yuv 0.8.13

High performance utilities for YUV format handling and conversion.
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
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279

/*
 * Copyright (c) Radzivon Bartoshyk, 10/2024. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 * 1.  Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 *
 * 2.  Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * 3.  Neither the name of the copyright holder nor the names of its
 * contributors may be used to endorse or promote products derived from
 * this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

use crate::avx2::avx2_utils::*;
use crate::yuv_support::{CbCrInverseTransform, YuvPacked444Format, YuvSourceChannels};
#[cfg(target_arch = "x86")]
use std::arch::x86::*;
#[cfg(target_arch = "x86_64")]
use std::arch::x86_64::*;

pub(crate) fn avx2_ayuv_to_rgba<const DESTINATION_CHANNELS: u8, const PACKING: u8>(
    ayuv: &[u8],
    rgba: &mut [u8],
    transform: &CbCrInverseTransform<i16>,
    bias_y: i16,
    bias_uv: i16,
    width: usize,
    use_premultiply: bool,
) {
    unsafe {
        avx2_ayuv_to_rgba_impl::<DESTINATION_CHANNELS, PACKING>(
            ayuv,
            rgba,
            transform,
            bias_y,
            bias_uv,
            width,
            use_premultiply,
        )
    }
}

#[inline(always)]
unsafe fn _mm256_load_deintl_ayuv<const PACKED: u8>(
    src: &[u8],
) -> (__m256i, __m256i, __m256i, __m256i) {
    let packing: YuvPacked444Format = PACKED.into();

    let r0 = _mm256_loadu_si256(src.as_ptr() as *const __m256i);
    let r1 = _mm256_loadu_si256(src.get_unchecked(32..).as_ptr() as *const __m256i);
    let r2 = _mm256_loadu_si256(src.get_unchecked(64..).as_ptr() as *const __m256i);
    let r3 = _mm256_loadu_si256(src.get_unchecked(96..).as_ptr() as *const __m256i);
    let (a0, a1, a2, a3) = _mm256_deinterleave_rgba_epi8(r0, r1, r2, r3);

    match packing {
        YuvPacked444Format::Ayuv => (a0, a1, a2, a3),
        YuvPacked444Format::Vuya => (a3, a2, a1, a0),
    }
}

#[target_feature(enable = "avx2")]
unsafe fn avx2_ayuv_to_rgba_impl<const DESTINATION_CHANNELS: u8, const PACKED: u8>(
    ayuv: &[u8],
    rgba: &mut [u8],
    transform: &CbCrInverseTransform<i16>,
    bias_y: i16,
    bias_uv: i16,
    width: usize,
    use_premultiply: bool,
) {
    let destination_channels: YuvSourceChannels = DESTINATION_CHANNELS.into();
    let channels = destination_channels.get_channels_count();

    let mut cx = 0usize;
    let rgba_ptr = rgba.as_mut_ptr();

    let y_corr = _mm256_set1_epi8(bias_y as i8);
    let uv_corr = _mm256_set1_epi16(((bias_uv) << 2) | ((bias_uv) >> 6));
    let v_luma_coeff = _mm256_set1_epi16(transform.y_coef);
    let v_cr_coeff = _mm256_set1_epi16(transform.cr_coef);
    let v_cb_coeff = _mm256_set1_epi16(transform.cb_coef);
    let v_g_coeff_1 = _mm256_set1_epi16(transform.g_coeff_1);
    let v_g_coeff_2 = _mm256_set1_epi16(transform.g_coeff_2);

    while cx + 32 <= width {
        let (a, mut y_vals, u, v) = _mm256_load_deintl_ayuv::<PACKED>(ayuv.get_unchecked(cx * 4..));

        y_vals = _mm256_subs_epu8(y_vals, y_corr);

        let u_high_u16 = _mm256_srli_epi16::<6>(_mm256_unpackhi_epi8(u, u));
        let v_high_u16 = _mm256_srli_epi16::<6>(_mm256_unpackhi_epi8(v, v));
        let u_low_u16 = _mm256_srli_epi16::<6>(_mm256_unpacklo_epi8(u, u));
        let v_low_u16 = _mm256_srli_epi16::<6>(_mm256_unpacklo_epi8(v, v));

