Struct Vec3Fix 

#[repr(C)]
pub struct Vec3Fix {
    pub x: Fix128,
    pub y: Fix128,
    pub z: Fix128,
}

3D vector using Fix128 components

Fields

x: Fix128

X component

y: Fix128

Y component

z: Fix128

Z component

Implementations

impl Vec3Fix

pub fn checked_scale(self, s: Fix128) -> Option<Self>

各成分を s 倍し、どれか 1 成分でも範囲外なら None を返す

Fix128::checked_mul の Vec3Fix 版 発散検出の投入点 (積分 position += velocity*dt と速度導出) で使う

impl Vec3Fix

pub const ZERO: Self

Zero vector

pub const UNIT_X: Self

Unit X vector

pub const UNIT_Y: Self

Unit Y vector

pub const UNIT_Z: Self

Unit Z vector

pub const fn new(x: Fix128, y: Fix128, z: Fix128) -> Self

Create new vector

pub const fn from_int(x: i64, y: i64, z: i64) -> Self

Create from integers

pub fn from_f32(x: f32, y: f32, z: f32) -> Self

Create from f32 components (for SDF bridge) Create from f32 components (for SDF bridge, not deterministic!)

pub fn to_f32(self) -> (f32, f32, f32)

Convert to f32 tuple (for SDF bridge)

pub fn dot(self, rhs: Self) -> Fix128

Dot product

pub fn cross(self, rhs: Self) -> Self

Cross product

pub fn length_squared(self) -> Fix128

Squared length (no sqrt)

pub fn length(self) -> Fix128

Length (magnitude)

pub fn normalize(self) -> Self

Normalize to unit length.

Returns Self::ZERO for zero-length vectors. Use Self::try_normalize when you need to distinguish a zero-length input from a valid unit vector.

pub fn try_normalize(self) -> Option<Self>

Try to normalize, returning None for zero-length vectors.

pub fn normalize_with_length(self) -> (Self, Fix128)

Normalize and return both the unit vector and the original length.

Avoids double sqrt when both normalized direction and distance are needed. Returns (Self::ZERO, Fix128::ZERO) if the vector is zero-length.

pub fn checked_add(self, rhs: Self) -> Option<Self>

成分ごとの Fix128::checked_add (+ と同じ和、どれかの成分が |·| ≥ 2⁶³ なら None)

Claims

• Some の時は self + rhs と bit 一致する
• 位置の加算のように wrap が反対側の端への移動になる経路だけが使う

pub fn checked_dot(self, rhs: Self) -> Option<Fix128>

Self::dot と同じ値を計算し、積か和のどれかが範囲外なら None を返す

Claims

• Some の時は dot と bit 一致する (同じ積を同じ順で足す)
• 3 つの積のどれか、または途中の和が |·| ≥ 2⁶³ なら None

pub fn checked_length_squared(self) -> Option<Fix128>

Self::length_squared の範囲検査版 (|v|² ≥ 2⁶³ で None)

Some の時は length_squared と bit 一致する

pub fn checked_length(self) -> Option<Fix128>

Self::length の範囲検査版 (|v| ≥ 2³¹·⁵ で None)

Some の時は length と bit 一致する 範囲外でも長さが要るなら Self::checked_length_scaled を使う

pub fn checked_normalize(self) -> Option<Self>

Self::try_normalize の範囲検査版

長さ 0 と |v| ≥ 2³¹·⁵ (長さの 2 乗が範囲外) のどちらも None Some の時は try_normalize と bit 一致する 範囲外でも向きが要るなら Self::try_normalize_scaled を使う

pub fn checked_length_scaled(self) -> Option<Fix128>

2 乗が範囲外になる大きさでも正しい長さ 結果自体が ≥ 2⁶³ (表せない) 時だけ None

Claims

• |v|² < 2⁶³ (length_squared が wrap しない範囲) では length と bit 一致する (同じ式をそのまま使う)
• それより大きい時は、最大成分を 2 の冪で [2²⁹, 2³⁰) に寄せてから 2 乗するので wrap せず、相対誤差は約 2⁻⁹² (成分の切り捨てと sqrt の floor)
• 結果が ≥ 2⁶³ (成分が 2⁶² 級) なら None

pub fn try_normalize_scaled(self) -> Option<Self>

2 乗が範囲外になる大きさでも向きを返す正規化 零ベクトルだけ None

Claims

• try_normalize が Some を返す入力では try_normalize と bit 一致
• try_normalize が範囲外 (|v| ≥ 2³¹·⁵) や長さの 2 乗の下位桁あふれ (|v| < 約 2⁻³²) で None を返す非零ベクトルには、最大成分を [2²⁹, 2³⁰) に寄せたベクトルを正規化して返す

pub fn scale(self, s: Fix128) -> Self

Scale by scalar

pub fn dot_simd(self, rhs: Self) -> Fix128

SIMD-accelerated dot product — safe public entry point.

On x86_64 with the simd feature enabled this dispatches to the SSE2 path (dot_simd_sse2). On every other platform it falls back to the scalar dot(). The result is bit-exact identical to dot() on every platform.

pub fn length_squared_simd(self) -> Fix128

Squared length using SIMD-accelerated dot product.

Equivalent to length_squared() but dispatches through dot_simd. Result is bit-exact identical to length_squared().

pub fn cross_simd(self, rhs: Self) -> Self

Cross product through the simd feature’s API surface — scalar-equivalent, bit-identical to cross() (see the module docs).

pub fn dot_batch_4(a: [Self; 4], b: [Self; 4]) -> [Fix128; 4]

Batch dot product for 4 vector pairs — scalar-equivalent (four dot() calls, bit-identical). The batch shape is the API a future AVX-512 / NEON path would fill in.

Trait Implementations

impl Add for Vec3Fix

type Output = Vec3Fix

The resulting type after applying the + operator.

fn add(self, rhs: Self) -> Self

Performs the + operation.

impl Clone for Vec3Fix

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for Vec3Fix

impl Debug for Vec3Fix

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Vec3Fix

fn default() -> Self

Returns the “default value” for a type.

impl Display for Vec3Fix

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Div<Fix128> for Vec3Fix

type Output = Vec3Fix

The resulting type after applying the / operator.

fn div(self, rhs: Fix128) -> Self

Performs the / operation.

impl Eq for Vec3Fix

impl From<Vec3Fix> for [Fix128; 3]

fn from(v: Vec3Fix) -> Self

Converts to this type from the input type.

impl From<[Fix128; 3]> for Vec3Fix

fn from(arr: [Fix128; 3]) -> Self

Converts to this type from the input type.

impl Mul<Fix128> for Vec3Fix

type Output = Vec3Fix

The resulting type after applying the * operator.

fn mul(self, rhs: Fix128) -> Self

Performs the * operation.

impl Neg for Vec3Fix

type Output = Vec3Fix

The resulting type after applying the - operator.

fn neg(self) -> Self

Performs the unary - operation.

impl PartialEq for Vec3Fix

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Vec3Fix

impl Sub for Vec3Fix

type Output = Vec3Fix

The resulting type after applying the - operator.

fn sub(self, rhs: Self) -> Self

Performs the - operation.