[Bug c++/18289] New: GCC3.4.2 selects incorrect function overload when an ellipsis overload is available

[Bug c++/18289] New: GCC3.4.2 selects incorrect function overload when an ellipsis overload is available

[aleethorp at hotmail dot com]

GCC 3.4.2 selects an overload that requires an inaccessible conversion when an
ellipsis function overload is available

Platform is Cygwin on Windows 2000.

See the example below.

class A {};
class B : public A {};
class C : private A {};

void f(...);
void f(A*);
void error()
{
  f(static_cast<B*>(0));
  f(static_cast<C*>(0)); // <- Line 42
}

Conversion.cc: In function `void error()':
Conversion.cc:42: error: `A' is an inaccessible base of `C'

-- 
           Summary: GCC3.4.2 selects incorrect function overload when an
                    ellipsis overload is available
           Product: gcc
           Version: 3.4.2
            Status: UNCONFIRMED
          Severity: minor
          Priority: P3
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: aleethorp at hotmail dot com
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18289

[Bug c++/18289] GCC3.4.2 selects incorrect function overload when an ellipsis overload is available

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-11-03 20:57 -------
Note ICC 8.0 also rejects this.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18289

[Bug c++/18289] GCC3.4.2 selects incorrect function overload when an ellipsis overload is available

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-11-03 21:03 -------
This is not a bug.

[13.3.3.2]/2 says that a standard conversion sequence is better
than an ellipsis conversion sequence.

The note in [13.3]/1 says:
The function selected by overload resolution is not guaranteed to
be appropriate for the context. Other restrictions, such as the
accessibility of the function, can make its use in the calling
context ill-formed.

Converting a pointer to a pointer of a base class is a standard
conversion. So gcc is right in choosing f(A*) and complaining
about it later.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |reichelt at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18289