        let y0_10 = _mm256_expand8_unordered_to_10(y_vals);

        let u_high = _mm256_sub_epi16(u_high_u16, uv_corr);
        let v_high = _mm256_sub_epi16(v_high_u16, uv_corr);
        let y_high = _mm256_mulhrs_epi16(y0_10.1, v_luma_coeff);

        let rhc = _mm256_mulhrs_epi16(v_high, v_cr_coeff);
        let bhc = _mm256_mulhrs_epi16(u_high, v_cb_coeff);
        let ghc0 = _mm256_mulhrs_epi16(v_high, v_g_coeff_1);
        let ghc1 = _mm256_mulhrs_epi16(u_high, v_g_coeff_2);

        let r_high = _mm256_add_epi16(y_high, rhc);
        let b_high = _mm256_add_epi16(y_high, bhc);
        let g_high = _mm256_sub_epi16(y_high, _mm256_add_epi16(ghc0, ghc1));

        let u_low = _mm256_sub_epi16(u_low_u16, uv_corr);
        let v_low = _mm256_sub_epi16(v_low_u16, uv_corr);
        let y_low = _mm256_mulhrs_epi16(y0_10.0, v_luma_coeff);

        let rlc = _mm256_mulhrs_epi16(v_low, v_cr_coeff);
        let blc = _mm256_mulhrs_epi16(u_low, v_cb_coeff);
        let glc0 = _mm256_mulhrs_epi16(v_low, v_g_coeff_1);
        let glc1 = _mm256_mulhrs_epi16(u_low, v_g_coeff_2);

        let r_low = _mm256_add_epi16(y_low, rlc);
        let b_low = _mm256_add_epi16(y_low, blc);
        let g_low = _mm256_sub_epi16(y_low, _mm256_add_epi16(glc0, glc1));

        let mut r_values = _mm256_packus_epi16(r_low, r_high);
        let mut g_values = _mm256_packus_epi16(g_low, g_high);
        let mut b_values = _mm256_packus_epi16(b_low, b_high);

        if use_premultiply {
            let a_high = _mm256_unpackhi_epi8(a, _mm256_setzero_si256());
            let a_low = _mm256_unpacklo_epi8(a, _mm256_setzero_si256());

            let r_low = _mm256_unpacklo_epi8(r_values, _mm256_setzero_si256());
            let r_high = _mm256_unpackhi_epi8(r_values, _mm256_setzero_si256());
            let g_low = _mm256_unpacklo_epi8(g_values, _mm256_setzero_si256());
            let g_high = _mm256_unpackhi_epi8(g_values, _mm256_setzero_si256());
            let b_low = _mm256_unpacklo_epi8(b_values, _mm256_setzero_si256());
            let b_high = _mm256_unpackhi_epi8(b_values, _mm256_setzero_si256());

            let (r_l, r_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(r_low, a_low),
                _mm256_mullo_epi16(r_high, a_high),
            );
            let (g_l, g_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(g_low, a_low),
                _mm256_mullo_epi16(g_high, a_high),
            );
            let (b_l, b_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(b_low, a_low),
                _mm256_mullo_epi16(b_high, a_high),
            );

            r_values = _mm256_packus_epi16(r_l, r_h);
            g_values = _mm256_packus_epi16(g_l, g_h);
            b_values = _mm256_packus_epi16(b_l, b_h);
        }

        let dst_shift = cx * channels;

