24 #error "Never use <avx512fintrin.h> directly; include <immintrin.h> instead."
27 #ifndef __AVX512FINTRIN_H
28 #define __AVX512FINTRIN_H
34 typedef long long __v8di
__attribute__((__vector_size__(64)));
38 typedef unsigned char __v64qu
__attribute__((__vector_size__(64)));
39 typedef unsigned short __v32hu
__attribute__((__vector_size__(64)));
40 typedef unsigned long long __v8du
__attribute__((__vector_size__(64)));
41 typedef unsigned int __v16su
__attribute__((__vector_size__(64)));
45 typedef long long __m512i
__attribute__((__vector_size__(64)));
51 #define _MM_FROUND_TO_NEAREST_INT 0x00
52 #define _MM_FROUND_TO_NEG_INF 0x01
53 #define _MM_FROUND_TO_POS_INF 0x02
54 #define _MM_FROUND_TO_ZERO 0x03
55 #define _MM_FROUND_CUR_DIRECTION 0x04
65 #define _MM_CMPINT_GE _MM_CMPINT_NLT
67 #define _MM_CMPINT_GT _MM_CMPINT_NLE
176 #define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("avx512f")))
183 return (__m512i)(__v8di){ 0, 0, 0, 0, 0, 0, 0, 0 };
186 #define _mm512_setzero_epi32 _mm512_setzero_si512
191 return (__m512d)__builtin_ia32_undef512();
197 return (__m512)__builtin_ia32_undef512();
203 return (__m512)__builtin_ia32_undef512();
209 return (__m512i)__builtin_ia32_undef512();
215 return (__m512i)__builtin_shufflevector((__v4si) __A,
217 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0);
223 return (__m512i)__builtin_ia32_selectd_512(__M,
231 return (__m512i)__builtin_ia32_selectd_512(__M,
239 return (__m512i)__builtin_shufflevector((__v2di) __A,
241 0, 0, 0, 0, 0, 0, 0, 0);
247 return (__m512i)__builtin_ia32_selectq_512(__M,
256 return (__m512i)__builtin_ia32_selectq_512(__M,
264 return (__m512i) __builtin_ia32_pbroadcastd512_gpr_mask (__A,
274 return (__m512i) __builtin_ia32_pbroadcastq512_gpr_mask (__A,
279 return (__m512i) __builtin_ia32_pbroadcastq512_mem_mask (__A,
289 return (__m512){ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
290 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 };
293 #define _mm512_setzero _mm512_setzero_ps
298 return (__m512d){ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 };
304 return (__m512){ __w, __w, __w, __w, __w, __w, __w, __w,
305 __w, __w, __w, __w, __w, __w, __w, __w };
311 return (__m512d){ __w, __w, __w, __w, __w, __w, __w, __w };
317 return (__m512i)(__v64qi){ __w, __w, __w, __w, __w, __w, __w, __w,
318 __w, __w, __w, __w, __w, __w, __w, __w,
319 __w, __w, __w, __w, __w, __w, __w, __w,
320 __w, __w, __w, __w, __w, __w, __w, __w,
321 __w, __w, __w, __w, __w, __w, __w, __w,
322 __w, __w, __w, __w, __w, __w, __w, __w,
323 __w, __w, __w, __w, __w, __w, __w, __w,
324 __w, __w, __w, __w, __w, __w, __w, __w };
330 return (__m512i)(__v32hi){ __w, __w, __w, __w, __w, __w, __w, __w,
331 __w, __w, __w, __w, __w, __w, __w, __w,
332 __w, __w, __w, __w, __w, __w, __w, __w,
333 __w, __w, __w, __w, __w, __w, __w, __w };
339 return (__m512i)(__v16si){ __s, __s, __s, __s, __s, __s, __s, __s,
340 __s, __s, __s, __s, __s, __s, __s, __s };
346 return (__m512i)(__v8di){ __d, __d, __d, __d, __d, __d, __d, __d };
352 return (__m512)__builtin_shufflevector((__v4sf) __A,
354 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0);
360 return (__m512i)(__v16si)
361 { __D, __C, __B, __A, __D, __C, __B, __A,
362 __D, __C, __B, __A, __D, __C, __B, __A };
369 return (__m512i) (__v8di)
370 { __D, __C, __B, __A, __D, __C, __B, __A };
377 { __D, __C, __B, __A, __D, __C, __B, __A };
384 { __D, __C, __B, __A, __D, __C, __B, __A,
385 __D, __C, __B, __A, __D, __C, __B, __A };
388 #define _mm512_setr4_epi32(e0,e1,e2,e3) \
389 _mm512_set4_epi32((e3),(e2),(e1),(e0))
391 #define _mm512_setr4_epi64(e0,e1,e2,e3) \
392 _mm512_set4_epi64((e3),(e2),(e1),(e0))
394 #define _mm512_setr4_pd(e0,e1,e2,e3) \
395 _mm512_set4_pd((e3),(e2),(e1),(e0))
397 #define _mm512_setr4_ps(e0,e1,e2,e3) \
398 _mm512_set4_ps((e3),(e2),(e1),(e0))
403 return (__m512d)__builtin_shufflevector((__v2df) __A,
405 0, 0, 0, 0, 0, 0, 0, 0);
413 return __builtin_shufflevector(__a, __a, 0, 1, 2, 3, -1, -1, -1, -1);
419 return __builtin_shufflevector(__a, __a, 0, 1, 2, 3, 4, 5, 6, 7,
420 -1, -1, -1, -1, -1, -1, -1, -1);
426 return __builtin_shufflevector(__a, __a, 0, 1);
432 return __builtin_shufflevector(__A, __A, 0, 1, 2, 3);
438 return __builtin_shufflevector(__a, __a, 0, 1, 2, 3);
444 return __builtin_shufflevector(__A, __A, 0, 1, 2, 3, 4, 5, 6, 7);
450 return (__m512) (__A);
456 return (__m512i) (__A);
462 return __builtin_shufflevector( __A, __A, 0, 1, -1, -1, -1, -1, -1, -1);
468 return (__m512d) (__A);
474 return (__m512i) (__A);
480 return __builtin_shufflevector( __A, __A, 0, 1, 2, 3, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1);
486 return __builtin_shufflevector( __A, __A, 0, 1, -1, -1, -1, -1, -1, -1);
492 return __builtin_shufflevector( __A, __A, 0, 1, 2, 3, -1, -1, -1, -1);
498 return (__m512) (__A);
504 return (__m512d) (__A);
510 return (__m128i)__builtin_shufflevector(__A, __A , 0, 1);
516 return (__m256i)__builtin_shufflevector(__A, __A , 0, 1, 2, 3);
522 return (__mmask16)__a;
547 return __builtin_shufflevector((__v2df)__a, (__v2df)
_mm_setzero_pd(), 0, 1, 2, 3, 2, 3, 2, 3);
566 return __builtin_shufflevector((__v4df)__a, (__v4df)
_mm256_setzero_pd(), 0, 1, 2, 3, 4, 5, 6, 7);
584 return __builtin_shufflevector((__v4sf)__a, (__v4sf)
_mm_setzero_ps(), 0, 1, 2, 3, 4, 5, 6, 7, 4, 5, 6, 7, 4, 5, 6, 7);
602 return __builtin_shufflevector((__v8sf)__a, (__v8sf)
_mm256_setzero_ps(), 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
620 return __builtin_shufflevector((__v2di)__a, (__v2di)
_mm_setzero_si128(), 0, 1, 2, 3, 2, 3, 2, 3);
638 return __builtin_shufflevector((__v4di)__a, (__v4di)
_mm256_setzero_si256(), 0, 1, 2, 3, 4, 5, 6, 7);
645 return (__m512i)((__v16su)__a & (__v16su)
__b);
651 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__k,
666 return (__m512i)((__v8du)__a & (__v8du)
__b);
672 return (__m512i) __builtin_ia32_selectq_512 ((__mmask8) __k,
687 return (__m512i)(~(__v8du)(__A) & (__v8du)__B);
693 return (__m512i)(~(__v16su)(__A) & (__v16su)__B);
699 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
714 return (__m512i)(~(__v8du)(__A) & (__v8du)__B);
720 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
735 return (__m512i)((__v16su)__a | (__v16su)
__b);
741 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__k,
755 return (__m512i)((__v8du)__a | (__v8du)
