[Bug c++/108218] [12/13 Regression] Constant arguments in the new expression is not checked in unevaluated operand since r12-5253-g4df7f8c79835d569

["ppalka at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108218

Patrick Palka <ppalka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ppalka at gcc dot gnu.org
           See Also|                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=108219

--- Comment #9 from Patrick Palka <ppalka at gcc dot gnu.org> ---
(In reply to Steven Sun from comment #7)
> I got one simple idea as a workaround. I do not have the resources to do the
> tests.

FWIW that seems like the right idea, but it regresses
gcc/testsuite/g++.dg/DRs/dr2392.C:

$ cat gcc/testsuite/g++.dg/DRs/dr2392.C
// DR 2392
// { dg-do compile { target c++11 } }

template <class T = void>
constexpr int
foo ()
{
  T t;
  return 1;
}

using V = decltype (new int[foo ()]);

$ gcc gcc/testsuite/g++.dg/DRs/dr2392.C
gcc/testsuite/g++.dg/DRs/dr2392.C: In instantiation of ‘constexpr int foo()
[with T = void]’:
gcc/testsuite/g++.dg/DRs/dr2392.C:12:33:   required from here
gcc/testsuite/g++.dg/DRs/dr2392.C:8:5: error: variable or field ‘t’ declared
void
    8 |   T t;
      |     ^