colconv 0.1.0

SIMD-dispatched color-conversion kernels covering the FFmpeg AVPixelFormat space, with a Sink-based API so consumers pick which derived outputs (RGB / Luma / HSV / custom) they want without paying for the ones they don't.
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
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201

//! AVX2 kernels for planar GBR sources (Tier 10).
//!
//! AVX2's planar→packed-RGB interleave has no widely-used native
//! 256-bit primitive (`vpshufb` works per-lane, but the 3-channel
//! pattern crosses 128-bit lanes); the cleanest approach is to reuse
//! the shared 128-bit [`super::write_rgb_16`] / [`super::write_rgba_16`]
//! helpers twice per iteration and process 32 pixels per outer-loop
//! step. Same scalar tail as the SSE4.1 kernel.

use core::arch::x86_64::*;

use super::*;

/// AVX2 G/B/R planar → packed `R, G, B`. Processes 32 pixels per
/// iteration via two calls to the shared 128-bit
/// [`super::write_rgb_16`] helper.
///
/// # Safety
///
/// 1. AVX2 must be available (caller obligation). AVX2 implies SSE4.1
///    (and SSSE3), so the underlying `_mm_shuffle_epi8` in
///    `write_rgb_16` is legal.
/// 2. `g.len()`, `b.len()`, `r.len()` ≥ `width`.
/// 3. `rgb_out.len()` ≥ `3 * width`.
#[inline]
#[target_feature(enable = "avx2")]
pub(crate) unsafe fn gbr_to_rgb_row(
  g: &[u8],
  b: &[u8],
  r: &[u8],
  rgb_out: &mut [u8],
  width: usize,
) {
  debug_assert!(g.len() >= width, "g row too short");
  debug_assert!(b.len() >= width, "b row too short");
  debug_assert!(r.len() >= width, "r row too short");
  debug_assert!(rgb_out.len() >= width * 3, "rgb_out row too short");

  // SAFETY: AVX2 (incl. SSSE3 superset) per caller obligation.
  unsafe {
    let mut x = 0usize;
    while x + 32 <= width {
      let g_lo = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_lo = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_lo = _mm_loadu_si128(r.as_ptr().add(x).cast());
      write_rgb_16(r_lo, g_lo, b_lo, rgb_out.as_mut_ptr().add(x * 3));

      let g_hi = _mm_loadu_si128(g.as_ptr().add(x + 16).cast());
      let b_hi = _mm_loadu_si128(b.as_ptr().add(x + 16).cast());
      let r_hi = _mm_loadu_si128(r.as_ptr().add(x + 16).cast());
      write_rgb_16(r_hi, g_hi, b_hi, rgb_out.as_mut_ptr().add((x + 16) * 3));

      x += 32;
    }
    // Process any remaining 16-pixel block before the scalar tail.
    if x + 16 <= width {
      let g_v = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_v = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_v = _mm_loadu_si128(r.as_ptr().add(x).cast());
      write_rgb_16(r_v, g_v, b_v, rgb_out.as_mut_ptr().add(x * 3));
      x += 16;
    }
    if x < width {
      scalar::gbr_to_rgb_row(
        &g[x..width],
        &b[x..width],
        &r[x..width],
        &mut rgb_out[x * 3..width * 3],
        width - x,
      );
    }
  }
}

/// AVX2 G/B/R/A planar → packed `R, G, B, A` (real alpha plane).
///
/// # Safety
///
/// Same as [`gbr_to_rgb_row`] plus `a.len()` ≥ `width`,
/// `rgba_out.len()` ≥ `4 * width`.
#[inline]
#[target_feature(enable = "avx2")]
pub(crate) unsafe fn gbra_to_rgba_row(
  g: &[u8],
  b: &[u8],
  r: &[u8],
  a: &[u8],
  rgba_out: &mut [u8],
  width: usize,
) {
  debug_assert!(g.len() >= width, "g row too short");
  debug_assert!(b.len() >= width, "b row too short");
  debug_assert!(r.len() >= width, "r row too short");
  debug_assert!(a.len() >= width, "a row too short");
  debug_assert!(rgba_out.len() >= width * 4, "rgba_out row too short");