__b);
761 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__k,
775 return (__m512i)((__v16su)__a ^ (__v16su)
__b);
781 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__k,
795 return (__m512i)((__v8du)__a ^ (__v8du)
__b);
801 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__k,
815 return (__m512i)((__v8du)__a & (__v8du)
__b);
821 return (__m512i)((__v8du)__a | (__v8du)
__b);
827 return (__m512i)((__v8du)__a ^ (__v8du)
__b);
835 return (__m512d)((__v8df)__a + (__v8df)
__b);
841 return (__m512)((__v16sf)__a + (__v16sf)
__b);
847 return (__m512d)((__v8df)__a * (__v8df)
__b);
853 return (__m512)((__v16sf)__a * (__v16sf)
__b);
859 return (__m512d)((__v8df)__a - (__v8df)
__b);
865 return (__m512)((__v16sf)__a - (__v16sf)
__b);
871 return (__m512i) ((__v8du) __A + (__v8du) __B);
877 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
885 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
893 return (__m512i) ((__v8du) __A - (__v8du) __B);
899 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
907 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
915 return (__m512i) ((__v16su) __A + (__v16su) __B);
921 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
929 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
937 return (__m512i) ((__v16su) __A - (__v16su) __B);
943 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
951 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
956 #define _mm512_mask_max_round_pd(W, U, A, B, R) __extension__ ({ \
957 (__m512d)__builtin_ia32_maxpd512_mask((__v8df)(__m512d)(A), \
958 (__v8df)(__m512d)(B), \
959 (__v8df)(__m512d)(W), (__mmask8)(U), \
962 #define _mm512_maskz_max_round_pd(U, A, B, R) __extension__ ({ \
963 (__m512d)__builtin_ia32_maxpd512_mask((__v8df)(__m512d)(A), \
964 (__v8df)(__m512d)(B), \
965 (__v8df)_mm512_setzero_pd(), \
966 (__mmask8)(U), (int)(R)); })
968 #define _mm512_max_round_pd(A, B, R) __extension__ ({ \
969 (__m512d)__builtin_ia32_maxpd512_mask((__v8df)(__m512d)(A), \
970 (__v8df)(__m512d)(B), \
971 (__v8df)_mm512_undefined_pd(), \
972 (__mmask8)-1, (int)(R)); })
977 return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
988 return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
998 return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
1006 #define _mm512_mask_max_round_ps(W, U, A, B, R) __extension__ ({ \
1007 (__m512)__builtin_ia32_maxps512_mask((__v16sf)(__m512)(A), \
1008 (__v16sf)(__m512)(B), \
1009 (__v16sf)(__m512)(W), (__mmask16)(U), \
1012 #define _mm512_maskz_max_round_ps(U, A, B, R) __extension__ ({ \
1013 (__m512)__builtin_ia32_maxps512_mask((__v16sf)(__m512)(A), \
1014 (__v16sf)(__m512)(B), \
1015 (__v16sf)_mm512_setzero_ps(), \
1016 (__mmask16)(U), (int)(R)); })
1018 #define _mm512_max_round_ps(A, B, R) __extension__ ({ \
1019 (__m512)__builtin_ia32_maxps512_mask((__v16sf)(__m512)(A), \
1020 (__v16sf)(__m512)(B), \
1021 (__v16sf)_mm512_undefined_ps(), \
1022 (__mmask16)-1, (int)(R)); })
1027 return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
1038 return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
1048 return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
1058 return (__m128) __builtin_ia32_maxss_round_mask ((__v4sf) __A,
1067 return (__m128) __builtin_ia32_maxss_round_mask ((__v4sf) __A,
1074 #define _mm_max_round_ss(A, B, R) __extension__ ({ \
1075 (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1076 (__v4sf)(__m128)(B), \
1077 (__v4sf)_mm_setzero_ps(), \
1078 (__mmask8)-1, (int)(R)); })
1080 #define _mm_mask_max_round_ss(W, U, A, B, R) __extension__ ({ \
1081 (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1082 (__v4sf)(__m128)(B), \
1083 (__v4sf)(__m128)(W), (__mmask8)(U), \
1086 #define _mm_maskz_max_round_ss(U, A, B, R) __extension__ ({ \
1087 (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1088 (__v4sf)(__m128)(B), \
1089 (__v4sf)_mm_setzero_ps(), \
1090 (__mmask8)(U), (int)(R)); })
1094 return (__m128d) __builtin_ia32_maxsd_round_mask ((__v2df) __A,
1103 return (__m128d) __builtin_ia32_maxsd_round_mask ((__v2df) __A,
1110 #define _mm_max_round_sd(A, B, R) __extension__ ({ \
1111 (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1112 (__v2df)(__m128d)(B), \
1113 (__v2df)_mm_setzero_pd(), \
1114 (__mmask8)-1, (int)(R)); })
1116 #define _mm_mask_max_round_sd(W, U, A, B, R) __extension__ ({ \
1117 (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1118 (__v2df)(__m128d)(B), \
1119 (__v2df)(__m128d)(W), \
1120 (__mmask8)(U), (int)(R)); })
1122 #define _mm_maskz_max_round_sd(U, A, B, R) __extension__ ({ \
1123 (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1124 (__v2df)(__m128d)(B), \
1125 (__v2df)_mm_setzero_pd(), \
1126 (__mmask8)(U), (int)(R)); })
1128 static __inline __m512i
1132 return (__m512i) __builtin_ia32_pmaxsd512_mask ((__v16si) __A,
1142 return (__m512i) __builtin_ia32_pmaxsd512_mask ((__v16si) __A,
1144 (__v16si) __W, __M);
1150 return (__m512i) __builtin_ia32_pmaxsd512_mask ((__v16si) __A,
1160 return (__m512i) __builtin_ia32_pmaxud512_mask ((__v16si) __A,
1170 return (__m512i) __builtin_ia32_pmaxud512_mask ((__v16si) __A,
1172 (__v16si) __W, __M);
1178 return (__m512i) __builtin_ia32_pmaxud512_mask ((__v16si) __A,
1188 return (__m512i) __builtin_ia32_pmaxsq512_mask ((__v8di) __A,
1198 return (__m512i) __builtin_ia32_pmaxsq512_mask ((__v8di) __A,
1206 return (__m512i) __builtin_ia32_pmaxsq512_mask ((__v8di) __A,
1216 return (__m512i) __builtin_ia32_pmaxuq512_mask ((__v8di) __A,
1226 return (__m512i) __builtin_ia32_pmaxuq512_mask ((__v8di) __A,
1234 return (__m512i) __builtin_ia32_pmaxuq512_mask ((__v8di) __A,
1241 #define _mm512_mask_min_round_pd(W, U, A, B, R) __extension__ ({ \
1242 (__m512d)__builtin_ia32_minpd512_mask((__v8df)(__m512d)(A), \
1243 (__v8df)(__m512d)(B), \
1244 (__v8df)(__m512d)(W), (__mmask8)(U), \
1247 #define _mm512_maskz_min_round_pd(U, A, B, R) __extension__ ({ \
1248 (__m512d)__builtin_ia32_minpd512_mask((__v8df)(__m512d)(A), \
1249 (__v8df)(__m512d)(B), \
1250 (__v8df)_mm512_setzero_pd(), \
1251 (__mmask8)(U), (int)(R)); })
1253 #define _mm512_min_round_pd(A, B, R) __extension__ ({ \
1254 (__m512d)__builtin_ia32_minpd512_mask((__v8df)(__m512d)(A), \
1255 (__v8df)(__m512d)(B), \
1256 (__v8df)_mm512_undefined_pd(), \
1257 (__mmask8)-1, (int)(R)); })
1262 return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
1273 return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
