| 1 |
/* PR target/79299 */ |
| 2 |
/* { dg-do assemble { target avx512vl } } */ |
| 3 |
/* { dg-require-effective-target masm_intel } */ |
| 4 |
/* { dg-options "-Ofast -mavx512vl -masm=intel" } */ |
| 5 |
|
| 6 |
#include <immintrin.h> |
| 7 |
|
| 8 |
__m512 |
| 9 |
f1 (__m512i x, void const *y) |
| 10 |
{ |
| 11 |
return _mm512_i32gather_ps (x, y, 1); |
| 12 |
} |
| 13 |
|
| 14 |
__m512 |
| 15 |
f2 (__m512 x, __mmask16 y, __m512i z, void const *w) |
| 16 |
{ |
| 17 |
return _mm512_mask_i32gather_ps (x, y, z, w, 1); |
| 18 |
} |
| 19 |
|
| 20 |
__m512d |
| 21 |
f3 (__m256i x, void const *y) |
| 22 |
{ |
| 23 |
return _mm512_i32gather_pd (x, y, 1); |
| 24 |
} |
| 25 |
|
| 26 |
__m512d |
| 27 |
f4 (__m512d x, __mmask8 y, __m256i z, void const *w) |
| 28 |
{ |
| 29 |
return _mm512_mask_i32gather_pd (x, y, z, w, 1); |
| 30 |
} |
| 31 |
|
| 32 |
__m256 |
| 33 |
f5 (__m512i x, void const *y) |
| 34 |
{ |
| 35 |
return _mm512_i64gather_ps (x, y, 1); |
| 36 |
} |
| 37 |
|
| 38 |
__m256 |
| 39 |
f6 (__m256 x, __mmask16 y, __m512i z, void const *w) |
| 40 |
{ |
| 41 |
return _mm512_mask_i64gather_ps (x, y, z, w, 1); |
| 42 |
} |
| 43 |
|
| 44 |
__m512d |
| 45 |
f7 (__m512i x, void const *y) |
| 46 |
{ |
| 47 |
return _mm512_i64gather_pd (x, y, 1); |
| 48 |
} |
| 49 |
|
| 50 |
__m512d |
| 51 |
f8 (__m512d x, __mmask8 y, __m512i z, void const *w) |
| 52 |
{ |
| 53 |
return _mm512_mask_i64gather_pd (x, y, z, w, 1); |
| 54 |
} |
| 55 |
|
| 56 |
__m512i |
| 57 |
f9 (__m512i x, void const *y) |
| 58 |
{ |
| 59 |
return _mm512_i32gather_epi32 (x, y, 1); |
| 60 |
} |
| 61 |
|
| 62 |
__m512i |
| 63 |
f10 (__m512i x, __mmask16 y, __m512i z, void const *w) |
| 64 |
{ |
| 65 |
return _mm512_mask_i32gather_epi32 (x, y, z, w, 1); |
| 66 |
} |
| 67 |
|
| 68 |
__m512i |
| 69 |
f11 (__m256i x, void const *y) |
| 70 |
{ |
| 71 |
return _mm512_i32gather_epi64 (x, y, 1); |
| 72 |
} |
| 73 |
|
| 74 |
__m512i |
| 75 |
f12 (__m512i x, __mmask8 y, __m256i z, void const *w) |
| 76 |
{ |
| 77 |
return _mm512_mask_i32gather_epi64 (x, y, z, w, 1); |
| 78 |
} |
| 79 |
|
| 80 |
__m256i |
| 81 |
f13 (__m512i x, void const *y) |
| 82 |
{ |
| 83 |
return _mm512_i64gather_epi32 (x, y, 1); |
| 84 |
} |
| 85 |
|
| 86 |
__m256i |
| 87 |
f14 (__m256i x, __mmask16 y, __m512i z, void const *w) |
| 88 |
{ |
| 89 |
return _mm512_mask_i64gather_epi32 (x, y, z, w, 1); |
| 90 |
} |
| 91 |
|
| 92 |
__m512i |
| 93 |
f15 (__m512i x, void const *y) |
| 94 |
{ |
| 95 |
return _mm512_i64gather_epi64 (x, y, 1); |
| 96 |
} |
| 97 |
|
| 98 |
__m512i |
| 99 |
f16 (__m512i x, __mmask8 y, __m512i z, void const *w) |
| 100 |
{ |
| 101 |
return _mm512_mask_i64gather_epi64 (x, y, z, w, 1); |
| 102 |
} |
| 103 |
|
| 104 |
