; };
template <template <typename> class View,
typename Block>
void operator+(float, View<Block> const&);
template <typename Block,
unsigned int Dim = Block::dim>
struct Lvalue_proxy { operator float() const; };
void test_1d (void) {
Lvalue_proxy<Dense> p;
float b;
b + p;
}
If Lvalue_proxy is allowed to bind to View, then the template operator+ will be used to perform addition; otherwise, Lvalue_proxy's implicit conversion to float, followed by the built-in addition on floats will be used.
Note (March, 2008):
The Evolution Working Group has accepted the intent of this issue and referred it to CWG for action (not for C++0x). See paper J16/07-0033 = WG21 N2173.
Notes from the June, 2008 meeting:
The CWG decided to take no action on this issue until an interested party produces a paper with analysis and a proposal.
The Standard does not appear to specify clearly the effect of a partial specialization of a member template of a class template. For example:
template<class T> struct B {
template<class U> struct A { // #1
void h() {}
};
template<class U> struct A<U*> { // #2
void f() {}
};
};
template<> template<class U> struct B<int>::A { // #3
void g() {}
};
void q(B<int>::A<char*>& p) {
p.f(); // #4
}
The explicit specialization at #3 replaces the primary member template #1 of B<int>; however, it is not clear whether the partial specialization #2 should be considered to apply to the explicitly-specialized member template of A<int> (thus allowing the call to p.f() at #4) or whether the partial specialization will be used only for specializations of B that are implicitly instantiated (meaning that #4 could call p.g() but not p.f()).
Given an example like
template<typename T, typename U>
struct Outer {
template<typename X, typename Y> struct Inner;
template<typename Y> struct Inner<T, Y> {};
template<typename Y> struct Inner<U, Y> {};
};
Outer<int, int> outer; // #1
Outer<int, int>::Inner<int, float> inner; // #2
Is #1 ill-formed because of the identical partial specializations? If not, presumably #2 is ill-formed because of the resulting ambiguity (14.6.5.1 [temp.class.spec.match] paragraph 1).
I get the following error diagnostic [from the EDG front end]:
line 8: error: function template "example<T>::foo<R,A>(A)" has
already been declared
R foo(const A);
^
when compiling this piece of code:
struct example {
template<class R, class A> // 1-st member template
R foo(A);
template<class R, class A> // 2-nd member template
const R foo(A&);
template<class R, class A> // 3-d member template
R foo(const A);
};
/*template<> template<>
int example<char>::foo(int&);*/
int main()
{
int (example<char>::* pf)(int&) =
&example<char>::foo;
}
The implementation complains that
template<class R, class A> // 1-st member template R foo(A); template<class R, class A> // 3-d member template R foo(const A);cannot be overloaded and I don't see any reason for it since it is function template specializations that are treated like ordinary non-template functions, meaning that the transformation of a parameter-declaration-clause into the corresponding parameter-type-list is applied to specializations (when determining its type) and not to function templates.
What makes me think so is the contents of 14.6.6.1 [temp.over.link] and the following sentence from 14.9.2.1 [temp.deduct.call] "If P is a cv-qualified type, the top level cv-qualifiers of P are ignored for type deduction". If the transformation was to be applied to function templates, then there would be no reason for having that sentence in 14.9.2.1 [temp.deduct.call].
14.9.2.2 [temp.deduct.funcaddr], which my example is based upon, says nothing about ignoring the top level cv-qualifiers of the function parameters of the function template whose address is being taken.
As a result, I expect that template argument deduction will fail for the 2-nd and 3-d member templates and the 1-st one will be used for the instantiation of the specialization.
Issue 1:
14.6.6.2 [temp.func.order] paragraph 2 says:
Given two overloaded function templates, whether one is more specialized than another can be determined by transforming each template in turn and using argument deduction (14.9.2 [temp.deduct] ) to compare it to the other.14.9.2 [temp.deduct] now has 4 subsections describing argument deduction in different situations. I think this paragraph should point to a subsection of 14.9.2 [temp.deduct] .
Rationale:
This is not a defect; it is not necessary to pinpoint cross-references to this level of detail.
Issue 2:
14.6.6.2 [temp.func.order] paragraph 4 says:
Using the transformed function parameter list, perform argument deduction against the other function template. The transformed template is at least as specialized as the other if, and only if, the deduction succeeds and the deduced parameter types are an exact match (so the deduction does not rely on implicit conversions).In "the deduced parameter types are an exact match", the terms exact match do not make it clear what happens when a type T is compared to the reference type T&. Is that an exact match?