1280 #define _mm512_mask_min_round_ps(W, U, A, B, R) __extension__ ({ \
1281 (__m512)__builtin_ia32_minps512_mask((__v16sf)(__m512)(A), \
1282 (__v16sf)(__m512)(B), \
1283 (__v16sf)(__m512)(W), (__mmask16)(U), \
1286 #define _mm512_maskz_min_round_ps(U, A, B, R) __extension__ ({ \
1287 (__m512)__builtin_ia32_minps512_mask((__v16sf)(__m512)(A), \
1288 (__v16sf)(__m512)(B), \
1289 (__v16sf)_mm512_setzero_ps(), \
1290 (__mmask16)(U), (int)(R)); })
1292 #define _mm512_min_round_ps(A, B, R) __extension__ ({ \
1293 (__m512)__builtin_ia32_minps512_mask((__v16sf)(__m512)(A), \
1294 (__v16sf)(__m512)(B), \
1295 (__v16sf)_mm512_undefined_ps(), \
1296 (__mmask16)-1, (int)(R)); })
1301 return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
1312 return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
1323 return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
1333 return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
1343 return (__m128) __builtin_ia32_minss_round_mask ((__v4sf) __A,
1352 return (__m128) __builtin_ia32_minss_round_mask ((__v4sf) __A,
1359 #define _mm_min_round_ss(A, B, R) __extension__ ({ \
1360 (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1361 (__v4sf)(__m128)(B), \
1362 (__v4sf)_mm_setzero_ps(), \
1363 (__mmask8)-1, (int)(R)); })
1365 #define _mm_mask_min_round_ss(W, U, A, B, R) __extension__ ({ \
1366 (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1367 (__v4sf)(__m128)(B), \
1368 (__v4sf)(__m128)(W), (__mmask8)(U), \
1371 #define _mm_maskz_min_round_ss(U, A, B, R) __extension__ ({ \
1372 (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1373 (__v4sf)(__m128)(B), \
1374 (__v4sf)_mm_setzero_ps(), \
1375 (__mmask8)(U), (int)(R)); })
1379 return (__m128d) __builtin_ia32_minsd_round_mask ((__v2df) __A,
1388 return (__m128d) __builtin_ia32_minsd_round_mask ((__v2df) __A,
1395 #define _mm_min_round_sd(A, B, R) __extension__ ({ \
1396 (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1397 (__v2df)(__m128d)(B), \
1398 (__v2df)_mm_setzero_pd(), \
1399 (__mmask8)-1, (int)(R)); })
1401 #define _mm_mask_min_round_sd(W, U, A, B, R) __extension__ ({ \
1402 (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1403 (__v2df)(__m128d)(B), \
1404 (__v2df)(__m128d)(W), \
1405 (__mmask8)(U), (int)(R)); })
1407 #define _mm_maskz_min_round_sd(U, A, B, R) __extension__ ({ \
1408 (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1409 (__v2df)(__m128d)(B), \
1410 (__v2df)_mm_setzero_pd(), \
1411 (__mmask8)(U), (int)(R)); })
1413 static __inline __m512i
1417 return (__m512i) __builtin_ia32_pminsd512_mask ((__v16si) __A,
1427 return (__m512i) __builtin_ia32_pminsd512_mask ((__v16si) __A,
1429 (__v16si) __W, __M);
1435 return (__m512i) __builtin_ia32_pminsd512_mask ((__v16si) __A,
1445 return (__m512i) __builtin_ia32_pminud512_mask ((__v16si) __A,
1455 return (__m512i) __builtin_ia32_pminud512_mask ((__v16si) __A,
1457 (__v16si) __W, __M);
1463 return (__m512i) __builtin_ia32_pminud512_mask ((__v16si) __A,
1473 return (__m512i) __builtin_ia32_pminsq512_mask ((__v8di) __A,
1483 return (__m512i) __builtin_ia32_pminsq512_mask ((__v8di) __A,
1491 return (__m512i) __builtin_ia32_pminsq512_mask ((__v8di) __A,
1501 return (__m512i) __builtin_ia32_pminuq512_mask ((__v8di) __A,
1511 return (__m512i) __builtin_ia32_pminuq512_mask ((__v8di) __A,
1519 return (__m512i) __builtin_ia32_pminuq512_mask ((__v8di) __A,
1529 return (__m512i)__builtin_ia32_pmuldq512((__v16si)__X, (__v16si) __Y);
1535 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
1543 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
1551 return (__m512i)__builtin_ia32_pmuludq512((__v16si)__X, (__v16si)__Y);
1557 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
1565 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
1573 return (__m512i) ((__v16su) __A * (__v16su) __B);
1579 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__M,
1587 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__M,
1592 #define _mm512_mask_sqrt_round_pd(W, U, A, R) __extension__ ({ \
1593 (__m512d)__builtin_ia32_sqrtpd512_mask((__v8df)(__m512d)(A), \
1594 (__v8df)(__m512d)(W), (__mmask8)(U), \
1597 #define _mm512_maskz_sqrt_round_pd(U, A, R) __extension__ ({ \
1598 (__m512d)__builtin_ia32_sqrtpd512_mask((__v8df)(__m512d)(A), \
1599 (__v8df)_mm512_setzero_pd(), \
1600 (__mmask8)(U), (int)(R)); })
1602 #define _mm512_sqrt_round_pd(A, R) __extension__ ({ \
1603 (__m512d)__builtin_ia32_sqrtpd512_mask((__v8df)(__m512d)(A), \
1604 (__v8df)_mm512_undefined_pd(), \
1605 (__mmask8)-1, (int)(R)); })
1610 return (__m512d)__builtin_ia32_sqrtpd512_mask((__v8df)__a,
1619 return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
1628 return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
1635 #define _mm512_mask_sqrt_round_ps(W, U, A, R) __extension__ ({ \
1636 (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)(__m512)(A), \
1637 (__v16sf)(__m512)(W), (__mmask16)(U), \
1640 #define _mm512_maskz_sqrt_round_ps(U, A, R) __extension__ ({ \
1641 (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)(__m512)(A), \
1642 (__v16sf)_mm512_setzero_ps(), \
1643 (__mmask16)(U), (int)(R)); })
1645 #define _mm512_sqrt_round_ps(A, R) __extension__ ({ \
1646 (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)(__m512)(A), \
1647 (__v16sf)_mm512_undefined_ps(), \
1648 (__mmask16)-1, (int)(R)); })
1653 return (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)__a,
1662 return (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)__A,
1671 return (__m512)__builtin_ia32_sqrtps512_mask((__v16sf)__A,
1680 return (__m512d) __builtin_ia32_rsqrt14pd512_mask ((__v8df) __A,
1688 return (__m512d) __builtin_ia32_rsqrt14pd512_mask ((__v8df) __A,
1696 return (__m512d) __builtin_ia32_rsqrt14pd512_mask ((__v8df) __A,
1705 return (__m512) __builtin_ia32_rsqrt14ps512_mask ((__v16sf) __A,
1714 return (__m512) __builtin_ia32_rsqrt14ps512_mask ((__v16sf) __A,
1722 return (__m512) __builtin_ia32_rsqrt14ps512_mask ((__v16sf) __A,
1731 return (__m128) __builtin_ia32_rsqrt14ss_mask ((__v4sf) __A,
1741 return (__m128) __builtin_ia32_rsqrt14ss_mask ((__v4sf) __A,
1750 return (__m128) __builtin_ia32_rsqrt14ss_mask ((__v4sf) __A,
1759 return (__m128d) __builtin_ia32_rsqrt14sd_mask ((__v2df) __A,
1769 return (__m128d) __builtin_ia32_rsqrt14sd_mask ( (__v2df) __A,