        _mm256_store_interleave_rgb_for_yuv::<DESTINATION_CHANNELS>(
            rgba_ptr.add(dst_shift),
            r_values,
            g_values,
            b_values,
            a,
        );

        cx += 32;
    }

    if cx < width {
        let diff = width - cx;

        assert!(diff <= 32);

        let mut dst_buffer: [u8; 32 * 4] = [0; 32 * 4];
        let mut src_buffer: [u8; 32 * 4] = [0; 32 * 4];

        std::ptr::copy_nonoverlapping(
            ayuv.get_unchecked(cx * 4..).as_ptr(),
            src_buffer.as_mut_ptr().cast(),
            diff * 4,
        );

        let (a, mut y_vals, u, v) = _mm256_load_deintl_ayuv::<PACKED>(src_buffer.as_slice());

        y_vals = _mm256_subs_epu8(y_vals, y_corr);

        let u_high_u16 = _mm256_srli_epi16::<6>(_mm256_unpackhi_epi8(u, u));
        let v_high_u16 = _mm256_srli_epi16::<6>(_mm256_unpackhi_epi8(v, v));
        let u_low_u16 = _mm256_srli_epi16::<6>(_mm256_unpacklo_epi8(u, u));
        let v_low_u16 = _mm256_srli_epi16::<6>(_mm256_unpacklo_epi8(v, v));

        let y0_10 = _mm256_expand8_unordered_to_10(y_vals);

        let u_high = _mm256_sub_epi16(u_high_u16, uv_corr);
        let v_high = _mm256_sub_epi16(v_high_u16, uv_corr);
        let y_high = _mm256_mulhrs_epi16(y0_10.1, v_luma_coeff);

        let rhc = _mm256_mulhrs_epi16(v_high, v_cr_coeff);
        let bhc = _mm256_mulhrs_epi16(u_high, v_cb_coeff);
        let ghc0 = _mm256_mulhrs_epi16(v_high, v_g_coeff_1);
        let ghc1 = _mm256_mulhrs_epi16(u_high, v_g_coeff_2);

        let r_high = _mm256_add_epi16(y_high, rhc);
        let b_high = _mm256_add_epi16(y_high, bhc);
        let g_high = _mm256_sub_epi16(y_high, _mm256_add_epi16(ghc0, ghc1));

        let u_low = _mm256_sub_epi16(u_low_u16, uv_corr);
        let v_low = _mm256_sub_epi16(v_low_u16, uv_corr);
        let y_low = _mm256_mulhrs_epi16(y0_10.0, v_luma_coeff);

        let rlc = _mm256_mulhrs_epi16(v_low, v_cr_coeff);
        let blc = _mm256_mulhrs_epi16(u_low, v_cb_coeff);
        let glc0 = _mm256_mulhrs_epi16(v_low, v_g_coeff_1);
        let glc1 = _mm256_mulhrs_epi16(u_low, v_g_coeff_2);

        let r_low = _mm256_add_epi16(y_low, rlc);
        let b_low = _mm256_add_epi16(y_low, blc);
        let g_low = _mm256_sub_epi16(y_low, _mm256_add_epi16(glc0, glc1));

        let (r_values, g_values, b_values);

        if use_premultiply {
            let a_high = _mm256_unpackhi_epi8(a, _mm256_setzero_si256());
            let a_low = _mm256_unpacklo_epi8(a, _mm256_setzero_si256());

            let (r_l, r_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(r_low, a_low),
                _mm256_mullo_epi16(r_high, a_high),
            );
            let (g_l, g_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(g_low, a_low),
                _mm256_mullo_epi16(g_high, a_high),
            );
            let (b_l, b_h) = avx2_div_by255_x2(
                _mm256_mullo_epi16(b_low, a_low),
                _mm256_mullo_epi16(b_high, a_high),
            );

            r_values = _mm256_packus_epi16(r_l, r_h);
            g_values = _mm256_packus_epi16(g_l, g_h);
            b_values = _mm256_packus_epi16(b_l, b_h);
        } else {
            r_values = _mm256_packus_epi16(r_low, r_high);
            g_values = _mm256_packus_epi16(g_low, g_high);
            b_values = _mm256_packus_epi16(b_low, b_high);
        }

        _mm256_store_interleave_rgb_for_yuv::<DESTINATION_CHANNELS>(
            dst_buffer.as_mut_ptr().cast(),
            r_values,
            g_values,
            b_values,
            a,
        );

        let dst_shift = cx * channels;
        std::ptr::copy_nonoverlapping(
            dst_buffer.as_mut_ptr().cast(),
            rgba.get_unchecked_mut(dst_shift..).as_mut_ptr(),
            diff * channels,
        );
    }
}