Created attachment 37094 [details] Original source file from Gentoo Bugzilla Found by Hunter Allen and reported at the Gentoo Bugzilla as https://bugs.gentoo.org/show_bug.cgi?id=568908. As requested in PR68761 by Jakub Jelinek I am opening this standalone report. Creduced case: namespace std { inline namespace __cxx11 {} template <typename> class allocator; namespace __cxx11 { template <typename _CharT, typename = _CharT, typename = allocator<_CharT>> class basic_string; typedef basic_string<char> string; } template <typename> class allocator { public: template <typename> struct rebind { typedef allocator other; }; }; template <typename _Alloc, typename _Tp> struct __alloctr_rebind { typedef typename _Alloc::template rebind<_Tp>::other __type; }; template <typename _Alloc> struct allocator_traits { typedef decltype(0) pointer; template <typename _Tp> using rebind_alloc = typename __alloctr_rebind<_Alloc, _Tp>::__type; }; } namespace __gnu_cxx { template <typename _Alloc> struct __alloc_traits : std::allocator_traits<_Alloc> { typedef std::allocator_traits<_Alloc> _Base_type; template <typename _Tp> struct rebind { typedef typename _Base_type::template rebind_alloc<_Tp> other; }; }; } namespace std { namespace __cxx11 { template <typename _CharT, typename, typename _Alloc> class basic_string { typedef typename __gnu_cxx::__alloc_traits<_Alloc>::template rebind<_CharT>::other _Char_alloc_type; typedef __gnu_cxx::__alloc_traits<_Char_alloc_type> _Alloc_traits; public: typedef _Char_alloc_type allocator_type; typedef typename _Alloc_traits::pointer pointer; struct _Alloc_hider { _Alloc_hider(pointer, _Alloc); } _M_dataplus; pointer _M_local_data(); allocator_type _M_get_allocator(); basic_string(const basic_string &) : _M_dataplus(_M_local_data(), _M_get_allocator()) {} basic_string(_CharT *, _Alloc __a = _Alloc()) : _M_dataplus(_M_local_data(), __a) {} }; } } auto vglambda = [](auto printer) { return [=](auto &&... ts) { return [=] { printer(ts...); }; }; }; main() { vglambda([](auto, auto, auto) {})(std::string(""), "", ""); }
What is going on is that cp_genericize_r first converts changes a by-reference RESULT_DECL into INDIRECT_REF of it, and then (dunno if because of tree sharing or some constexpr stuff) at some point that IL is passed again to cp_genericize_r again and again performs the same change, so we in the end have two nested INDIRECT_REFs and the types don't match any longer.
Fixed with r242766 and r243407 for 6.x. *** This bug has been marked as a duplicate of bug 77739 ***