x86intrin 0.4.5

x86intrin implementation
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
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316

use super::*;
use super::simd_shuffle16;

extern "platform-intrinsic" {
    fn x86_mm_abs_epi8(x: i8x16) -> i8x16;
    fn x86_mm_abs_epi16(x: i16x8) -> i16x8;
    fn x86_mm_abs_epi32(x: i32x4) -> i32x4;
    fn x86_mm_hadd_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_hadd_epi32(x: i32x4, y: i32x4) -> i32x4;
    fn x86_mm_hadds_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_hsub_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_hsub_epi32(x: i32x4, y: i32x4) -> i32x4;
    fn x86_mm_hsubs_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_maddubs_epi16(x: u8x16, y: i8x16) -> i16x8;
    fn x86_mm_mulhrs_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_shuffle_epi8(x: i8x16, y: i8x16) -> i8x16;
    fn x86_mm_sign_epi8(x: i8x16, y: i8x16) -> i8x16;
    fn x86_mm_sign_epi16(x: i16x8, y: i16x8) -> i16x8;
    fn x86_mm_sign_epi32(x: i32x4, y: i32x4) -> i32x4;
}

// pabsw
// __m128i _mm_abs_epi16 (__m128i a)
#[inline]
pub fn mm_abs_epi16(a: m128i) -> m128i {
    unsafe { x86_mm_abs_epi16(a.as_i16x8()).as_m128i() }
}

// pabsd
// __m128i _mm_abs_epi32 (__m128i a)
#[inline]
pub fn mm_abs_epi32(a: m128i) -> m128i {
    unsafe { x86_mm_abs_epi32(a.as_i32x4()).as_m128i() }
}

// pabsb
// __m128i _mm_abs_epi8 (__m128i a)
#[inline]
pub fn mm_abs_epi8(a: m128i) -> m128i {
    unsafe { x86_mm_abs_epi8(a.as_i8x16()).as_m128i() }
}

// palignr
// __m128i _mm_alignr_epi8 (__m128i a, __m128i b, int count)
#[inline]
pub fn mm_alignr_epi8(a: m128i, b: m128i, count: i32) -> m128i {
    let ai = a.as_i8x16();
    let bi = b.as_i8x16();
    let zi = mm_setzero_si128().as_i8x16();
    unsafe {
        let c: i8x16 = match count {
            0 => simd_shuffle16(ai, bi, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]),
            1 => simd_shuffle16(ai, bi, [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]),
            2 => simd_shuffle16(ai, bi, [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]),
            3 => simd_shuffle16(ai, bi, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]),
            4 => simd_shuffle16(ai, bi, [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]),
            5 => simd_shuffle16(ai, bi, [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]),
            6 => simd_shuffle16(ai, bi, [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]),
            7 => simd_shuffle16(ai, bi, [7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22]),
            8 => simd_shuffle16(ai, bi, [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]),
            9 => simd_shuffle16(ai, bi, [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24]),
            10 => simd_shuffle16(ai, bi, [10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]),
            11 => simd_shuffle16(ai, bi, [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]),
            12 => simd_shuffle16(ai, bi, [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]),
            13 => simd_shuffle16(ai, bi, [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]),
            14 => simd_shuffle16(ai, bi, [14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]),
            15 => simd_shuffle16(ai, bi, [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]),
            16 => simd_shuffle16(bi, zi, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]),
            17 => simd_shuffle16(bi, zi, [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]),
            18 => simd_shuffle16(bi, zi, [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]),
            19 => simd_shuffle16(bi, zi, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]),
            20 => simd_shuffle16(bi, zi, [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]),
            21 => simd_shuffle16(bi, zi, [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]),
            22 => simd_shuffle16(bi, zi, [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]),
            23 => simd_shuffle16(bi, zi, [7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22]),
            24 => simd_shuffle16(bi, zi, [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]),
            25 => simd_shuffle16(bi, zi, [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24]),
            26 => simd_shuffle16(bi, zi, [10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]),
            27 => simd_shuffle16(bi, zi, [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]),
            28 => simd_shuffle16(bi, zi, [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]),
            29 => simd_shuffle16(bi, zi, [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]),
            30 => simd_shuffle16(bi, zi, [14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]),
            31 => simd_shuffle16(bi, zi, [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]),
            _ => zi,
        };
        c.as_m128i()
    }
}

// phaddw
// __m128i _mm_hadd_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_hadd_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hadd_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// phaddd
// __m128i _mm_hadd_epi32 (__m128i a, __m128i b)
#[inline]
pub fn mm_hadd_epi32(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hadd_epi32(a.as_i32x4(), b.as_i32x4()).as_m128i() }
}