Issue 3:
14.6.6.2 [temp.func.order] paragraph 5 says:
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.What happens in this case:
template<class T> void f(T,int);
template<class T> void f(T, T);
void f(1,1);
For the first function template, there is no type deduction for the second
parameter. So the rules in this clause seem to imply that the second function
template will be chosen.
Rationale:
This is not a defect; the standard unambiguously makes the above example ill-formed due to ambiguity.
This was split off from issue 214 at the April 2003 meeting.
Nathan Sidwell: John Spicer's proposed resolution does not make the following well-formed.
template <typename T> int Foo (T const *) {return 1;} //#1
template <unsigned I> int Foo (char const (&)[I]) {return 2;} //#2
int main ()
{
return Foo ("a") != 2;
}
Both #1 and #2 can deduce the "a" argument, #1 deduces T as char and #2 deduces I as 2. However, neither is more specialized because the proposed rules do not have any array to pointer decay.
#1 is only deduceable because of the rules in 14.9.2.1 [temp.deduct.call] paragraph 2 that decay array and function type arguments when the template parameter is not a reference. Given that such behaviour happens in deduction, I believe there should be equivalent behaviour during partial ordering. #2 should be resolved as more specialized as #1. The following alteration to the proposed resolution of DR214 will do that.
Insert before,
the following
For the example above, this change results in deducing 'T const *' against 'char const *' in one direction (which succeeds), and 'char [I]' against 'T const *' in the other (which fails).
John Spicer: I don't consider this a shortcoming of my proposed wording, as I don't think this is part of the current rules. In other words, the resolution of 214 might make it clearer how this case is handled (i.e., clearer that it is not allowed), but I don't believe it represents a change in the language.
I'm not necessarily opposed to such a change, but I think it should be reviewed by the core group as a related change and not a defect in the proposed resolution to 214.
Notes from the October 2003 meeting:
There was some sentiment that it would be desirable to have this case ordered, but we don't think it's worth spending the time to work on it now. If we look at some larger partial ordering changes at some point, we will consider this again.
The standard prohibits a class template from having the same name as one of its template parameters (14.7.1 [temp.local] paragraph 4). This prohibits
template <class X> class X;
for the reason that the template name would hide the parameter, and
such hiding is in general prohibited.
Presumably, we should also prohibit
template <template <class T> class T> struct A;
for the same reason.
Currently, member of nondependent base classes hide references to template parameters in the definition of a derived class template.
Consider the following example:
class B {
typedef void *It; // (1)
// ...
};
class M: B {};
template<typename> X {};
template<typename It> struct S // (2)
: M, X<It> { // (3)
S(It, It); // (4)
// ...
};
As the C++ language currently stands, the name "It" in line (3) refers to the template parameter declared in line (2), but the name "It" in line (4) refers to the typedef in the private base class (declared in line (1)).
This situation is both unintuitive and a hindrance to sound software engineering. (See also the Usenet discussion at http://tinyurl.com/32q8d .) Among other things, it implies that the private section of a base class may change the meaning of the derived class, and (unlike other cases where such things happen) there is no way for the writer of the derived class to defend the code against such intrusion (e.g., by using a qualified name).
Changing this can break code that is valid today. However, such code would have to:
It has been suggested to make situations like these ill-formed. That solution is unattractive however because it still leaves the writer of a derived class template without defense against accidental name conflicts with base members. (Although at least the problem would be guaranteed to be caught at compile time.) Instead, since just about everyone's intuition agrees, I would like to see the rules changed to make class template parameters hide members of the same name in a base class.
See also issue 458.
Notes from the March 2004 meeting:
We have some sympathy for a change, but the current rules fall straightforwardly out of the lookup rules, so they're not “wrong.” Making private members invisible also would solve this problem. We'd be willing to look at a paper proposing that.
Additional discussion (April, 2005):
John Spicer: Base class members are more-or-less treated as members of the class, [so] it is only natural that the base [member] would hide the template parameter.
Daveed Vandevoorde: Are base class members really “more or less” members of the class from a lookup perspective? After all, derived class members can hide base class members of the same name. So there is some pretty definite boundary between those two sets of names. IMO, the template parameters should either sit between those two sets, or they should (for lookup purposes) be treated as members of the class they parameterize (I cannot think of a practical difference between those two formulations).
