[Bug c++/123189] [16 Regression] ICE in gromacs since r16-6140-g4a718209d448f6

[dhruvc at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123189

--- Comment #2 from Dhruv Chawla <dhruvc at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #1)
> If before the commit the reduced testcase is rejected it's not valid.  I
> suspect this is over-reduced.  You can adjust the reduction script to verify
> the reduced input is still accepted before the offending rev.

Hi Richi, I tried reducing this again:

===
namespace {
namespace {
template <typename a> class b { friend b c<>() noexcept(a ::d); };
} // namespace
} // namespace
===

This no longer gives an error with the passing commit.