// phaddsw
// __m128i _mm_hadds_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_hadds_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hadds_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// phsubw
// __m128i _mm_hsub_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_hsub_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hsub_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// phsubd
// __m128i _mm_hsub_epi32 (__m128i a, __m128i b)
#[inline]
pub fn mm_hsub_epi32(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hsub_epi32(a.as_i32x4(), b.as_i32x4()).as_m128i() }
}

// phsubsw
// __m128i _mm_hsubs_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_hsubs_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_hsubs_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// pmaddubsw
// __m128i _mm_maddubs_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_maddubs_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_maddubs_epi16(a.as_u8x16(), b.as_i8x16()).as_m128i() }
}

// pmulhrsw
// __m128i _mm_mulhrs_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_mulhrs_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_mulhrs_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// pshufb
// __m128i _mm_shuffle_epi8 (__m128i a, __m128i b)
#[inline]
pub fn mm_shuffle_epi8(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_shuffle_epi8(a.as_i8x16(), b.as_i8x16()).as_m128i() }
}

// psignw
// __m128i _mm_sign_epi16 (__m128i a, __m128i b)
#[inline]
pub fn mm_sign_epi16(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_sign_epi16(a.as_i16x8(), b.as_i16x8()).as_m128i() }
}

// psignd
// __m128i _mm_sign_epi32 (__m128i a, __m128i b)
#[inline]
pub fn mm_sign_epi32(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_sign_epi32(a.as_i32x4(), b.as_i32x4()).as_m128i() }
}

// psignb
// __m128i _mm_sign_epi8 (__m128i a, __m128i b)
#[inline]
pub fn mm_sign_epi8(a: m128i, b: m128i) -> m128i {
    unsafe { x86_mm_sign_epi8(a.as_i8x16(), b.as_i8x16()).as_m128i() }
}

// pabsw
// __m64 _mm_abs_pi16 (__m64 a)
// pabsd
// __m64 _mm_abs_pi32 (__m64 a)
// pabsb
// __m64 _mm_abs_pi8 (__m64 a)
// palignr
// __m64 _mm_alignr_pi8 (__m64 a, __m64 b, int count)
// phaddw
// __m64 _mm_hadd_pi16 (__m64 a, __m64 b)
// phaddw
// __m64 _mm_hadd_pi32 (__m64 a, __m64 b)
// phaddsw
// __m64 _mm_hadds_pi16 (__m64 a, __m64 b)
// phsubw
// __m64 _mm_hsub_pi16 (__m64 a, __m64 b)
// phsubd
// __m64 _mm_hsub_pi32 (__m64 a, __m64 b)
// phsubsw
// __m64 _mm_hsubs_pi16 (__m64 a, __m64 b)
// pmaddubsw
// __m64 _mm_maddubs_pi16 (__m64 a, __m64 b)
// pmulhrsw
// __m64 _mm_mulhrs_pi16 (__m64 a, __m64 b)
// pshufb
// __m64 _mm_shuffle_pi8 (__m64 a, __m64 b)
// psignw
// __m64 _mm_sign_pi16 (__m64 a, __m64 b)
// psignd
// __m64 _mm_sign_pi32 (__m64 a, __m64 b)
// psignb
// __m64 _mm_sign_pi8 (__m64 a, __m64 b)

#[cfg(test)]
mod tests {
    use super::super::*;

    #[test]
    fn test_abs() {
        let x8 = mm_setr_epi8(1, 2, 3, 4, 5, 6, 7, 8, -1, -2, -3, -4, -5, -6, -7, -0x80);
        let x16 = mm_setr_epi16(1, 2, 3, 4, -1, -2, -3, -0x8000);
        let x32 = mm_setr_epi32(1, 2, -1, -0x80000000);

        assert_eq!(mm_abs_epi8(x8).as_i8x16().as_array(), [1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, -0x80]);
        assert_eq!(mm_abs_epi16(x16).as_i16x8().as_array(), [1, 2, 3, 4, 1, 2, 3, -0x8000]);
        assert_eq!(mm_abs_epi32(x32).as_i32x4().as_array(), [1, 2, 1, -0x80000000]);
    }

    #[test]
    fn test_hadd() {
        let x16 = mm_setr_epi16(1, 2, 0x7000, 0x7000, -1, -2, -0x7000, -0x7000);
        let x32 = mm_setr_epi32(1, 2, -1, -2);

