https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69350
--- Comment #3 from Jonathan Wakely <redi at gcc dot gnu.org> --- I'm inclined to close this one as WONTFIX. We should fix PR 11196 but I don't really care about this C++98 non-conformance now, and we've been doing it this way since before C++11 was even a thing. Defining those <math.h> functions before C++11 is not the same as dumping a load of totally non-standard GNU extensions into the global namespace.
