[llvm-bugs] [Bug 42330] New: cosntexpr constructor breaks out expression SFINAE
via llvm-bugs
llvm-bugs at lists.llvm.org
Wed Jun 19 09:15:32 PDT 2019
https://bugs.llvm.org/show_bug.cgi?id=42330
Bug ID: 42330
Summary: cosntexpr constructor breaks out expression SFINAE
Product: clang
Version: trunk
Hardware: PC
OS: Windows NT
Status: NEW
Severity: enhancement
Priority: P
Component: C++
Assignee: unassignedclangbugs at nondot.org
Reporter: lichray at gmail.com
CC: blitzrakete at gmail.com, dgregor at apple.com,
erik.pilkington at gmail.com, llvm-bugs at lists.llvm.org,
richard-llvm at metafoo.co.uk
Created attachment 22124
--> https://bugs.llvm.org/attachment.cgi?id=22124&action=edit
reproduce
Clang does not compile the following program:
#include <type_traits>
struct A {};
struct B {};
template <class T>
struct MyPtr {
MyPtr() = default;
template <class U>
constexpr
MyPtr(MyPtr<U> v) : ptr_(v.ptr_) {}
T* ptr_;
};
template <class T>
struct __identity { using type = T; };
template <class... _Types>
struct __overload;
template <>
struct __overload<> { void operator()() const; };
template <class _Tp, class... _Types>
struct __overload<_Tp, _Types...> : __overload<_Types...> {
using __overload<_Types...>::operator();
static auto __test(_Tp (&&)[1]) -> __identity<_Tp>;
template <class _Up>
auto operator()(_Tp, _Up&& __t) const
-> decltype(__test({ std::forward<_Up>(__t) }));
};
template <class _Tp, class... _Types>
using __best_match_t =
typename std::invoke_result_t<__overload<_Types...>, _Tp, _Tp>::type;
void test() {
using t = __best_match_t<MyPtr<A>, MyPtr<A>, MyPtr<B>>;
}
<source>:11:23: error: cannot initialize a member subobject of type 'B *' with
an lvalue of type 'A *'
MyPtr(MyPtr<U> v) : ptr_(v.ptr_) {}
^ ~~~~~~
<source>:32:26: note: in instantiation of function template specialization
'MyPtr<B>::MyPtr<A>' requested here
-> decltype(__test({ std::forward<_Up>(__t) }));
^
[...]/include/c++/v1/type_traits:4425:23: note: while substituting deduced
template arguments into function template 'operator()' [with _Up = MyPtr<A>]
Essentially it is saying that a constexpr constructor is called and failed
during substitution, but that shouldn't happen. What should happen though is
MyPtr<A> being considered convertible to MyPtr<B>, but it's a worse match.
MSVC gives the right answer.
Commenting out the `constexpr` on line 10 fixes this issue.
--
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20190619/2129b8e8/attachment.html>
More information about the llvm-bugs
mailing list