John Spicer: How is [hiding template parameters] different from the fact that namespace members can be hidden by private parts of a base class? The addition of int C to N::A breaks the code in namespace M in this example:
namespace N {
class A {
private:
int C;
};
}
namespace M {
typedef int C;
class B : public N::A {
void f() {
C c;
}
};
}
Daveed Vandevoorde: C++ has a mechanism in place to handle such situations: qualified names. There is no such mechanism in place for template parameters.
Nathan Myers: What I see as obviously incorrect ... is simply that a name defined right where I can see it, and directly attached to the textual scope of B's class body, is ignored in favor of something found in some other file. I don't care that C1 is defined in A, I have a C1 right here that I have chosen to use. If I want A::C1, I can say so.
I doubt you'll find any regular C++ coder who doesn't find the standard behavior bizarre. If the meaning of any code is changed by fixing this behavior, the overwhelming majority of cases will be mysterious bugs magically fixed.
John Spicer: I have not heard complaints that this is actually a cause of problems in real user code. Where is the evidence that the status quo is actually causing problems?
In this example, the T2 that is found is the one from the base class. I would argue that this is natural because base class members are found as part of the lookup in class B:
struct A {
typedef int T2;
};
template <class T2> struct B : public A {
typedef int T1;
T1 t1;
T2 t2;
};
This rule that base class members hide template parameters was formalized about a dozen years ago because it fell out of the principle that base class members should be found at the same stage of lookup as derived class members, and that to do otherwise would be surprising.
Gabriel Dos Reis: The bottom line is that:
Unless presented with real major programming problems the current rules exhibit, I do not think the simple rule “scopes nest” needs a change that silently mutates program meaning.
Mike Miller: The rationale for the current specification is really very simple:
That's it. Because template parameters are not members, they are hidden by member names (whether inherited or not). I don't find that “bizarre,” or even particularly surprising.
I believe these rules are straightforward and consistent, so I would be opposed to changing them. However, I am not unsympathetic toward Daveed's concern about name hijacking from base classes. How about a rule that would make a program ill-formed if a direct or inherited member hides a template parameter?
Unless this problem is a lot more prevalent than I've heard so far, I would not want to change the lookup rules; making this kind of collision a diagnosable error, however, would prevent hijacking without changing the lookup rules.
Erwin Unruh: I have a different approach that is consistent and changes the interpretation of the questionable code. At present lookup is done in this sequence:
If we change this order to
it is still consistent in that no lookup is placed between the base class and the derived class. However, it introduces another inconsistency: now scopes do not nest the same way as curly braces nest — but base classes are already inconsistent this way.
Nathan Myers: This looks entirely satisfactory. If even this seems like too big a change, it would suffice to say that finding a different name by this search order makes the program ill-formed. Of course, a compiler might issue only a portability warning in that case and use the name found Erwin's way, anyhow.
Gabriel Dos Reis: It is a simple fact, even without templates, that a writer of a derived class cannot protect himself against declaration changes in the base class.
Richard Corden: If a change is to be made, then making it ill-formed is better than just changing the lookup rules.
struct B
{
typedef int T;
virtual void bar (T const & );
};
template <typename T>
struct D : public B
{
virtual void bar (T const & );
};
template class D<float>;
I think changing the semantics of the above code silently would result in very difficult-to-find problems.
Mike Miller: Another case that may need to be considered in deciding on Erwin's suggestion or the “ill-formed” alternative is the treatment of friend declarations described in 3.4.1 [basic.lookup.unqual] paragraph 10:
struct A {
typedef int T;
void f(T);
};
template<typename T> struct B {
friend void A::f(T); // Currently T is A::T
};
Notes from the October, 2005 meeting:
The CWG decided not to consider a change to the existing rules at this time without a paper exploring the issue in more detail.
Consider the following example:
template<class T>
struct A {
template<class U>
friend struct A; // Which A?
};
Presumably the lookup for A in the friend declaration finds the injected-class-name of the template. However, according to 14.7.1 [temp.local] paragraph 1,
The injected-class-name can be used with or without a template-argument-list. When it is used without a template-argument-list, it is equivalent to the injected-class-name followed by the template-parameters of the class template enclosed in <>. When it is used with a template-argument-list, it refers to the specified class template specialization, which could be the current specialization or another specialization.
If that rule applies, then this example is ill-formed (because you can't have a template-argument-list in a class template declaration that is not a partial specialization).
Mike Miller: The injected-class-name has a dual nature, as described in 14.7.1 [temp.local], acting as either a template name or a class name, depending on the context; a template argument list forces the name to be interpreted as a template. It seems reasonable that in this example the injected-class-name has to be understood as referring to the class template; a template header is at least as strong a contextual indicator as a template argument list. However, the current wording doesn't say that.