1778 return (__m128d) __builtin_ia32_rsqrt14sd_mask ( (__v2df) __A,
1787 return (__m512d) __builtin_ia32_rcp14pd512_mask ((__v8df) __A,
1796 return (__m512d) __builtin_ia32_rcp14pd512_mask ((__v8df) __A,
1804 return (__m512d) __builtin_ia32_rcp14pd512_mask ((__v8df) __A,
1813 return (__m512) __builtin_ia32_rcp14ps512_mask ((__v16sf) __A,
1822 return (__m512) __builtin_ia32_rcp14ps512_mask ((__v16sf) __A,
1830 return (__m512) __builtin_ia32_rcp14ps512_mask ((__v16sf) __A,
1839 return (__m128) __builtin_ia32_rcp14ss_mask ((__v4sf) __A,
1849 return (__m128) __builtin_ia32_rcp14ss_mask ((__v4sf) __A,
1858 return (__m128) __builtin_ia32_rcp14ss_mask ((__v4sf) __A,
1867 return (__m128d) __builtin_ia32_rcp14sd_mask ((__v2df) __A,
1877 return (__m128d) __builtin_ia32_rcp14sd_mask ( (__v2df) __A,
1886 return (__m128d) __builtin_ia32_rcp14sd_mask ( (__v2df) __A,
1895 return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A,
1904 return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A,
1913 return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A,
1922 return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A,
1931 return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A,
1940 return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A,
1949 return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A,
1958 return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A,
1967 return (__m512i) __builtin_ia32_pabsq512_mask ((__v8di) __A,
1976 return (__m512i) __builtin_ia32_pabsq512_mask ((__v8di) __A,
1984 return (__m512i) __builtin_ia32_pabsq512_mask ((__v8di) __A,
1993 return (__m512i) __builtin_ia32_pabsd512_mask ((__v16si) __A,
2002 return (__m512i) __builtin_ia32_pabsd512_mask ((__v16si) __A,
2010 return (__m512i) __builtin_ia32_pabsd512_mask ((__v16si) __A,
2018 return (__m128) __builtin_ia32_addss_round_mask ((__v4sf) __A,
2027 return (__m128) __builtin_ia32_addss_round_mask ((__v4sf) __A,
2034 #define _mm_add_round_ss(A, B, R) __extension__ ({ \
2035 (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
2036 (__v4sf)(__m128)(B), \
2037 (__v4sf)_mm_setzero_ps(), \
2038 (__mmask8)-1, (int)(R)); })
2040 #define _mm_mask_add_round_ss(W, U, A, B, R) __extension__ ({ \
2041 (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
2042 (__v4sf)(__m128)(B), \
2043 (__v4sf)(__m128)(W), (__mmask8)(U), \
2046 #define _mm_maskz_add_round_ss(U, A, B, R) __extension__ ({ \
2047 (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
2048 (__v4sf)(__m128)(B), \
2049 (__v4sf)_mm_setzero_ps(), \
2050 (__mmask8)(U), (int)(R)); })
2054 return (__m128d) __builtin_ia32_addsd_round_mask ((__v2df) __A,
2063 return (__m128d) __builtin_ia32_addsd_round_mask ((__v2df) __A,
2069 #define _mm_add_round_sd(A, B, R) __extension__ ({ \
2070 (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
2071 (__v2df)(__m128d)(B), \
2072 (__v2df)_mm_setzero_pd(), \
2073 (__mmask8)-1, (int)(R)); })
2075 #define _mm_mask_add_round_sd(W, U, A, B, R) __extension__ ({ \
2076 (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
2077 (__v2df)(__m128d)(B), \
2078 (__v2df)(__m128d)(W), \
2079 (__mmask8)(U), (int)(R)); })
2081 #define _mm_maskz_add_round_sd(U, A, B, R) __extension__ ({ \
2082 (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
2083 (__v2df)(__m128d)(B), \
2084 (__v2df)_mm_setzero_pd(), \
2085 (__mmask8)(U), (int)(R)); })
2089 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2096 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2103 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2110 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2115 #define _mm512_add_round_pd(A, B, R) __extension__ ({ \
2116 (__m512d)__builtin_ia32_addpd512_mask((__v8df)(__m512d)(A), \
2117 (__v8df)(__m512d)(B), \
2118 (__v8df)_mm512_setzero_pd(), \
2119 (__mmask8)-1, (int)(R)); })
2121 #define _mm512_mask_add_round_pd(W, U, A, B, R) __extension__ ({ \
2122 (__m512d)__builtin_ia32_addpd512_mask((__v8df)(__m512d)(A), \
2123 (__v8df)(__m512d)(B), \
2124 (__v8df)(__m512d)(W), (__mmask8)(U), \
2127 #define _mm512_maskz_add_round_pd(U, A, B, R) __extension__ ({ \
2128 (__m512d)__builtin_ia32_addpd512_mask((__v8df)(__m512d)(A), \
2129 (__v8df)(__m512d)(B), \
2130 (__v8df)_mm512_setzero_pd(), \
2131 (__mmask8)(U), (int)(R)); })
2133 #define _mm512_add_round_ps(A, B, R) __extension__ ({ \
2134 (__m512)__builtin_ia32_addps512_mask((__v16sf)(__m512)(A), \
2135 (__v16sf)(__m512)(B), \
2136 (__v16sf)_mm512_setzero_ps(), \
2137 (__mmask16)-1, (int)(R)); })
2139 #define _mm512_mask_add_round_ps(W, U, A, B, R) __extension__ ({ \
2140 (__m512)__builtin_ia32_addps512_mask((__v16sf)(__m512)(A), \
2141 (__v16sf)(__m512)(B), \
2142 (__v16sf)(__m512)(W), (__mmask16)(U), \
2145 #define _mm512_maskz_add_round_ps(U, A, B, R) __extension__ ({ \
2146 (__m512)__builtin_ia32_addps512_mask((__v16sf)(__m512)(A), \
2147 (__v16sf)(__m512)(B), \
2148 (__v16sf)_mm512_setzero_ps(), \
2149 (__mmask16)(U), (int)(R)); })
2153 return (__m128) __builtin_ia32_subss_round_mask ((__v4sf) __A,
2162 return (__m128) __builtin_ia32_subss_round_mask ((__v4sf) __A,
2168 #define _mm_sub_round_ss(A, B, R) __extension__ ({ \
2169 (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2170 (__v4sf)(__m128)(B), \
2171 (__v4sf)_mm_setzero_ps(), \
2172 (__mmask8)-1, (int)(R)); })
2174 #define _mm_mask_sub_round_ss(W, U, A, B, R) __extension__ ({ \
2175 (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2176 (__v4sf)(__m128)(B), \
2177 (__v4sf)(__m128)(W), (__mmask8)(U), \
2180 #define _mm_maskz_sub_round_ss(U, A, B, R) __extension__ ({ \
2181 (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2182 (__v4sf)(__m128)(B), \
2183 (__v4sf)_mm_setzero_ps(), \
2184 (__mmask8)(U), (int)(R)); })
2188 return (__m128d) __builtin_ia32_subsd_round_mask ((__v2df) __A,
2197 return (__m128d) __builtin_ia32_subsd_round_mask ((__v2df) __A,
2204 #define _mm_sub_round_sd(A, B, R) __extension__ ({ \
2205 (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2206 (__v2df)(__m128d)(B), \
2207 (__v2df)_mm_setzero_pd(), \
2208 (__mmask8)-1, (int)(R)); })
2210 #define _mm_mask_sub_round_sd(W, U, A, B, R) __extension__ ({ \
2211 (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2212 (__v2df)(__m128d)(B), \
2213 (__v2df)(__m128d)(W), \
2214 (__mmask8)(U), (int)(R)); })
2216 #define _mm_maskz_sub_round_sd(U, A, B, R) __extension__ ({ \