__m256 |
| 105 |
f17 (__m256 x, __mmask8 y, __m256i z, void const *w) |
| 106 |
{ |
| 107 |
return _mm256_mmask_i32gather_ps (x, y, z, w, 1); |
| 108 |
} |
| 109 |
|
| 110 |
__m128 |
| 111 |
f18 (__m128 x, __mmask8 y, __m128i z, void const *w) |
| 112 |
{ |
| 113 |
return _mm_mmask_i32gather_ps (x, y, z, w, 1); |
| 114 |
} |
| 115 |
|
| 116 |
__m256d |
| 117 |
f19 (__m256d x, __mmask8 y, __m128i z, void const *w) |
| 118 |
{ |
| 119 |
return _mm256_mmask_i32gather_pd (x, y, z, w, 1); |
| 120 |
} |
| 121 |
|
| 122 |
__m128d |
| 123 |
f20 (__m128d x, __mmask8 y, __m128i z, void const *w) |
| 124 |
{ |
| 125 |
return _mm_mmask_i32gather_pd (x, y, z, w, 1); |
| 126 |
} |
| 127 |
|
| 128 |
__m128 |
| 129 |
f21 (__m128 x, __mmask8 y, __m256i z, void const *w) |
| 130 |
{ |
| 131 |
return _mm256_mmask_i64gather_ps (x, y, z, w, 1); |
| 132 |
} |
| 133 |
|
| 134 |
__m128 |
| 135 |
f22 (__m128 x, __mmask8 y, __m128i z, void const *w) |
| 136 |
{ |
| 137 |
return _mm_mmask_i64gather_ps (x, y, z, w, 1); |
| 138 |
} |
| 139 |
|
| 140 |
__m256d |
| 141 |
f23 (__m256d x, __mmask8 y, __m256i z, void const *w) |
| 142 |
{ |
| 143 |
return _mm256_mmask_i64gather_pd (x, y, z, w, 1); |
| 144 |
} |
| 145 |
|
| 146 |
__m128d |
| 147 |
f24 (__m128d x, __mmask8 y, __m128i z, void const *w) |
| 148 |
{ |
| 149 |
return _mm_mmask_i64gather_pd (x, y, z, w, 1); |
| 150 |
} |
| 151 |
|
| 152 |
__m256i |
| 153 |
f25 (__m256i x, __mmask8 y, __m256i z, void const *w) |
| 154 |
{ |
| 155 |
return _mm256_mmask_i32gather_epi32 (x, y, z, w, 1); |
| 156 |
} |
| 157 |
|
| 158 |
__m128i |
| 159 |
f26 (__m128i x, __mmask8 y, __m128i z, void const *w) |
| 160 |
{ |
| 161 |
return _mm_mmask_i32gather_epi32 (x, y, z, w, 1); |
| 162 |
} |
| 163 |
|
| 164 |
__m256i |
| 165 |
f27 (__m256i x, __mmask8 y, __m128i z, void const *w) |
| 166 |
{ |
| 167 |
return _mm256_mmask_i32gather_epi64 (x, y, z, w, 1); |
| 168 |
} |
| 169 |
|
| 170 |
__m128i |
| 171 |
f28 (__m128i x, __mmask8 y, __m128i z, void const *w) |
| 172 |
{ |
| 173 |
return _mm_mmask_i32gather_epi64 (x, y, z, w, 1); |
| 174 |
} |
| 175 |
|
| 176 |
__m128i |
| 177 |
f29 (__m128i x, __mmask8 y, __m256i z, void const *w) |
| 178 |
{ |
| 179 |
return _mm256_mmask_i64gather_epi32 (x, y, z, w, 1); |
| 180 |
} |
| 181 |
|
| 182 |
__m128i |
| 183 |
f30 (__m128i x, __mmask8 y, __m128i z, void const *w) |
| 184 |
{ |
| 185 |
return _mm_mmask_i64gather_epi32 (x, y, z, w, 1); |
| 186 |
} |
| 187 |
|
| 188 |
__m256i |
| 189 |
f31 (__m256i x, __mmask8 y, __m256i z, void const *w) |
| 190 |
{ |
| 191 |
return _mm256_mmask_i64gather_epi64 (x, y, z, w, 1); |
| 192 |
} |
| 193 |
|
| 194 |
__m128i |
| 195 |
f32 (__m128i x, __mmask8 y, __m128i z, void const *w) |
| 196 |