Is the following example well-formed?
template<class T> struct A {
typedef int M;
struct B {
typedef void M;
struct C;
};
};
template<class T> struct A<T>::B::C : A<T> {
M // A<T>::M or A<T>::B::M?
p[2];
};
14.7.2 [temp.dep] paragraph 3 says the use of M should refer to A<T>::B::M because the base class A<T> is not searched because it's dependent. But in this case A<T> is also the current instantiation (14.7.2.1 [temp.dep.type]) so it seems like it should be searched.
In 14.7.2.1 [temp.dep.type] paragraph 5 we have:
A name is a member of an unknown specialization if the name is a qualified-id in which the nested-name-specifier names a dependent type that is not the current instantiation.
So given:
template<class T> struct A {
struct B {
struct C {
A<T>::B::C f();
};
};
};
it appears that the name A<T>::B::C should be taken as a member of an unknown specialization, because the WP refers to “the” current instantiation, implying that there can be at most one at any given time. At the declaration of f(), the current instantiation is C, so A<T>::B is not the current instantiation.
Would it be better to refer to “a known instantiation” instead of “the current instantiation?”
Mike Miller:
I agree that there is a problem here, but I don't think the “current instantiation” terminology needs to be replaced. By way of background, paragraph 1 makes it clear that A<T>::B “refers to” the current instantiation:
In the definition of a class template, a nested class of a class template, a member of a class template, or a member of a nested class of a class template, a name refers to the current instantiation if it is
the injected-class-name (9 [class]) of the class template or nested class,
in the definition of a primary class template, the name of the class template followed by the template argument list of the primary template (as described below) enclosed in <>,
in the definition of a nested class of a class template, the name of the nested class referenced as a member of the current instantiation...
A<T>::B satisfies bullet 3. Paragraph 4 says,
A name is a member of the current instantiation if it is
An unqualified name that, when looked up, refers to a member of a class template. [Note: this can only occur when looking up a name in a scope enclosed by the definition of a class template. —end note]
A qualified-id in which the nested-name-specifier refers to the current instantiation.
So clearly by paragraphs 1 and 4, A<T>::B::C is a member of the current instantiation. The problem is in the phrasing of paragraph 5, which incorrectly requires that the nested-name-specifier “be” the current instantiation rather than simply “referring to” the current instantiation, which would be the correct complement to paragraph 4. Perhaps paragraph 5 could simply be rephrased as, “...a dependent type and it is not a member of the current instantiation.”
(Paragraph 1 may require a bit more wordsmithing to make it truly recursive across multiple levels of nested classes; as it stands, it's not clear whether the name of a nested class of a nested class of a class template is covered or not.)
Consider the following example:
void f(int*);
void f(...);
template <int N> void g() {
f(N);
}
int main() {
g<0>();
g<1>();
}
The call to f in g is not type-dependent, so the overload resolution must be done at definition time rather than at instantiation time. As a result, both of the calls to g will result in calls to f(...), i.e., N will not be a null pointer constant, even if the value of N is 0.
It would be most consistent to adopt a rule that a value-dependent expression can never be a null pointer constant, even in cases like
template <int N> void g() {
int* p = N;
}
This would always be ill-formed, even when N is 0.
John Spicer: It's clear that this treatment is required for overload resolution, but it seems too expansive given that there are other cases in which the value of a template parameter can affect the validity of the program, and an implementation is forbidden to issue a diagnostic on a template definition unless there are no possible valid specializations.
Notes from the July, 2009 meeting:
There was a strong consensus among the CWG that only the literal 0 should be considered a null pointer constant, not any arbitrary zero-valued constant expression as is currently specified.
The intent is that it is a permissible implementation technique to do template instantiation at the end of a translation unit rather than at an actual point of instantiation. This idea is not reflected in the current rules, however.
14.8.2 [temp.explicit] defines an explicit instantiation as
Syntactically, that allows things like:
template int S<int>::i = 5, S<int>::j = 7;
which isn't what anyone actually expects. As far as I can tell, nothing in the standard explicitly forbids this, as written. Syntactically, this also allows:
template namespace N { void f(); }
although perhaps the surrounding context is enough to suggest that this is invalid.
Suggested resolution:
I think we should say:
[Steve Adamczyk: presumably, this should have template at the beginning.]
and then say that:
There are similar problems in 14.8.3 [temp.expl.spec]:
Here, I think we want:
with similar restrictions as above.
