1 /* SPDX-License-Identifier: MIT
2 *
3 * Permission is hereby granted, free of charge, to any person
4 * obtaining a copy of this software and associated documentation
5 * files (the "Software"), to deal in the Software without
6 * restriction, including without limitation the rights to use, copy,
7 * modify, merge, publish, distribute, sublicense, and/or sell copies
8 * of the Software, and to permit persons to whom the Software is
9 * furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice shall be
12 * included in all copies or substantial portions of the Software.
13 *
14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
15 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
16 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
17 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
18 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
19 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
20 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 * SOFTWARE.
22 *
23 * Copyright:
24 * 2020 Evan Nemerson <evan@nemerson.com>
25 * 2020 Hidayat Khan <huk2209@gmail.com>
26 */
27
28 #if !defined(SIMDE_X86_AVX512_SRLI_H)
29 #define SIMDE_X86_AVX512_SRLI_H
30
31 #include "types.h"
32 #include "../avx2.h"
33 #include "mov.h"
34 #include "setzero.h"
35
36 HEDLEY_DIAGNOSTIC_PUSH
37 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
38 SIMDE_BEGIN_DECLS_
39
40 SIMDE_FUNCTION_ATTRIBUTES
41 simde__m512i
simde_mm512_srli_epi16(simde__m512i a,const unsigned int imm8)42 simde_mm512_srli_epi16 (simde__m512i a, const unsigned int imm8)
43 SIMDE_REQUIRE_CONSTANT_RANGE(imm8, 0, 255) {
44 #if defined(SIMDE_X86_AVX512BW_NATIVE) && (defined(HEDLEY_GCC_VERSION) && ((__GNUC__ == 5 && __GNUC_MINOR__ == 5) || (__GNUC__ == 6 && __GNUC_MINOR__ >= 4)))
45 simde__m512i r;
46
47 SIMDE_CONSTIFY_16_(_mm512_srli_epi16, r, simde_mm512_setzero_si512(), imm8, a);
48
49 return r;
50 #elif defined(SIMDE_X86_AVX512BW_NATIVE)
51 return SIMDE_BUG_IGNORE_SIGN_CONVERSION(_mm512_srli_epi16(a, imm8));
52 #else
53 simde__m512i_private
54 r_,
55 a_ = simde__m512i_to_private(a);
56
57 if (HEDLEY_STATIC_CAST(unsigned int, imm8) > 15)
58 return simde_mm512_setzero_si512();
59
60 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
61 r_.u16 = a_.u16 >> SIMDE_CAST_VECTOR_SHIFT_COUNT(16, imm8);
62 #else
63 SIMDE_VECTORIZE
64 for (size_t i = 0 ; i < (sizeof(r_.u16) / sizeof(r_.u16[0])) ; i++) {
65 r_.u16[i] = a_.u16[i] >> imm8;
66 }
67 #endif
68
69 return simde__m512i_from_private(r_);
70 #endif
71 }
72 #if defined(SIMDE_X86_AVX512BW_NATIVE)
73 #define simde_mm512_srli_epi16(a, imm8) _mm512_srli_epi16(a, imm8)
74 #endif
75 #if defined(SIMDE_X86_AVX512BW_ENABLE_NATIVE_ALIASES)
76 #undef _mm512_srli_epi16
77 #define _mm512_srli_epi16(a, imm8) simde_mm512_srli_epi16(a, imm8)
78 #endif
79
80 SIMDE_FUNCTION_ATTRIBUTES
81 simde__m512i
simde_mm512_srli_epi32(simde__m512i a,unsigned int imm8)82 simde_mm512_srli_epi32 (simde__m512i a, unsigned int imm8) {
83 #if defined(SIMDE_X86_AVX512F_NATIVE) && (defined(HEDLEY_GCC_VERSION) && ((__GNUC__ == 5 && __GNUC_MINOR__ == 5) || (__GNUC__ == 6 && __GNUC_MINOR__ >= 4)))
84 simde__m512i r;
85
86 SIMDE_CONSTIFY_32_(_mm512_srli_epi32, r, simde_mm512_setzero_si512(), imm8, a);
87
88 return r;
89 #elif defined(SIMDE_X86_AVX512F_NATIVE)
90 return SIMDE_BUG_IGNORE_SIGN_CONVERSION(_mm512_srli_epi32(a, imm8));
