[][src]Function directx_math::XMVector3Normalize

pub fn XMVector3Normalize(V: FXMVECTOR) -> FXMVECTOR

Returns the normalized version of a 3D vector.

Parameters

V 3D vector.

Return value

Returns the normalized version of V.

Remarks

For a vector of length 0, this function returns a zero vector. For a vector with infinite length, it returns a vector of QNaN.

Note that for most graphics applications, ensuring the vectors have well-defined lengths that don't cause problems for normalization is common practice. However, if you need a robust normalization that works for all floating-point inputs, you can use the following code instead:

inline XMVECTOR XMVector3NormalizeRobust( FXMVECTOR V )
{
    // Compute the maximum absolute value component.
    XMVECTOR vAbs = XMVectorAbs(V);
    XMVECTOR max0 = XMVectorSplatX(vAbs);
    XMVECTOR max1 = XMVectorSplatY(vAbs);
    XMVECTOR max2 = XMVectorSplatZ(vAbs);
    max0 = XMVectorMax(max0, max1);
    max0 = XMVectorMax(max0, max2);

    // Divide by the maximum absolute component.
    XMVECTOR normalized = XMVectorDivide(V, max0);

    // Set to zero when the original length is zero.
    XMVECTOR mask = XMVectorNotEqual(g_XMZero, max0);
    normalized = XMVectorAndInt(normalized, mask);

    XMVECTOR t0 = XMVector3LengthSq(normalized);
    XMVECTOR length = XMVectorSqrt(t0);

    // Divide by the length to normalize.
    normalized = XMVectorDivide(normalized, length);

    // Set to zero when the original length is zero or infinity.  In the
    // latter case, this is considered to be an unexpected condition.
    normalized = XMVectorAndInt(normalized, mask);
    return normalized;
}

Reference

https://docs.microsoft.com/en-us/windows/win32/api/directxmath/nf-directxmath-XMVector3Normalize