| Issue |
83409
|
| Summary |
clang: initialization with assignment _expression_ possibly triggers wrong template instantiation
|
| Labels |
clang
|
| Assignees |
|
| Reporter |
bebuch
|
```cpp
template <typename T>
struct holder {
holder() = default;
constexpr ~holder() {
static_assert(sizeof(T) || true);
}
};
struct Incomplete;
struct Class {
Class();
~Class();
holder<Incomplete> a{}; // Reject: MSVC; Accept: GCC, clang
holder<Incomplete> b = {}; // Reject: MSVC, GCC; Accept: clang
holder<Incomplete> c = holder<Incomplete>{}; // Reject: MSVC, GCC, clang
};
int main() {
[[maybe_unused]] Class v;
}
```
Live Code: [Case `a`](https://godbolt.org/z/cYWTTMeoa), [Case `b`](https://godbolt.org/z/xf45795b1), [Case `c`](https://godbolt.org/z/Y764zPqKa).
I assume that all 3 cases should be valid in C++20/23, but I'm not sure.
Note that this compiles fine with MSVC and GCC, only clang reject:
```cpp
template <typename T>
struct holder {
holder() = default;
constexpr ~holder() {
static_assert(sizeof(T) || true);
}
};
struct Incomplete;
template <typename=void> // make Class a template
struct Class {
Class();
~Class();
holder<Incomplete> a{};
holder<Incomplete> b = {};
holder<Incomplete> c = holder<Incomplete>{};
};
int main() {
[[maybe_unused]] Class v;
}
```
[Live Code](https://godbolt.org/z/1so75fa5x)
There are these Bug reports to MSVC:
- https://developercommunity.visualstudio.com/t/10514504
- https://github.com/microsoft/STL/issues/4417
- https://developercommunity.visualstudio.com/t/10604135
And this one to GCC:
- https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114076
Note comment from Jiang An:
> I don't see why there's even a temporary value [in case `c`] since C++17. The prvalue is used to initialize the data member (via temporary materialization). The potential invocation of destructor should be in the body of constructors.
I also asked about this at StackOverflow:
- https://stackoverflow.com/questions/78040562
The original `holder` class template was `std::unique_ptr`, which became `constexpr` in C++23.
![]()
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