[Steve Adamczyk: This also needs to have template <> at the beginning, possibly repeated.]
14.8.2 [temp.explicit] paragraph 5 has an example that reads, in significant part,
namespace N {
template<class T> class Y {
void mf() { }
};
}
using N::Y;
template class Y<int>; // OK: explicit instantiation in namespace N
In fact, paragraph 2 requires that an explicit instantiation with an unqualified name must appear in the same namespace in which the template was declared, so the example is ill-formed.
The note in paragraph 5 of 14.9.1 [temp.arg.explicit] makes clear that explicit template arguments cannot be supplied in invocations of constructors and conversion functions because they are called without using a name. However, there is nothing in the current wording of the Standard that makes declaring a constructor or conversion operator that is unusable because of nondeduced parameters (i.e., that would need to be specified explicitly) ill-formed. It would be a service to the programmer to diagnose this useless construct as early as possible.
Nicolai Josuttis sent me an example like the following:
template <typename RET, typename T1, typename T2>
const RET& min (const T1& a, const T2& b)
{
return (a < b ? a : b);
}
template const int& min<int>(const int&,const int&); // #1
template const int& min(const int&,const int&); // #2
Among the questions was whether explicit instantiation #2 is valid, where deduction is required to determine the type of RET.
The first thing I realized when researching this is that the standard does not really spell out the rules for deduction in declarative contexts (friend declarations, explicit specializations, and explicit instantiations). For explicit instantiations, 14.8.2 [temp.explicit] paragraph 2 does mention deduction, but it doesn't say which set of deduction rules from 14.9.2 [temp.deduct] should be applied.
Second, Nicolai pointed out that 14.8.2 [temp.explicit] paragraph 6 says
A trailing template-argument can be left unspecified in an explicit instantiation provided it can be deduced from the type of a function parameter (14.9.2 [temp.deduct]).
This prohibits cases like #2, but I believe this was not considered in the wording as there is no reason not to include the return type in the deduction process.
I think there may have been some confusion because the return type is excluded when doing deduction on a function call. But there are contexts where the return type is included in deduction, for example, when taking the address of a function template specialization.
Suggested resolution:
Andrei Iltchenko points out that the standard has no wording that defines how to determine which template is specialized by an explicit specialization of a function template. He suggests "template argument deduction in such cases proceeds in the same way as when taking the address of a function template, which is described in 14.9.2.2 [temp.deduct.funcaddr]."
John Spicer points out that the same problem exists for all similar declarations, i.e., friend declarations and explicit instantiation directives. Finding a corresponding placement operator delete may have a similar problem.
John Spicer: There are two aspects of "determining which template" is referred to by a declaration: determining the function template associated with the named specialization, and determining the values of the template arguments of the specialization.
template <class T> void f(T); #1
template <class T> void f(T*); #2
template <> void f(int*);
In other words, which f is being specialized (#1 or #2)? And then, what are the deduced template arguments?
14.6.6.2 [temp.func.order] does say that partial ordering is done in contexts such as this. Is this sufficient, or do we need to say more about the selection of the function template to be selected?
14.9.2 [temp.deduct] probably needs a new section to cover argument deduction for cases like this.
14.9.2 [temp.deduct] is all about function types, but these rules also apply, e.g., when matching a class template partial specialization. We should add a note stating that we could be doing substitution into the template-id for a class template partial specialization.
Additional note (August 2008):
According to 14.6.5.1 [temp.class.spec.match] paragraph 2, argument deduction is used to determine whether a given partial specialization matches a given argument list. However, there is nothing in 14.6.5.1 [temp.class.spec.match] nor in 14.9.2 [temp.deduct] and its subsections that describes exactly how argument deduction is to be performed in this case. It would seem that more than just a note is required to clarify this processing.
Consider the following program:
template <typename T> int ref (T&) { return 0; }
template <typename T> int ref (const T&) { return 1; }
template <typename T> int ref (const volatile T&) { return 2; }
template <typename T> int ref (volatile T&) { return 4; }
template <typename T> int ptr (T*) { return 0; }
template <typename T> int ptr (const T*) { return 8; }
template <typename T> int ptr (const volatile T*) { return 16; }
template <typename T> int ptr (volatile T*) { return 32; }
void foo() {}
int main()
{
return ref(foo) + ptr(&foo);
}
The Standard appears to specify that the value returned from main is 2. The reason for this result is that references and pointers are handled differently in template argument deduction.