{ |
| 197 |
return _mm_mmask_i64gather_epi64 (x, y, z, w, 1); |
| 198 |
} |
| 199 |
|
| 200 |
__m256 |
| 201 |
f33 (__m256 x, __m256i z, void const *w) |
| 202 |
{ |
| 203 |
return _mm256_mmask_i32gather_ps (x, -1, z, w, 1); |
| 204 |
} |
| 205 |
|
| 206 |
__m128 |
| 207 |
f34 (__m128 x, __m128i z, void const *w) |
| 208 |
{ |
| 209 |
return _mm_mmask_i32gather_ps (x, -1, z, w, 1); |
| 210 |
} |
| 211 |
|
| 212 |
__m256d |
| 213 |
f35 (__m256d x, __m128i z, void const *w) |
| 214 |
{ |
| 215 |
return _mm256_mmask_i32gather_pd (x, -1, z, w, 1); |
| 216 |
} |
| 217 |
|
| 218 |
__m128d |
| 219 |
f36 (__m128d x, __m128i z, void const *w) |
| 220 |
{ |
| 221 |
return _mm_mmask_i32gather_pd (x, -1, z, w, 1); |
| 222 |
} |
| 223 |
|
| 224 |
__m128 |
| 225 |
f37 (__m128 x, __m256i z, void const *w) |
| 226 |
{ |
| 227 |
return _mm256_mmask_i64gather_ps (x, -1, z, w, 1); |
| 228 |
} |
| 229 |
|
| 230 |
__m128 |
| 231 |
f38 (__m128 x, __m128i z, void const *w) |
| 232 |
{ |
| 233 |
return _mm_mmask_i64gather_ps (x, -1, z, w, 1); |
| 234 |
} |
| 235 |
|
| 236 |
__m256d |
| 237 |
f39 (__m256d x, __m256i z, void const *w) |
| 238 |
{ |
| 239 |
return _mm256_mmask_i64gather_pd (x, -1, z, w, 1); |
| 240 |
} |
| 241 |
|
| 242 |
__m128d |
| 243 |
f40 (__m128d x, __m128i z, void const *w) |
| 244 |
{ |
| 245 |
return _mm_mmask_i64gather_pd (x, -1, z, w, 1); |
| 246 |
} |
| 247 |
|
| 248 |
__m256i |
| 249 |
f41 (__m256i x, __m256i z, void const *w) |
| 250 |
{ |
| 251 |
return _mm256_mmask_i32gather_epi32 (x, -1, z, w, 1); |
| 252 |
} |
| 253 |
|
| 254 |
__m128i |
| 255 |
f42 (__m128i x, __m128i z, void const *w) |
| 256 |
{ |
| 257 |
return _mm_mmask_i32gather_epi32 (x, -1, z, w, 1); |
| 258 |
} |
| 259 |
|
| 260 |
__m256i |
| 261 |
f43 (__m256i x, __m128i z, void const *w) |
| 262 |
{ |
| 263 |
return _mm256_mmask_i32gather_epi64 (x, -1, z, w, 1); |
| 264 |
} |
| 265 |
|
| 266 |
__m128i |
| 267 |
f44 (__m128i x, __m128i z, void const *w) |
| 268 |
{ |
| 269 |
return _mm_mmask_i32gather_epi64 (x, -1, z, w, 1); |
| 270 |
} |
| 271 |
|
| 272 |
__m128i |
| 273 |
f45 (__m128i x, __m256i z, void const *w) |
| 274 |
{ |
| 275 |
return _mm256_mmask_i64gather_epi32 (x, -1, z, w, 1); |
| 276 |
} |
| 277 |
|
| 278 |
__m128i |
| 279 |
f46 (__m128i x, __m128i z, void const *w) |
| 280 |
{ |
| 281 |
return _mm_mmask_i64gather_epi32 (x, -1, z, w, 1); |
| 282 |
} |
| 283 |
|
| 284 |
__m256i |
| 285 |
f47 (__m256i x, __m256i z, void const *w) |
| 286 |
{ |
| 287 |
return _mm256_mmask_i64gather_epi64 (x, -1, z, w, 1); |
| 288 |
} |
| 289 |
|
| 290 |
__m128i |
| 291 |
f48 (__m128i x, __m128i z, void const *w) |
| 292 |
{ |
| 293 |
return _mm_mmask_i64gather_epi64 (x, -1, z, w, 1); |
| 294 |
} |