FastInvSqrt.h File Reference

Detailed Description

Optimized implementation of 1/sqrt( double )

Author

Martin Bauer marti.nosp@m.n.ba.nosp@m.uer@f.nosp@m.au.d.nosp@m.e

#include "core/DataTypes.h"

Classes
union	walberla::math::DoubleAndLongLong
union	walberla::math::FloatAndLong

Namespaces
namespace	walberla
	Storage for detected contacts which can be used to perform actions for all contacts, e.g.
namespace	walberla::math

Functions
template<unsigned int numIter>
double	walberla::math::fastInvSqrt (double y)
	Computes 1.0 / sqrt(x) with limited accuracy.
template<unsigned int numIter>
float	walberla::math::fastInvSqrt (float y)
	Float version, for documentation see above.