For the reference case, 14.9.2.1 [temp.deduct.call] paragraph 3 says that “If P is a reference type, the type referred to by P is used for type deduction.” Because of issue 295, all four of the types for the ref function parameters are the same, with no cv-qualification; overload resolution does not find a best match among the parameters and thus the most-specialized function is selected.
For the pointer type, argument deduction does not get as far as forming a cv-qualified function type; instead, argument deduction fails in the cv-qualified cases because of the cv-qualification mismatch, and only the cv-unqualified version of ptr survives as a viable function.
I think the choice of ignoring cv-qualifiers in the reference case but not the pointer case is very troublesome. The reason is that when one considers function objects as function parameters, it introduces a semantic difference whether the function parameter is declared a reference or a pointer. In all other contexts, it does not matter: a function name decays to a pointer and the resulting semantics are the same.
The current wording of 15.3 [except.handle] paragraph 16 is:
The object declared in an exception-declaration or, if the exception-declaration does not specify a name, a temporary (12.2 [class.temporary]) is copy-initialized (8.5 [dcl.init]) from the exception object. The object shall not have an abstract class type. The object is destroyed when the handler exits, after the destruction of any automatic objects initialized within the handler.
There are two problems with this. First, it's not clear what it means for the handler's “parameter” to be a temporary. This possibility is briefly mentioned in 12.2 [class.temporary], but the lifetime of such a temporary is not defined there; the discussion of lifetime is restricted to those temporaries that arise during the evaluation of an expression, and this is not such a case.
Second, this wording assumes that there will be an object to be destroyed and thus ignores the possibility that the exception-declaration declares a reference.
It was tentatively agreed at the Santa Cruz meeting that exception specifications should fully participate in the type system. This change would address gaps in the current static checking of exception specifications such as
void (*p)() throw(int);
void (**pp)() throw() = &p; // not currently an error
This is such a major change that it deserves to be a separate issue.
See also issues 25, 87, and 133.
A type used in an exception specification must be complete (15.4 [except.spec] paragraph 2). The resolution of issue 437 stated that a class type appearing in an exception specification inside its own member-specification is considered to be complete. Should this also apply to exception specifications in class templates instantiated because of a reference inside the member-specification of a class? For example,
template<class T> struct X {
void f() throw(T) {}
};
struct S {
X<S> xs;
};
Exception specifications have proven close to worthless in practice, while adding a measurable overhead to programs. The feature should be deprecated. The one exception to the rule is the empty throw specification which could serve a legitimate optimizing role if the requirement to call std::unexpected were relaxed in this case.
Notes from the July, 2009 meeting:
The consensus of the CWG was in favor of deprecating exception specifications. Further discussion, and with a wider constituency, is needed to determine a position on the status of throw().
(See also issue 814.)
Destructors that throw can easily cause programs to terminate, with no possible defense. Example: Given
struct XY { X x; Y y; };
Assume that X::~X() is the only destructor in the entire program that can throw. Assume further that Y construction is the only other operation in the whole program that can throw. Then XY cannot be used safely, in any context whatsoever, period — even simply declaring an XY object can crash the program:
XY xy; // construction attempt might terminate program:
// 1. construct x -- succeeds
// 2. construct y -- fails, throws exception
// 3. clean up by destroying x -- fails, throws exception,
// but an exception is already active, so call
// std::terminate() (oops)
// there is no defense
So it is highly dangerous to have even one destructor that could throw.
Suggested Resolution:
Fix the above problem in one of the following two ways. I prefer the first.
Fergus Henderson: I disagree. Code using XY may well be safe, if X::~X() only throws if std::uncaught_exception() is false.
I think the current exception handling scheme in C++ is certainly flawed, but the flaws are IMHO design flaws, not minor technical defects, and I don't think they can be solved by minor tweaks to the existing design. I think that at this point it is probably better to keep the standard stable, and learn to live with the existing flaws, rather than trying to solve them via TC.
Bjarne Stroustrup: I strongly prefer to have the call to std::terminate() be conforming. I see std::terminate() as a proper way to blow away "the current mess" and get to the next level of error handling. I do not want that escape to be non-conforming — that would imply that programs relying on a error handling based on serious errors being handled by terminating a process (which happens to be a C++ program) in std::terminate() becomes non-conforming. In many systems, there are — and/or should be — error-handling and recovery mechanisms beyond what is offered by a single C++ program.
Andy Koenig: If we were to prohibit writing a destructor that can throw, how would I solve the following problem?
I want to write a class that does buffered output. Among the other properties of that class is that destroying an object of that class writes the last buffer on the output device before freeing memory.
