http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51079
Bug #: 51079
Summary: [C++0x] Conversion function template ignored in the
presence of non-template conversion function
Classification: Unclassified
Product: gcc
Version: 4.7.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: daniel.krueg...@googlemail.com
CC: ja...@redhat.com
gcc 4.7.0 20111105 (experimental) accepts the following code if compiled with
-Wall -std=c++0x:
//---
struct C1
{
template <class T>
operator T() = delete;
operator bool() { return false; }
} c1;
struct C2
{
private:
template <class T>
operator T();
public:
operator bool() { return false; }
} c2;
int ic1 = c1; // #1a
int ic2 = c2; // #1b
int ac1 = c1 + c1; // #2a
int ac2 = c2 + c2; // #2b
//---
After acceptance of the defect report
http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#495
the lines #1a/b should be rejected, because the conversion function template
should deduce operator int as a better match, making these accesses ill-formed.
Similarly, #2a/b should be rejected, because there should be a manifold of
possible candidates for operator+.
Bug 49531 looks related, but I believe there are enough differences in the test
case to justify a separate report.
