simdeez 3.0.1

SIMD library to abstract over different instruction sets and widths
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

use super::*;

impl_op! {
    fn bit_and<binary> {
        for Avx512(a: __m512, b: __m512) -> __m512 {
            _mm512_and_ps(a, b)
        }
        for Avx2(a: __m256, b: __m256) -> __m256 {
            _mm256_and_ps(a, b)
        }
        for Sse41(a: __m128, b: __m128) -> __m128 {
            _mm_and_ps(a, b)
        }
        for Sse2(a: __m128, b: __m128) -> __m128 {
            _mm_and_ps(a, b)
        }
        for Scalar(a: u64, b: u64) -> u64 {
            a & b
        }
        for Neon(a: int8x16_t, b: int8x16_t) -> int8x16_t {
            vandq_s8(a, b)
        }
        for Wasm(a: v128, b: v128) -> v128 {
            v128_and(a, b)
        }
    }
}

impl_op! {
    fn bit_or<binary> {
        for Avx512(a: __m512, b: __m512) -> __m512 {
            _mm512_or_ps(a, b)
        }
        for Avx2(a: __m256, b: __m256) -> __m256 {
            _mm256_or_ps(a, b)
        }
        for Sse41(a: __m128, b: __m128) -> __m128 {
            _mm_or_ps(a, b)
        }
        for Sse2(a: __m128, b: __m128) -> __m128 {
            _mm_or_ps(a, b)
        }
        for Scalar(a: u64, b: u64) -> u64 {
            a | b
        }
        for Neon(a: int8x16_t, b: int8x16_t) -> int8x16_t {
            vorrq_s8(a, b)
        }
        for Wasm(a: v128, b: v128) -> v128 {
            v128_or(a, b)
        }
    }
}

impl_op! {
    fn bit_xor<binary> {
        for Avx512(a: __m512, b: __m512) -> __m512 {
            _mm512_xor_ps(a, b)
        }
        for Avx2(a: __m256, b: __m256) -> __m256 {
            _mm256_xor_ps(a, b)
        }
        for Sse41(a: __m128, b: __m128) -> __m128 {
            _mm_xor_ps(a, b)
        }
        for Sse2(a: __m128, b: __m128) -> __m128 {
            _mm_xor_ps(a, b)
        }
        for Scalar(a: u64, b: u64) -> u64 {
            a ^ b
        }
        for Neon(a: int8x16_t, b: int8x16_t) -> int8x16_t {
            veorq_s8(a, b)
        }
        for Wasm(a: v128, b: v128) -> v128 {
            v128_xor(a, b)
        }
    }
}

impl_op! {
    fn bit_not<binary> {
        for Avx512(a: __m512) -> __m512 {
            _mm512_xor_ps(a, _mm512_castsi512_ps(_mm512_set1_epi32(-1)))
        }
        for Avx2(a: __m256) -> __m256 {
            let all1 = _mm256_set1_epi32(-1);
            _mm256_castsi256_ps(_mm256_xor_si256(_mm256_castps_si256(a), all1))
        }
        for Sse41(a: __m128) -> __m128 {
            let all1 = _mm_set1_epi32(-1);
            _mm_castsi128_ps(_mm_xor_si128(_mm_castps_si128(a), all1))
        }
        for Sse2(a: __m128) -> __m128 {
            let all1 = _mm_set1_epi32(-1);
            _mm_castsi128_ps(_mm_xor_si128(_mm_castps_si128(a), all1))
        }
        for Scalar(a: u64) -> u64 {
            !a
        }
        for Neon(a: int8x16_t) -> int8x16_t {
            vreinterpretq_s8_u8(vmvnq_u8(vreinterpretq_u8_s8(a)))
        }
        for Wasm(a: v128) -> v128 {
            v128_not(a)
        }
    }
}

impl_op! {
    fn bit_andnot<binary> {
        for Avx512(a: __m512, b: __m512) -> __m512 {
            _mm512_andnot_ps(a, b)
        }
        for Avx2(a: __m256, b: __m256) -> __m256 {
            _mm256_andnot_ps(a, b)
        }
        for Sse41(a: __m128, b: __m128) -> __m128 {
            _mm_andnot_ps(a, b)
        }
        for Sse2(a: __m128, b: __m128) -> __m128 {
            _mm_andnot_ps(a, b)
        }
        for Scalar(a: u64, b: u64) -> u64 {
            !a & b
        }
        for Neon(a: int8x16_t, b: int8x16_t) -> int8x16_t {
            vandq_s8(vreinterpretq_s8_u8(vmvnq_u8(vreinterpretq_u8_s8(a))), b)
        }
        for Wasm(a: v128, b: v128) -> v128 {
            v128_andnot(b, a)
        }
    }
}