91 #else
92 simde__m512i_private
93 r_,
94 a_ = simde__m512i_to_private(a);
95
96 #if defined(SIMDE_X86_AVX2_NATIVE)
97 r_.m256i[0] = simde_mm256_srli_epi32(a_.m256i[0], HEDLEY_STATIC_CAST(int, imm8));
98 r_.m256i[1] = simde_mm256_srli_epi32(a_.m256i[1], HEDLEY_STATIC_CAST(int, imm8));
99 #elif defined(SIMDE_X86_SSE2_NATIVE)
100 r_.m128i[0] = simde_mm_srli_epi32(a_.m128i[0], HEDLEY_STATIC_CAST(int, imm8));
101 r_.m128i[1] = simde_mm_srli_epi32(a_.m128i[1], HEDLEY_STATIC_CAST(int, imm8));
102 r_.m128i[2] = simde_mm_srli_epi32(a_.m128i[2], HEDLEY_STATIC_CAST(int, imm8));
103 r_.m128i[3] = simde_mm_srli_epi32(a_.m128i[3], HEDLEY_STATIC_CAST(int, imm8));
104 #else
105 if (imm8 > 31) {
106 simde_memset(&r_, 0, sizeof(r_));
107 } else {
108 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
109 r_.u32 = a_.u32 >> imm8;
110 #else
111 SIMDE_VECTORIZE
112 for (size_t i = 0 ; i < (sizeof(r_.u32) / sizeof(r_.u32[0])) ; i++) {
113 r_.u32[i] = a_.u32[i] >> imm8;
114 }
115 #endif
116 }
117 #endif
118
119 return simde__m512i_from_private(r_);
120 #endif
121 }
122 #if defined(SIMDE_X86_AVX512F_ENABLE_NATIVE_ALIASES)
123 #undef _mm512_srli_epi32
124 #define _mm512_srli_epi32(a, imm8) simde_mm512_srli_epi32(a, imm8)
125 #endif
126
127 SIMDE_FUNCTION_ATTRIBUTES
128 simde__m512i
simde_mm512_srli_epi64(simde__m512i a,unsigned int imm8)129 simde_mm512_srli_epi64 (simde__m512i a, unsigned int imm8) {
130 #if defined(SIMDE_X86_AVX512F_NATIVE) && (defined(HEDLEY_GCC_VERSION) && ((__GNUC__ == 5 && __GNUC_MINOR__ == 5) || (__GNUC__ == 6 && __GNUC_MINOR__ >= 4)))
131 simde__m512i r;
132
133 SIMDE_CONSTIFY_64_(_mm512_srli_epi64, r, simde_mm512_setzero_si512(), imm8, a);
134
135 return r;
136 #elif defined(SIMDE_X86_AVX512F_NATIVE)
137 return SIMDE_BUG_IGNORE_SIGN_CONVERSION(_mm512_srli_epi64(a, imm8));
138 #else
139 simde__m512i_private
140 r_,
141 a_ = simde__m512i_to_private(a);
142
143 #if defined(SIMDE_X86_AVX2_NATIVE)
144 r_.m256i[0] = simde_mm256_srli_epi64(a_.m256i[0], HEDLEY_STATIC_CAST(int, imm8));
145 r_.m256i[1] = simde_mm256_srli_epi64(a_.m256i[1], HEDLEY_STATIC_CAST(int, imm8));
146 #elif defined(SIMDE_X86_SSE2_NATIVE)
147 r_.m128i[0] = simde_mm_srli_epi64(a_.m128i[0], HEDLEY_STATIC_CAST(int, imm8));
148 r_.m128i[1] = simde_mm_srli_epi64(a_.m128i[1], HEDLEY_STATIC_CAST(int, imm8));
149 r_.m128i[2] = simde_mm_srli_epi64(a_.m128i[2], HEDLEY_STATIC_CAST(int, imm8));
150 r_.m128i[3] = simde_mm_srli_epi64(a_.m128i[3], HEDLEY_STATIC_CAST(int, imm8));
151 #else
152 /* The Intel Intrinsics Guide says that only the 8 LSBits of imm8 are
153 * used. In this case we should do "imm8 &= 0xff" here. However in
154 * practice all bits are used. */
155 if (imm8 > 63) {
156 simde_memset(&r_, 0, sizeof(r_));
157 } else {
158 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
159 r_.u64 = a_.u64 >> imm8;
160 #else
161 SIMDE_VECTORIZE
162 for (size_t i = 0 ; i < (sizeof(r_.u64) / sizeof(r_.u64[0])) ; i++) {
163 r_.u64[i] = a_.u64[i] >> imm8;
164 }
165 #endif
166 }
167 #endif
168
169 return simde__m512i_from_private(r_);
170 #endif
171 }
172 #if defined(SIMDE_X86_AVX512F_ENABLE_NATIVE_ALIASES)
173 #undef _mm512_srli_epi64
174 #define _mm512_srli_epi64(a, imm8) simde_mm512_srli_epi64(a, imm8)
175 #endif
176
177 SIMDE_END_DECLS_
178 HEDLEY_DIAGNOSTIC_POP
179
180 #endif /* !defined(SIMDE_X86_AVX512_SRLI_H) */
181