  // SAFETY: see `gbr_to_rgb_row`.
  unsafe {
    let mut x = 0usize;
    while x + 32 <= width {
      let g_lo = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_lo = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_lo = _mm_loadu_si128(r.as_ptr().add(x).cast());
      let a_lo = _mm_loadu_si128(a.as_ptr().add(x).cast());
      write_rgba_16(r_lo, g_lo, b_lo, a_lo, rgba_out.as_mut_ptr().add(x * 4));

      let g_hi = _mm_loadu_si128(g.as_ptr().add(x + 16).cast());
      let b_hi = _mm_loadu_si128(b.as_ptr().add(x + 16).cast());
      let r_hi = _mm_loadu_si128(r.as_ptr().add(x + 16).cast());
      let a_hi = _mm_loadu_si128(a.as_ptr().add(x + 16).cast());
      write_rgba_16(
        r_hi,
        g_hi,
        b_hi,
        a_hi,
        rgba_out.as_mut_ptr().add((x + 16) * 4),
      );

      x += 32;
    }
    if x + 16 <= width {
      let g_v = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_v = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_v = _mm_loadu_si128(r.as_ptr().add(x).cast());
      let a_v = _mm_loadu_si128(a.as_ptr().add(x).cast());
      write_rgba_16(r_v, g_v, b_v, a_v, rgba_out.as_mut_ptr().add(x * 4));
      x += 16;
    }
    if x < width {
      scalar::gbra_to_rgba_row(
        &g[x..width],
        &b[x..width],
        &r[x..width],
        &a[x..width],
        &mut rgba_out[x * 4..width * 4],
        width - x,
      );
    }
  }
}

/// AVX2 G/B/R planar → packed `R, G, B, A` with constant `α = 0xFF`.
///
/// # Safety
///
/// Same as [`gbr_to_rgb_row`] plus `rgba_out.len()` ≥ `4 * width`.
#[inline]
#[target_feature(enable = "avx2")]
pub(crate) unsafe fn gbr_to_rgba_opaque_row(
  g: &[u8],
  b: &[u8],
  r: &[u8],
  rgba_out: &mut [u8],
  width: usize,
) {
  debug_assert!(g.len() >= width, "g row too short");
  debug_assert!(b.len() >= width, "b row too short");
  debug_assert!(r.len() >= width, "r row too short");
  debug_assert!(rgba_out.len() >= width * 4, "rgba_out row too short");

  // SAFETY: see `gbr_to_rgb_row`.
  unsafe {
    let opaque = _mm_set1_epi8(-1);
    let mut x = 0usize;
    while x + 32 <= width {
      let g_lo = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_lo = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_lo = _mm_loadu_si128(r.as_ptr().add(x).cast());
      write_rgba_16(r_lo, g_lo, b_lo, opaque, rgba_out.as_mut_ptr().add(x * 4));

      let g_hi = _mm_loadu_si128(g.as_ptr().add(x + 16).cast());
      let b_hi = _mm_loadu_si128(b.as_ptr().add(x + 16).cast());
      let r_hi = _mm_loadu_si128(r.as_ptr().add(x + 16).cast());
      write_rgba_16(
        r_hi,
        g_hi,
        b_hi,
        opaque,
        rgba_out.as_mut_ptr().add((x + 16) * 4),
      );

      x += 32;
    }
    if x + 16 <= width {
      let g_v = _mm_loadu_si128(g.as_ptr().add(x).cast());
      let b_v = _mm_loadu_si128(b.as_ptr().add(x).cast());
      let r_v = _mm_loadu_si128(r.as_ptr().add(x).cast());
      write_rgba_16(r_v, g_v, b_v, opaque, rgba_out.as_mut_ptr().add(x * 4));
      x += 16;
    }
    if x < width {
      scalar::gbr_to_rgba_opaque_row(
        &g[x..width],
        &b[x..width],
        &r[x..width],
        &mut rgba_out[x * 4..width * 4],
        width - x,
      );
    }
  }
}