[Bug c++/13963] New: gcc rejects array of unknown bounds as parameter of function template specialization

[austern at apple dot com]

Every version of g++ that I've tested (2.95, 3.1, 3.3, mainline) rejects the following code:

  template<typename X> void h(X x) { }

  void foo()
  {
    int A[9];
    h<int[]>(A);
  }

EDG accepts this code.  I believe EDG is correct.  When instantiating h, int[] should be treated as a 
synonym for int*.

My analysis:
 - First, just to clear away an irrelevancy: there's a standard array-to-pointer conversion, but it 
doesn't matter.  We're talking about the legality of a declaration.
 - Second, again mostly as a sanity check: void f(int a[]) { } is legal C++, just like it's legal C.  But 
the reason it's legal matters: clause 8.3.5, paragraph 3, says: The type of each parameter is 
determined from its own decl-specifier-seq and declarator. After determining the type of each 
parameter, any parameter of type “array of T” or “function returningT” is adjusted to be “pointer 
toT” or “pointer to function returningT,” respectively.
 - Finally, the only remaining question is: when instantiating a function template whose template 
arguments are explicitly provided, is the compiler supposed to perform the same parameter 
adjustments as in an ordinary function declaration?  The answer is yes.  14.8.2, paragraph 3, says: 
After this substitution is performed, the function parameter type adjustments described in 8.3.5 
are performed.

-- 
           Summary: gcc rejects array of unknown bounds as parameter of
                    function template specialization
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: austern at apple dot com
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i386-redhat-linux
  GCC host triplet: i386-redhat-linux
GCC target triplet: i386-redhat-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13963