2217 (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2218 (__v2df)(__m128d)(B), \
2219 (__v2df)_mm_setzero_pd(), \
2220 (__mmask8)(U), (int)(R)); })
2224 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2231 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2238 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2245 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2250 #define _mm512_sub_round_pd(A, B, R) __extension__ ({ \
2251 (__m512d)__builtin_ia32_subpd512_mask((__v8df)(__m512d)(A), \
2252 (__v8df)(__m512d)(B), \
2253 (__v8df)_mm512_setzero_pd(), \
2254 (__mmask8)-1, (int)(R)); })
2256 #define _mm512_mask_sub_round_pd(W, U, A, B, R) __extension__ ({ \
2257 (__m512d)__builtin_ia32_subpd512_mask((__v8df)(__m512d)(A), \
2258 (__v8df)(__m512d)(B), \
2259 (__v8df)(__m512d)(W), (__mmask8)(U), \
2262 #define _mm512_maskz_sub_round_pd(U, A, B, R) __extension__ ({ \
2263 (__m512d)__builtin_ia32_subpd512_mask((__v8df)(__m512d)(A), \
2264 (__v8df)(__m512d)(B), \
2265 (__v8df)_mm512_setzero_pd(), \
2266 (__mmask8)(U), (int)(R)); })
2268 #define _mm512_sub_round_ps(A, B, R) __extension__ ({ \
2269 (__m512)__builtin_ia32_subps512_mask((__v16sf)(__m512)(A), \
2270 (__v16sf)(__m512)(B), \
2271 (__v16sf)_mm512_setzero_ps(), \
2272 (__mmask16)-1, (int)(R)); })
2274 #define _mm512_mask_sub_round_ps(W, U, A, B, R) __extension__ ({ \
2275 (__m512)__builtin_ia32_subps512_mask((__v16sf)(__m512)(A), \
2276 (__v16sf)(__m512)(B), \
2277 (__v16sf)(__m512)(W), (__mmask16)(U), \
2280 #define _mm512_maskz_sub_round_ps(U, A, B, R) __extension__ ({ \
2281 (__m512)__builtin_ia32_subps512_mask((__v16sf)(__m512)(A), \
2282 (__v16sf)(__m512)(B), \
2283 (__v16sf)_mm512_setzero_ps(), \
2284 (__mmask16)(U), (int)(R)); });
2288 return (__m128) __builtin_ia32_mulss_round_mask ((__v4sf) __A,
2297 return (__m128) __builtin_ia32_mulss_round_mask ((__v4sf) __A,
2303 #define _mm_mul_round_ss(A, B, R) __extension__ ({ \
2304 (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2305 (__v4sf)(__m128)(B), \
2306 (__v4sf)_mm_setzero_ps(), \
2307 (__mmask8)-1, (int)(R)); })
2309 #define _mm_mask_mul_round_ss(W, U, A, B, R) __extension__ ({ \
2310 (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2311 (__v4sf)(__m128)(B), \
2312 (__v4sf)(__m128)(W), (__mmask8)(U), \
2315 #define _mm_maskz_mul_round_ss(U, A, B, R) __extension__ ({ \
2316 (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2317 (__v4sf)(__m128)(B), \
2318 (__v4sf)_mm_setzero_ps(), \
2319 (__mmask8)(U), (int)(R)); })
2323 return (__m128d) __builtin_ia32_mulsd_round_mask ((__v2df) __A,
2332 return (__m128d) __builtin_ia32_mulsd_round_mask ((__v2df) __A,
2339 #define _mm_mul_round_sd(A, B, R) __extension__ ({ \
2340 (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2341 (__v2df)(__m128d)(B), \
2342 (__v2df)_mm_setzero_pd(), \
2343 (__mmask8)-1, (int)(R)); })
2345 #define _mm_mask_mul_round_sd(W, U, A, B, R) __extension__ ({ \
2346 (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2347 (__v2df)(__m128d)(B), \
2348 (__v2df)(__m128d)(W), \
2349 (__mmask8)(U), (int)(R)); })
2351 #define _mm_maskz_mul_round_sd(U, A, B, R) __extension__ ({ \
2352 (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2353 (__v2df)(__m128d)(B), \
2354 (__v2df)_mm_setzero_pd(), \
2355 (__mmask8)(U), (int)(R)); })
2359 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2366 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2373 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2380 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2385 #define _mm512_mul_round_pd(A, B, R) __extension__ ({ \
2386 (__m512d)__builtin_ia32_mulpd512_mask((__v8df)(__m512d)(A), \
2387 (__v8df)(__m512d)(B), \
2388 (__v8df)_mm512_setzero_pd(), \
2389 (__mmask8)-1, (int)(R)); })
2391 #define _mm512_mask_mul_round_pd(W, U, A, B, R) __extension__ ({ \
2392 (__m512d)__builtin_ia32_mulpd512_mask((__v8df)(__m512d)(A), \
2393 (__v8df)(__m512d)(B), \
2394 (__v8df)(__m512d)(W), (__mmask8)(U), \
2397 #define _mm512_maskz_mul_round_pd(U, A, B, R) __extension__ ({ \
2398 (__m512d)__builtin_ia32_mulpd512_mask((__v8df)(__m512d)(A), \
2399 (__v8df)(__m512d)(B), \
2400 (__v8df)_mm512_setzero_pd(), \
2401 (__mmask8)(U), (int)(R)); })
2403 #define _mm512_mul_round_ps(A, B, R) __extension__ ({ \
2404 (__m512)__builtin_ia32_mulps512_mask((__v16sf)(__m512)(A), \
2405 (__v16sf)(__m512)(B), \
2406 (__v16sf)_mm512_setzero_ps(), \
2407 (__mmask16)-1, (int)(R)); })
2409 #define _mm512_mask_mul_round_ps(W, U, A, B, R) __extension__ ({ \
2410 (__m512)__builtin_ia32_mulps512_mask((__v16sf)(__m512)(A), \
2411 (__v16sf)(__m512)(B), \
2412 (__v16sf)(__m512)(W), (__mmask16)(U), \
2415 #define _mm512_maskz_mul_round_ps(U, A, B, R) __extension__ ({ \
2416 (__m512)__builtin_ia32_mulps512_mask((__v16sf)(__m512)(A), \
2417 (__v16sf)(__m512)(B), \
2418 (__v16sf)_mm512_setzero_ps(), \
2419 (__mmask16)(U), (int)(R)); });
2423 return (__m128) __builtin_ia32_divss_round_mask ((__v4sf) __A,
2432 return (__m128) __builtin_ia32_divss_round_mask ((__v4sf) __A,
2439 #define _mm_div_round_ss(A, B, R) __extension__ ({ \
2440 (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2441 (__v4sf)(__m128)(B), \
2442 (__v4sf)_mm_setzero_ps(), \
2443 (__mmask8)-1, (int)(R)); })
2445 #define _mm_mask_div_round_ss(W, U, A, B, R) __extension__ ({ \
2446 (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2447 (__v4sf)(__m128)(B), \
2448 (__v4sf)(__m128)(W), (__mmask8)(U), \
2451 #define _mm_maskz_div_round_ss(U, A, B, R) __extension__ ({ \
2452 (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2453 (__v4sf)(__m128)(B), \
2454 (__v4sf)_mm_setzero_ps(), \
2455 (__mmask8)(U), (int)(R)); })
2459 return (__m128d) __builtin_ia32_divsd_round_mask ((__v2df) __A,
2468 return (__m128d) __builtin_ia32_divsd_round_mask ((__v2df) __A,
2475 #define _mm_div_round_sd(A, B, R) __extension__ ({ \
2476 (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2477 (__v2df)(__m128d)(B), \
2478 (__v2df)_mm_setzero_pd(), \
2479 (__mmask8)-1, (int)(R)); })
2481 #define _mm_mask_div_round_sd(W, U, A, B, R) __extension__ ({ \
2482 (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2483 (__v2df)(__m128d)(B), \
2484 (__v2df)(__m128d)(W), \
2485 (__mmask8)(U), (int)(R)); })
2487 #define _mm_maskz_div_round_sd(U, A, B, R) __extension__ ({ \
2488 (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2489 (__v2df)(__m128d)(B), \
