https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104485
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jsm28 at gcc dot gnu.org
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
To some extent fmod (x, 1.) looks equivalent to C99 modf (x, &tem) where modf
might be cheaper on the libm side. With -ffast-math one might be
tempted to calculate fmod (x, 1.) as x - (long)x or sth similar (though the
excessive exponent of doubles make this flawed).
