[llvm-bugs] [Bug 45088] New: Compound requirement type constraint uses decltype(E) instead of decltype((E))

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=45088

            Bug ID: 45088
           Summary: Compound requirement type constraint uses decltype(E)
                    instead of decltype((E))
           Product: clang
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: C++2a
          Assignee: unassignedclangbugs at nondot.org
          Reporter: kjcamann.lists at gmail.com
                CC: blitzrakete at gmail.com, erik.pilkington at gmail.com,
                    llvm-bugs at lists.llvm.org, richard-llvm at metafoo.co.uk

The following program should compile in clang but does not. It does work in gcc
10:

```
#include <type_traits>

template <typename T>
concept reference = std::is_reference_v<T>;

struct S {
  int i;
};

template <typename T>
concept C = requires(T &t) {
  { t.i } -> reference;
};

static_assert( C<S> ); // static_assert fails
```

The issue here is that according to [expr.prim.req.compound]p1.3.2, the type
substituted into the type constraint should be decltype((E)), with the extra
parentheses to inhibit the corner case rule for decltype
([dcl.type.decltype]p1.3) that makes id expressions and class member access
expressions into just T rather than the normal T& for lvalue category
expressions. clang uses `decltype(E)` rather than `decltype((E))` and even says
so in the diagnostic:

"note: because 'decltype(t.i)' (aka 'int')  does not satisfy 'reference'"

This bug is present in both clang 10 and in trunk.

-- 
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/20200303/3180275d/attachment.html>