2490 (__v2df)_mm_setzero_pd(), \
2491 (__mmask8)(U), (int)(R)); })
2496 return (__m512d)((__v8df)__a/(__v8df)
__b);
2501 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2508 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
2516 return (__m512)((__v16sf)__a/(__v16sf)
__b);
2521 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2528 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
2533 #define _mm512_div_round_pd(A, B, R) __extension__ ({ \
2534 (__m512d)__builtin_ia32_divpd512_mask((__v8df)(__m512d)(A), \
2535 (__v8df)(__m512d)(B), \
2536 (__v8df)_mm512_setzero_pd(), \
2537 (__mmask8)-1, (int)(R)); })
2539 #define _mm512_mask_div_round_pd(W, U, A, B, R) __extension__ ({ \
2540 (__m512d)__builtin_ia32_divpd512_mask((__v8df)(__m512d)(A), \
2541 (__v8df)(__m512d)(B), \
2542 (__v8df)(__m512d)(W), (__mmask8)(U), \
2545 #define _mm512_maskz_div_round_pd(U, A, B, R) __extension__ ({ \
2546 (__m512d)__builtin_ia32_divpd512_mask((__v8df)(__m512d)(A), \
2547 (__v8df)(__m512d)(B), \
2548 (__v8df)_mm512_setzero_pd(), \
2549 (__mmask8)(U), (int)(R)); })
2551 #define _mm512_div_round_ps(A, B, R) __extension__ ({ \
2552 (__m512)__builtin_ia32_divps512_mask((__v16sf)(__m512)(A), \
2553 (__v16sf)(__m512)(B), \
2554 (__v16sf)_mm512_setzero_ps(), \
2555 (__mmask16)-1, (int)(R)); })
2557 #define _mm512_mask_div_round_ps(W, U, A, B, R) __extension__ ({ \
2558 (__m512)__builtin_ia32_divps512_mask((__v16sf)(__m512)(A), \
2559 (__v16sf)(__m512)(B), \
2560 (__v16sf)(__m512)(W), (__mmask16)(U), \
2563 #define _mm512_maskz_div_round_ps(U, A, B, R) __extension__ ({ \
2564 (__m512)__builtin_ia32_divps512_mask((__v16sf)(__m512)(A), \
2565 (__v16sf)(__m512)(B), \
2566 (__v16sf)_mm512_setzero_ps(), \
2567 (__mmask16)(U), (int)(R)); });
2569 #define _mm512_roundscale_ps(A, B) __extension__ ({ \
2570 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(B), \
2571 (__v16sf)(__m512)(A), (__mmask16)-1, \
2572 _MM_FROUND_CUR_DIRECTION); })
2574 #define _mm512_mask_roundscale_ps(A, B, C, imm) __extension__ ({\
2575 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2576 (__v16sf)(__m512)(A), (__mmask16)(B), \
2577 _MM_FROUND_CUR_DIRECTION); })
2579 #define _mm512_maskz_roundscale_ps(A, B, imm) __extension__ ({\
2580 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2581 (__v16sf)_mm512_setzero_ps(), \
2583 _MM_FROUND_CUR_DIRECTION); })
2585 #define _mm512_mask_roundscale_round_ps(A, B, C, imm, R) __extension__ ({ \
2586 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2587 (__v16sf)(__m512)(A), (__mmask16)(B), \
2590 #define _mm512_maskz_roundscale_round_ps(A, B, imm, R) __extension__ ({ \
2591 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2592 (__v16sf)_mm512_setzero_ps(), \
2593 (__mmask16)(A), (int)(R)); })
2595 #define _mm512_roundscale_round_ps(A, imm, R) __extension__ ({ \
2596 (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(imm), \
2597 (__v16sf)_mm512_undefined_ps(), \
2598 (__mmask16)-1, (int)(R)); })
2600 #define _mm512_roundscale_pd(A, B) __extension__ ({ \
2601 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(B), \
2602 (__v8df)(__m512d)(A), (__mmask8)-1, \
2603 _MM_FROUND_CUR_DIRECTION); })
2605 #define _mm512_mask_roundscale_pd(A, B, C, imm) __extension__ ({\
2606 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2607 (__v8df)(__m512d)(A), (__mmask8)(B), \
2608 _MM_FROUND_CUR_DIRECTION); })
2610 #define _mm512_maskz_roundscale_pd(A, B, imm) __extension__ ({\
2611 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2612 (__v8df)_mm512_setzero_pd(), \
2614 _MM_FROUND_CUR_DIRECTION); })
2616 #define _mm512_mask_roundscale_round_pd(A, B, C, imm, R) __extension__ ({ \
2617 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2618 (__v8df)(__m512d)(A), (__mmask8)(B), \
2621 #define _mm512_maskz_roundscale_round_pd(A, B, imm, R) __extension__ ({ \
2622 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2623 (__v8df)_mm512_setzero_pd(), \
2624 (__mmask8)(A), (int)(R)); })
2626 #define _mm512_roundscale_round_pd(A, imm, R) __extension__ ({ \
2627 (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(imm), \
2628 (__v8df)_mm512_undefined_pd(), \
2629 (__mmask8)-1, (int)(R)); })
2631 #define _mm512_fmadd_round_pd(A, B, C, R) __extension__ ({ \
2632 (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2633 (__v8df)(__m512d)(B), \
2634 (__v8df)(__m512d)(C), (__mmask8)-1, \
2638 #define _mm512_mask_fmadd_round_pd(A, U, B, C, R) __extension__ ({ \
2639 (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2640 (__v8df)(__m512d)(B), \
2641 (__v8df)(__m512d)(C), \
2642 (__mmask8)(U), (int)(R)); })
2645 #define _mm512_mask3_fmadd_round_pd(A, B, C, U, R) __extension__ ({ \
2646 (__m512d)__builtin_ia32_vfmaddpd512_mask3((__v8df)(__m512d)(A), \
2647 (__v8df)(__m512d)(B), \
2648 (__v8df)(__m512d)(C), \
2649 (__mmask8)(U), (int)(R)); })
2652 #define _mm512_maskz_fmadd_round_pd(U, A, B, C, R) __extension__ ({ \
2653 (__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2654 (__v8df)(__m512d)(B), \
2655 (__v8df)(__m512d)(C), \
2656 (__mmask8)(U), (int)(R)); })
2659 #define _mm512_fmsub_round_pd(A, B, C, R) __extension__ ({ \
2660 (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2661 (__v8df)(__m512d)(B), \
2662 -(__v8df)(__m512d)(C), \
2663 (__mmask8)-1, (int)(R)); })
2666 #define _mm512_mask_fmsub_round_pd(A, U, B, C, R) __extension__ ({ \
2667 (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2668 (__v8df)(__m512d)(B), \
2669 -(__v8df)(__m512d)(C), \
2670 (__mmask8)(U), (int)(R)); })
2673 #define _mm512_maskz_fmsub_round_pd(U, A, B, C, R) __extension__ ({ \
2674 (__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2675 (__v8df)(__m512d)(B), \
2676 -(__v8df)(__m512d)(C), \
2677 (__mmask8)(U), (int)(R)); })
2680 #define _mm512_fnmadd_round_pd(A, B, C, R) __extension__ ({ \
2681 (__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2682 (__v8df)(__m512d)(B), \
2683 (__v8df)(__m512d)(C), (__mmask8)-1, \
2687 #define _mm512_mask3_fnmadd_round_pd(A, B, C, U, R) __extension__ ({ \
2688 (__m512d)__builtin_ia32_vfmaddpd512_mask3(-(__v8df)(__m512d)(A), \
2689 (__v8df)(__m512d)(B), \
2690 (__v8df)(__m512d)(C), \
2691 (__mmask8)(U), (int)(R)); })
2694 #define _mm512_maskz_fnmadd_round_pd(U, A, B, C, R) __extension__ ({ \
2695 (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2696 (__v8df)(__m512d)(B), \
