https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77776
g.peterh...@t-online.de changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |g.peterh...@t-online.de --- Comment #11 from g.peterh...@t-online.de --- Would this be a good implementation for hypot3 in cmath? #define GCC_UNLIKELY(x) __builtin_expect(x, 0) #define GCC_LIKELY(x) __builtin_expect(x, 1) namespace __detail { template <typename _Tp> inline _GLIBCXX_CONSTEXPR typename enable_if<is_floating_point<_Tp>::value, bool>::type __isinf3(const _Tp __x, const _Tp __y, const _Tp __z) noexcept { return bool(int(std::isinf(__x)) | int(std::isinf(__y)) | int(std::isinf(__z))); } template <typename _Tp> inline _GLIBCXX_CONSTEXPR typename enable_if<is_floating_point<_Tp>::value, _Tp>::type __hypot3(_Tp __x, _Tp __y, _Tp __z) noexcept { __x = std::fabs(__x); __y = std::fabs(__y); __z = std::fabs(__z); const _Tp __max = std::fmax(std::fmax(__x, __y), __z); if (GCC_UNLIKELY(__max == _Tp{})) { return __max; } else { __x /= __max; __y /= __max; __z /= __max; return std::sqrt(__x*__x + __y*__y + __z*__z) * __max; } } } // __detail template <typename _Tp> inline _GLIBCXX_CONSTEXPR typename enable_if<is_floating_point<_Tp>::value, _Tp>::type __hypot3(const _Tp __x, const _Tp __y, const _Tp __z) noexcept { return (GCC_UNLIKELY(__detail::__isinf3(__x, __y, __z))) ? numeric_limits<_Tp>::infinity() : __detail::__hypot3(__x, __y, __z); } #undef GCC_UNLIKELY #undef GCC_LIKELY How does it work? * Basically, I first pull out the special case INFINITY (see https://en.cppreference.com/w/cpp/numeric/math/hypot). * As an additional safety measure (to prevent misuse) the functions are defined by enable_if. constexpr * The hypot3 functions can thus be defined as _GLIBCXX_CONSTEXPR. Questions * To get a better runtime behavior I define GCC_(UN)LIKELY. Are there already such macros (which I have overlooked)? * The functions are noexcept. Does that make sense? If yes: why are the math functions not noexcept? thx Gero