What should my class do if writing that last buffer indicates a hardware output error? My user had the option to flush the last buffer explicitly before destroying the object, but didn't do so, and therefore did not anticipate such a problem. Unfortunately, the problem happened anyway. Should I be required to suppress this error indication anyway? In all cases?
Herb Sutter (June, 2007): IMO, it's fine to suppress it. The user had the option of flushing the buffer and thus being notified of the problem and chose not to use it. If the caller didn't flush, then likely the caller isn't ready for an exception from the destructor, either. You could also put an assert into the destructor that would trigger if flush() had not been called, to force callers to use the interface that would report the error.
In practice, I would rather thrown an exception, even at the risk of crashing the program if we happen to be in the middle of stack unwinding. The reason is that the program would crash only if a hardware error occurred in the middle of cleaning up from some other error that was in the process of being handled. I would rather have such a bizarre coincidence cause a crash, which stands a chance of being diagnosed later, than to be ignored entirely and leave the system in a state where the ignore error could cause other trouble later that is even harder to diagnose.
If I'm not allowed to throw an exception when I detect this problem, what are my options?
Herb Sutter: I understand that some people might feel that "a failed dtor during stack unwinding is preferable in certain cases" (e.g., when recovery can be done beyond the scope of the program), but the problem is "says who?" It is the application program that should be able to decide whether or not such semantics are correct for it, and the problem here is that with the status quo a program cannot defend itself against a std::terminate() — period. The lower-level code makes the decision for everyone. In the original example, the mere existence of an XY object puts at risk every program that uses it, whether std::terminate() makes sense for that program or not, and there is no way for a program to protect itself.
That the "it's okay if the process goes south should a rare combination of things happen" decision should be made by lower-level code (e.g., X dtor) for all apps that use it, and which doesn't even understand the context of any of the hundreds of apps that use it, just cannot be correct.
When a function throws an exception that is not in its exception-specification, std::unexpected() is called. According to 15.5.2 [except.unexpected] paragraph 2,
If [std::unexpected()] throws or rethrows an exception that the exception-specification does not allow then the following happens: If the exception-specification does not include the class std::bad_exception (18.8.2.1 [bad.exception]) then the function std::terminate() is called, otherwise the thrown exception is replaced by an implementation-defined object of the type std::bad_exception, and the search for another handler will continue at the call of the function whose exception-specification was violated.
The “replaced by” wording is imprecise and undefined. For example, does this mean that the destructor is called for the existing exception object, or is it simply abandoned? Is the replacement in situ, so that a pointer to the existing exception object will now point to the std::bad_exception object?
Mike Miller: The call to std::unexpected() is not described as analogous to invoking a handler, but if it were, that would resolve this question; it is clearly specified what happens to the previous exception object when a new exception is thrown from a handler (15.1 [except.throw] paragraph 4).
This approach would also clarify other questions that have been raised regarding the requirements for stack unwinding. For example, 15.5.1 [except.terminate] paragraph 2 says that
In the situation where no matching handler is found, it is implementation-defined whether or not the stack is unwound before std::terminate() is called.
This requirement could be viewed as in conflict with the statement in 15.5.2 [except.unexpected] paragraph 1 that
If a function with an exception-specification throws an exception that is not listed in the exception-specification, the function std::unexpected() is called (18.8.2 [exception.unexpected]) immediately after completing the stack unwinding for the former function.
If it is implementation-defined whether stack unwinding occurs before calling std::terminate() and std::unexpected() is called only after doing stack unwinding, does that mean that it is implementation-defined whether std::unexpected() is called if there is ultimately no handler found?
Again, if invoking std::unexpected() were viewed as essentially invoking a handler, the answer to this would be clear, because unwinding occurs before invoking a handler.
According to 16.1 [cpp.cond] paragraph 4,
The resulting tokens comprise the controlling constant expression which is evaluated according to the rules of 5.19 [expr.const] using arithmetic that has at least the ranges specified in 18.3 [support.limits], except that all signed and unsigned integer types act as if they have the same representation as, respectively, intmax_t or uintmax_t (18.4.2 [stdinth]). This includes interpreting character literals, which may involve converting escape sequences into execution character set members.