2697 (__v8df)(__m512d)(C), \
2698 (__mmask8)(U), (int)(R)); })
2701 #define _mm512_fnmsub_round_pd(A, B, C, R) __extension__ ({ \
2702 (__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2703 (__v8df)(__m512d)(B), \
2704 -(__v8df)(__m512d)(C), \
2705 (__mmask8)-1, (int)(R)); })
2708 #define _mm512_maskz_fnmsub_round_pd(U, A, B, C, R) __extension__ ({ \
2709 (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2710 (__v8df)(__m512d)(B), \
2711 -(__v8df)(__m512d)(C), \
2712 (__mmask8)(U), (int)(R)); })
2718 return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
2728 return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
2738 return (__m512d) __builtin_ia32_vfmaddpd512_mask3 ((__v8df) __A,
2748 return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
2758 return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
2768 return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
2778 return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
2788 return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
2798 return (__m512d) __builtin_ia32_vfmaddpd512_mask3 (-(__v8df) __A,
2808 return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
2818 return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
2828 return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
2835 #define _mm512_fmadd_round_ps(A, B, C, R) __extension__ ({ \
2836 (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2837 (__v16sf)(__m512)(B), \
2838 (__v16sf)(__m512)(C), (__mmask16)-1, \
2842 #define _mm512_mask_fmadd_round_ps(A, U, B, C, R) __extension__ ({ \
2843 (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2844 (__v16sf)(__m512)(B), \
2845 (__v16sf)(__m512)(C), \
2846 (__mmask16)(U), (int)(R)); })
2849 #define _mm512_mask3_fmadd_round_ps(A, B, C, U, R) __extension__ ({ \
2850 (__m512)__builtin_ia32_vfmaddps512_mask3((__v16sf)(__m512)(A), \
2851 (__v16sf)(__m512)(B), \
2852 (__v16sf)(__m512)(C), \
2853 (__mmask16)(U), (int)(R)); })
2856 #define _mm512_maskz_fmadd_round_ps(U, A, B, C, R) __extension__ ({ \
2857 (__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2858 (__v16sf)(__m512)(B), \
2859 (__v16sf)(__m512)(C), \
2860 (__mmask16)(U), (int)(R)); })
2863 #define _mm512_fmsub_round_ps(A, B, C, R) __extension__ ({ \
2864 (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2865 (__v16sf)(__m512)(B), \
2866 -(__v16sf)(__m512)(C), \
2867 (__mmask16)-1, (int)(R)); })
2870 #define _mm512_mask_fmsub_round_ps(A, U, B, C, R) __extension__ ({ \
2871 (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2872 (__v16sf)(__m512)(B), \
2873 -(__v16sf)(__m512)(C), \
2874 (__mmask16)(U), (int)(R)); })
2877 #define _mm512_maskz_fmsub_round_ps(U, A, B, C, R) __extension__ ({ \
2878 (__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2879 (__v16sf)(__m512)(B), \
2880 -(__v16sf)(__m512)(C), \
2881 (__mmask16)(U), (int)(R)); })
2884 #define _mm512_fnmadd_round_ps(A, B, C, R) __extension__ ({ \
2885 (__m512)__builtin_ia32_vfmaddps512_mask(-(__v16sf)(__m512)(A), \
2886 (__v16sf)(__m512)(B), \
2887 (__v16sf)(__m512)(C), (__mmask16)-1, \
2891 #define _mm512_mask3_fnmadd_round_ps(A, B, C, U, R) __extension__ ({ \
2892 (__m512)__builtin_ia32_vfmaddps512_mask3(-(__v16sf)(__m512)(A), \
2893 (__v16sf)(__m512)(B), \
2894 (__v16sf)(__m512)(C), \
2895 (__mmask16)(U), (int)(R)); })
2898 #define _mm512_maskz_fnmadd_round_ps(U, A, B, C, R) __extension__ ({ \
2899 (__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2900 (__v16sf)(__m512)(B), \
2901 (__v16sf)(__m512)(C), \
2902 (__mmask16)(U), (int)(R)); })
2905 #define _mm512_fnmsub_round_ps(A, B, C, R) __extension__ ({ \
2906 (__m512)__builtin_ia32_vfmaddps512_mask(-(__v16sf)(__m512)(A), \
2907 (__v16sf)(__m512)(B), \
2908 -(__v16sf)(__m512)(C), \
2909 (__mmask16)-1, (int)(R)); })
2912 #define _mm512_maskz_fnmsub_round_ps(U, A, B, C, R) __extension__ ({ \
2913 (__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2914 (__v16sf)(__m512)(B), \
2915 -(__v16sf)(__m512)(C), \
2916 (__mmask16)(U), (int)(R)); })
2922 return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
2932 return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
2942 return (__m512) __builtin_ia32_vfmaddps512_mask3 ((__v16sf) __A,
2952 return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
2962 return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
2972 return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
2982 return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
2992 return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
3002 return (__m512) __builtin_ia32_vfmaddps512_mask3 (-(__v16sf) __A,
3012 return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
3022 return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
3032 return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
3039 #define _mm512_fmaddsub_round_pd(A, B, C, R) __extension__ ({ \
3040 (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
3041 (__v8df)(__m512d)(B), \
3042 (__v8df)(__m512d)(C), \
3043 (__mmask8)-1, (int)(R)); })
3046 #define _mm512_mask_fmaddsub_round_pd(A, U, B, C, R) __extension__ ({ \
3047 (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
3048 (__v8df)(__m512d)(B), \
3049 (__v8df)(__m512d)(C), \
3050 (__mmask8)(U), (int)(R)); })
3053 #define _mm512_mask3_fmaddsub_round_pd(A, B, C, U, R) __extension__ ({ \
3054 (__m512d)__builtin_ia32_vfmaddsubpd512_mask3((__v8df)(__m512d)(A), \
3055 (__v8df)(__m512d)(B), \
3056 (__v8df)(__m512d)(C), \
3057 (__mmask8)(U), (int)(R)); })
3060 #define _mm512_maskz_fmaddsub_round_pd(U, A, B, C, R) __extension__ ({ \
3061 (__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
3062 (__v8df)(__m512d)(B), \
3063 (__v8df)(__m512d)(C), \
3064 (__mmask8)(U), (int)(R)); })
3067 #define _mm512_fmsubadd_round_pd(A, B, C, R) __extension__ ({ \
3068 (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
3069 (__v8df)(__m512d)(B), \
3070 -(__v8df)(__m512d)(C), \
3071 (__mmask8)-1, (int)(R)); })
3074 #define _mm512_mask_fmsubadd_round_pd(A, U, B, C, R) __extension__ ({ \
3075 (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
3076 (__v8df)(__m512d)(B), \
3077 -(__v8df)(__m512d)(C), \
3078 (__mmask8)(U), (int)(R)); })
3081 #define _mm512_maskz_fmsubadd_round_pd(U, A, B, C, R) __extension__ ({ \
3082 (__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
3083 (__v8df)(__m512d)(B), \
