https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78420
--- Comment #13 from Jason Merrill <jason at gcc dot gnu.org> --- (In reply to Ville Voutilainen from comment #11) > Ah, the plot thickens. Jens Maurer wrote: > > "Regarding the std::less<T*> issue, it seems a bug in the standard > to require that it be constexpr and deliver a total order. After > all, the addresses of global objects are defined by the linker, > so it doesn't seem plausible to get a compile-time answer for > std::less<T*> that is the same as a later run-time answer." That doesn't make sense to me; you can call a constexpr function with non-constant arguments and get a non-constant result. There's no requirement that it give a total order at compile time, is there?