        // 0x7000 + 0x7000 = 0xE000
        let e = 0xE000u16 as i16;
        assert_eq!(mm_hadd_epi16(x16, x16).as_i16x8().as_array(), [3, e, -3, -e, 3, e, -3, -e]);
        assert_eq!(mm_hadd_epi32(x32, x32).as_i32x4().as_array(), [3, -3, 3, -3]);
        assert_eq!(mm_hadds_epi16(x16, x16).as_i16x8().as_array(), [3, 0x7FFF, -3, -0x8000, 3, 0x7FFF, -3, -0x8000]);
    }

    #[test]
    fn test_hsub() {
        let x16 = mm_setr_epi16(1, 2, 0x7000, -0x7000, -1, -2, -0x7000, 0x7000);
        let x32 = mm_setr_epi32(1, 2, -1, -2);

        // 0x7000 + 0x7000 = 0xE000
        let e = 0xE000u16 as i16;
        assert_eq!(mm_hsub_epi16(x16, x16).as_i16x8().as_array(), [-1, e, 1, -e, -1, e, 1, -e]);
        assert_eq!(mm_hsub_epi32(x32, x32).as_i32x4().as_array(), [-1, 1, -1, 1]);
        assert_eq!(mm_hsubs_epi16(x16, x16).as_i16x8().as_array(), [-1, 0x7FFF, 1, -0x8000, -1, 0x7FFF, 1, -0x8000]);
    }

    #[test]
    fn test_maddubs() {
        let xi8 = mm_setr_epi8(1, -2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 127, 127);
        let xu8 = mm_setr_epi8(1, -2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, -1, -1);

        assert_eq!(mm_maddubs_epi16(xi8, xu8).as_i16x8().as_array(),
                   [1 * 1 + -2 * 254,
                    3 * 3 + 4 * 4,
                    5 * 5 + 6 * 6,
                    7 * 7 + 8 * 8,
                    9 * 9 + 10 * 10,
                    11 * 11 + 12 * 12,
                    13 * 13 + 14 * 14,
                    -254]);
    }

    #[test]
    fn test_mulhrs() {
        let x16 = mm_setr_epi16(-0x5CEE, 0x0105, 0x3DA9, -0x7FFF, 0x7FFF, 0x1111, -0x219D, -0x1DBC);
        let y16 = mm_setr_epi16(0x4000, -0x510A, 0x209D, -0x7FFF, 0x0000, 0x2222, 0x1027, 0x7AEF);

        assert_eq!(mm_mulhrs_epi16(x16, y16).as_i16x8().as_array(), [-11895, -165, 4022, 32766, 0, 1165, -1086, -7311]);
    }

    #[test]
    fn test_shuffle_epi8() {
        let x8 = mm_setr_epi8(51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66);
        let idx = mm_setr_epi8(4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11);

        assert_eq!(mm_shuffle_epi8(x8, idx).as_i8x16().as_array(),
                   [55, 56, 57, 58, 51, 52, 53, 54, 63, 64, 65, 66, 59, 60, 61, 62]);
    }

    #[test]
    fn test_sign() {
        let x8 = mm_setr_epi8(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
        let x16 = mm_setr_epi16(1, 2, 3, 4, 5, 6, 7, 8);
        let x32 = mm_setr_epi32(1, 2, 3, 4);

        let idx8 = mm_setr_epi8(0, 1, 2, 3, -1, -2, -3, -4, 0, 1, 2, 3, -1, -2, -3, -4);
        let idx16 = mm_setr_epi16(0, 1, -1, 0, 1, -1, 0, 1);
        let idx32 = mm_setr_epi32(0, 1, -1, 0);

        assert_eq!(mm_sign_epi8(x8, idx8).as_i8x16().as_array(),
                   [0, 2, 3, 4, -5, -6, -7, -8, 0, 10, 11, 12, -13, -14, -15, -16]);
        assert_eq!(mm_sign_epi16(x16, idx16).as_i16x8().as_array(),
                   [0, 2, -3, 0, 5, -6, 0, 8]);
        assert_eq!(mm_sign_epi32(x32, idx32).as_i32x4().as_array(),
                   [0, 2, -3, 0]);
    }

    #[test]
    fn test_palignr() {
        let a = mm_setr_epi8(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
        let b = mm_setr_epi8(17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32);

        assert_eq!(mm_alignr_epi8(a, b, 1).as_i8x16().as_array(), [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]);
        assert_eq!(mm_alignr_epi8(a, b, 31).as_i8x16().as_array(), [32, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]);

        for i in 0..33 {
            let c = mm_alignr_epi8(a, b, i).as_i8x16().as_array();
            for j in 0..16 {
                if i + j + 1 > 32 {
                    assert_eq!(c[j as usize], 0);
                } else {
                    assert_eq!(c[j as usize], (i + j + 1) as i8);
                }
            }
        }
    }
}