glam 0.33.7

A simple and fast 3D math library for games and graphics
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71

#[cfg(target_arch = "wasm32")]
use core::arch::wasm32::*;
#[cfg(target_arch = "wasm64")]
use core::arch::wasm64::*;

pub const fn v128_from_f32x4(a: [f32; 4]) -> v128 {
    f32x4(a[0], a[1], a[2], a[3])
}

const PS_SIGN_MASK: v128 = v128_from_f32x4([-0.0; 4]);

/// Rounds each lane to the nearest integer, rounding half-way cases away from
/// zero, matching `f32::round`.
#[inline]
pub(crate) fn f32x4_round(v: v128) -> v128 {
    // `f32x4_nearest` rounds half-way cases to the nearest even integer. The two
    // roundings only differ on half-way cases that it rounded towards zero, and
    // those are exactly the lanes where the difference between the input and the
    // result is `0.5` with the sign of the input, so move those one further away
    // from zero.
    let rounded = f32x4_nearest(v);
    let sign = v128_and(v, PS_SIGN_MASK);
    let ties_down = f32x4_eq(f32x4_sub(v, rounded), v128_or(f32x4_splat(0.5), sign));
    v128_bitselect(
        f32x4_add(rounded, v128_or(f32x4_splat(1.0), sign)),
        rounded,
        ties_down,
    )
}

/// Calculates the vector 3 dot product and returns answer in x lane of v128.
#[inline(always)]
pub(crate) fn dot3_in_x(lhs: v128, rhs: v128) -> v128 {
    let x2_y2_z2_w2 = f32x4_mul(lhs, rhs);
    let y2_0_0_0 = i32x4_shuffle::<1, 0, 0, 0>(x2_y2_z2_w2, x2_y2_z2_w2);
    let z2_0_0_0 = i32x4_shuffle::<2, 0, 0, 0>(x2_y2_z2_w2, x2_y2_z2_w2);
    let x2y2_0_0_0 = f32x4_add(x2_y2_z2_w2, y2_0_0_0);
    f32x4_add(x2y2_0_0_0, z2_0_0_0)
}

/// Calculates the vector 4 dot product and returns answer in x lane of v128.
#[inline(always)]
pub(crate) fn dot4_in_x(lhs: v128, rhs: v128) -> v128 {
    let x2_y2_z2_w2 = f32x4_mul(lhs, rhs);
    let z2_w2_0_0 = i32x4_shuffle::<2, 3, 0, 0>(x2_y2_z2_w2, x2_y2_z2_w2);
    let x2z2_y2w2_0_0 = f32x4_add(x2_y2_z2_w2, z2_w2_0_0);
    let y2w2_0_0_0 = i32x4_shuffle::<1, 0, 0, 0>(x2z2_y2w2_0_0, x2z2_y2w2_0_0);
    f32x4_add(x2z2_y2w2_0_0, y2w2_0_0_0)
}

#[inline]
pub(crate) fn dot3(lhs: v128, rhs: v128) -> f32 {
    f32x4_extract_lane::<0>(dot3_in_x(lhs, rhs))
}

#[inline]
pub(crate) fn dot3_into_v128(lhs: v128, rhs: v128) -> v128 {
    let dot_in_x = dot3_in_x(lhs, rhs);
    i32x4_shuffle::<0, 0, 0, 0>(dot_in_x, dot_in_x)
}

#[inline]
pub(crate) fn dot4(lhs: v128, rhs: v128) -> f32 {
    f32x4_extract_lane::<0>(dot4_in_x(lhs, rhs))
}

#[inline]
pub(crate) fn dot4_into_v128(lhs: v128, rhs: v128) -> v128 {
    let dot_in_x = dot4_in_x(lhs, rhs);
    i32x4_shuffle::<0, 0, 0, 0>(dot_in_x, dot_in_x)
}