floatUtils.hpp File Reference

Floating-point classification utilities that remain reliable under fast-math optimization. More...

Floating-point classification utilities that remain reliable under fast-math optimization.

Author

Jared R. Males

Definition in file floatUtils.hpp.

#include <bit>
#include <cstdint>
#include <limits>
#include <type_traits>

Go to the source code of this file.

Namespaces
namespace	mx
	The mxlib c++ namespace.

Functions
template<typename realT>
double	mx::math::floatUtils_detail::normalizedDouble (realT value)
	Convert an extended floating-point value to a classifiable double without overflowing finite values.
template<typename realT>
bool	mx::math::isNan (realT value)
	Test whether a floating-point value is NaN, including under finite-math-only optimization.
template<typename realT>
bool	mx::math::isFinite (realT value)
	Test whether a floating-point value is finite, including under finite-math-only optimization.

Function Documentation

normalizedDouble()

template<typename realT>

double mx::math::floatUtils_detail::normalizedDouble

(

realT

value

)

Convert an extended floating-point value to a classifiable double without overflowing finite values.

Parameters

[in]

value

floating-point value to normalize

Definition at line 25 of file floatUtils.hpp.