Ordinary character literals with a single c-char have the type char, which is neither a signed nor an unsigned integer type. Although 4.5 [conv.prom] paragraph 1 is clear that char values promote to int, regardless of whether the implementation treats char as having the values of signed char or unsigned char, 16.1 [cpp.cond] paragraph 4 isn't clear on whether character literals should be treated as signed or unsigned values. In C99, such literals have type int, so the question does not arise. If an implementation in which plain char has the values of unsigned char were to treat character literals as unsigned, an expression like '0'-'1' would thus have different values in C and C++, namely -1 in C and some large unsigned value in C++.
It is not clear from the Standard what the result of the following example should be:
#define NIL(xxx) xxx #define G_0(arg) NIL(G_1)(arg) #define G_1(arg) NIL(arg) G_0(42)
The relevant text from the Standard is found in 16.3.4 [cpp.rescan] paragraph 2:
If the name of the macro being replaced is found during this scan of the replacement list (not including the rest of the source file's preprocessing tokens), it is not replaced. Further, if any nested replacements encounter the name of the macro being replaced, it is not replaced. These nonreplaced macro name preprocessing tokens are no longer available for further replacement even if they are later (re)examined in contexts in which that macro name preprocessing token would otherwise have been replaced.
The sequence of expansion of G0(42) is as follows:
G0(42) NIL(G_1)(42) G_1(42) NIL(42)
The question is whether the use of NIL in the last line of this sequence qualifies for non-replacement under the cited text. If it does, the result will be NIL(42). If it does not, the result will be simply 42.
The original intent of the J11 committee in this text was that the result should be 42, as demonstrated by the original pseudo-code description of the replacement algorithm provided by Dave Prosser, its author. The English description, however, omits some of the subtleties of the pseudo-code and thus arguably gives an incorrect answer for this case.
Suggested resolution (Mike Miller): Replace the cited paragraph with the following:
As long as the scan involves only preprocessing tokens from a given macro's replacement list, or tokens resulting from a replacement of those tokens, an occurrence of the macro's name will not result in further replacement, even if it is later (re)examined in contexts in which that macro name preprocessing token would otherwise have been replaced.
Once the scan reaches the preprocessing token following a macro's replacement list — including as part of the argument list for that or another macro — the macro's name is once again available for replacement. [Example:
#define NIL(xxx) xxx #define G_0(arg) NIL(G_1)(arg) #define G_1(arg) NIL(arg) G_0(42) // result is 42, not NIL(42)The reason that NIL(42) is replaced is that (42) comes from outside the replacement list of NIL(G_1), hence the occurrence of NIL within the replacement list for NIL(G_1) (via the replacement of G_1(42)) is not marked as nonreplaceable. —end example]
(Note: The resolution of this issue must be coordinated with J11/WG14.)
Notes (via Tom Plum) from April, 2004 WG14 Meeting:
Back in the 1980's it was understood by several WG14 people that there were tiny differences between the "non-replacement" verbiage and the attempts to produce pseudo-code. The committee's decision was that no realistic programs "in the wild" would venture into this area, and trying to reduce the uncertainties is not worth the risk of changing conformance status of implementations or programs.
C99 is very clear that a #error directive causes a translation to fail: Clause 4 paragraph 4 says,
The implementation shall not successfully translate a preprocessing translation unit containing a #error preprocessing directive unless it is part of a group skipped by conditional inclusion.
C++, on the other hand, simply says that a #error directive “renders the program ill-formed” (16.5 [cpp.error]), and the only requirement for an ill-formed program is that a diagnostic be issued; the translation may continue and succeed. (Noted in passing: if this difference between C99 and C++ is addressed, it would be helpful for synchronization purposes in other contexts as well to introduce the term “preprocessing translation unit.”)
The specification of how the string-literal in a _Pragma operator is handled does not deal with the new kinds of string literals. 16.9 [cpp.pragma.op] says,
The string literal is destringized by deleting the L prefix, if present, deleting the leading and trailing double-quotes, replacing each escape sequence...
The various other prefixes should either be handled or prohibited.
During the discussion of issues 167 and 174, it became apparent that there was no consensus on the meaning of deprecation. Some thought that deprecating a feature reflected an intent to remove it from the language. Others viewed it more as an encouragement to programmers not to use certain constructs, even though they might be supported in perpetuity.
There is a formal-sounding definition of deprecation in Annex D [depr] paragraph 2:
deprecated is defined as: Normative for the current edition of the Standard, but not guaranteed to be part of the Standard in future revisions.However, this definition would appear to say that any non-deprecated feature is "guaranteed to be part of the Standard in future revisions." It's not clear that that implication was intended, so this definition may need to be amended.
This issue is intended to provide an avenue for discussing and resolving those questions, after which the original issues may be reopened if that is deemed desirable.