3084 -(__v8df)(__m512d)(C), \
3085 (__mmask8)(U), (int)(R)); })
3091 return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
3101 return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
3111 return (__m512d) __builtin_ia32_vfmaddsubpd512_mask3 ((__v8df) __A,
3121 return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
3131 return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
3141 return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
3151 return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
3158 #define _mm512_fmaddsub_round_ps(A, B, C, R) __extension__ ({ \
3159 (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
3160 (__v16sf)(__m512)(B), \
3161 (__v16sf)(__m512)(C), \
3162 (__mmask16)-1, (int)(R)); })
3165 #define _mm512_mask_fmaddsub_round_ps(A, U, B, C, R) __extension__ ({ \
3166 (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
3167 (__v16sf)(__m512)(B), \
3168 (__v16sf)(__m512)(C), \
3169 (__mmask16)(U), (int)(R)); })
3172 #define _mm512_mask3_fmaddsub_round_ps(A, B, C, U, R) __extension__ ({ \
3173 (__m512)__builtin_ia32_vfmaddsubps512_mask3((__v16sf)(__m512)(A), \
3174 (__v16sf)(__m512)(B), \
3175 (__v16sf)(__m512)(C), \
3176 (__mmask16)(U), (int)(R)); })
3179 #define _mm512_maskz_fmaddsub_round_ps(U, A, B, C, R) __extension__ ({ \
3180 (__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
3181 (__v16sf)(__m512)(B), \
3182 (__v16sf)(__m512)(C), \
3183 (__mmask16)(U), (int)(R)); })
3186 #define _mm512_fmsubadd_round_ps(A, B, C, R) __extension__ ({ \
3187 (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
3188 (__v16sf)(__m512)(B), \
3189 -(__v16sf)(__m512)(C), \
3190 (__mmask16)-1, (int)(R)); })
3193 #define _mm512_mask_fmsubadd_round_ps(A, U, B, C, R) __extension__ ({ \
3194 (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
3195 (__v16sf)(__m512)(B), \
3196 -(__v16sf)(__m512)(C), \
3197 (__mmask16)(U), (int)(R)); })
3200 #define _mm512_maskz_fmsubadd_round_ps(U, A, B, C, R) __extension__ ({ \
3201 (__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
3202 (__v16sf)(__m512)(B), \
3203 -(__v16sf)(__m512)(C), \
3204 (__mmask16)(U), (int)(R)); })
3210 return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
3220 return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
3230 return (__m512) __builtin_ia32_vfmaddsubps512_mask3 ((__v16sf) __A,
3240 return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
3250 return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
3260 return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
3270 return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
3277 #define _mm512_mask3_fmsub_round_pd(A, B, C, U, R) __extension__ ({ \
3278 (__m512d)__builtin_ia32_vfmsubpd512_mask3((__v8df)(__m512d)(A), \
3279 (__v8df)(__m512d)(B), \
3280 (__v8df)(__m512d)(C), \
3281 (__mmask8)(U), (int)(R)); })
3287 return (__m512d) __builtin_ia32_vfmsubpd512_mask3 ((__v8df) __A,
3294 #define _mm512_mask3_fmsub_round_ps(A, B, C, U, R) __extension__ ({ \
3295 (__m512)__builtin_ia32_vfmsubps512_mask3((__v16sf)(__m512)(A), \
3296 (__v16sf)(__m512)(B), \
3297 (__v16sf)(__m512)(C), \
3298 (__mmask16)(U), (int)(R)); })
3304 return (__m512) __builtin_ia32_vfmsubps512_mask3 ((__v16sf) __A,
3311 #define _mm512_mask3_fmsubadd_round_pd(A, B, C, U, R) __extension__ ({ \
3312 (__m512d)__builtin_ia32_vfmsubaddpd512_mask3((__v8df)(__m512d)(A), \
3313 (__v8df)(__m512d)(B), \
3314 (__v8df)(__m512d)(C), \
3315 (__mmask8)(U), (int)(R)); })
3321 return (__m512d) __builtin_ia32_vfmsubaddpd512_mask3 ((__v8df) __A,
3328 #define _mm512_mask3_fmsubadd_round_ps(A, B, C, U, R) __extension__ ({ \
3329 (__m512)__builtin_ia32_vfmsubaddps512_mask3((__v16sf)(__m512)(A), \
3330 (__v16sf)(__m512)(B), \
3331 (__v16sf)(__m512)(C), \
3332 (__mmask16)(U), (int)(R)); })
3338 return (__m512) __builtin_ia32_vfmsubaddps512_mask3 ((__v16sf) __A,
3345 #define _mm512_mask_fnmadd_round_pd(A, U, B, C, R) __extension__ ({ \
3346 (__m512d)__builtin_ia32_vfnmaddpd512_mask((__v8df)(__m512d)(A), \
3347 (__v8df)(__m512d)(B), \
3348 (__v8df)(__m512d)(C), \
3349 (__mmask8)(U), (int)(R)); })
3355 return (__m512d) __builtin_ia32_vfnmaddpd512_mask ((__v8df) __A,
3362 #define _mm512_mask_fnmadd_round_ps(A, U, B, C, R) __extension__ ({ \
3363 (__m512)__builtin_ia32_vfnmaddps512_mask((__v16sf)(__m512)(A), \
3364 (__v16sf)(__m512)(B), \
3365 (__v16sf)(__m512)(C), \
3366 (__mmask16)(U), (int)(R)); })
3372 return (__m512) __builtin_ia32_vfnmaddps512_mask ((__v16sf) __A,
3379 #define _mm512_mask_fnmsub_round_pd(A, U, B, C, R) __extension__ ({ \
3380 (__m512d)__builtin_ia32_vfnmsubpd512_mask((__v8df)(__m512d)(A), \
3381 (__v8df)(__m512d)(B), \
3382 (__v8df)(__m512d)(C), \
3383 (__mmask8)(U), (int)(R)); })
3386 #define _mm512_mask3_fnmsub_round_pd(A, B, C, U, R) __extension__ ({ \
3387 (__m512d)__builtin_ia32_vfnmsubpd512_mask3((__v8df)(__m512d)(A), \
3388 (__v8df)(__m512d)(B), \
3389 (__v8df)(__m512d)(C), \
3390 (__mmask8)(U), (int)(R)); })
3396 return (__m512d) __builtin_ia32_vfnmsubpd512_mask ((__v8df) __A,
3406 return (__m512d) __builtin_ia32_vfnmsubpd512_mask3 ((__v8df) __A,
3413 #define _mm512_mask_fnmsub_round_ps(A, U, B, C, R) __extension__ ({ \
3414 (__m512)__builtin_ia32_vfnmsubps512_mask((__v16sf)(__m512)(A), \
3415 (__v16sf)(__m512)(B), \
3416 (__v16sf)(__m512)(C), \
3417 (__mmask16)(U), (int)(R)); })
3420 #define _mm512_mask3_fnmsub_round_ps(A, B, C, U, R) __extension__ ({ \
3421 (__m512)__builtin_ia32_vfnmsubps512_mask3((__v16sf)(__m512)(A), \
3422 (__v16sf)(__m512)(B), \
3423 (__v16sf)(__m512)(C), \
3424 (__mmask16)(U), (int)(R)); })
3430 return (__m512) __builtin_ia32_vfnmsubps512_mask ((__v16sf) __A,
3440 return (__m512) __builtin_ia32_vfnmsubps512_mask3 ((__v16sf) __A,
3454 return (__m512i) __builtin_ia32_vpermt2vard512_mask ((__v16si) __I
3463 __m512i __I, __m512i __B)
3465 return (__m512i) __builtin_ia32_vpermt2vard512_mask ((__v16si) __I
3474 __m512i __I, __m512i __B)
3476 return (__m512i) __builtin_ia32_vpermt2vard512_maskz ((__v16si) __I
3486 return (__m512i) __builtin_ia32_vpermt2varq512_mask ((__v8di) __I
3497 return (__m512i) __builtin_ia32_vpermt2varq512_mask ((__v8di) __I
3507 __m512i __I, __m512i __B)
3509 return (__m512i) __builtin_ia32_vpermt2varq512_maskz ((__v8di) __I
3516 #define _mm512_alignr_epi64(A, B, I) __extension__ ({ \
3517 (__m512i)__builtin_shufflevector((__v8di)(__m512i)(B), \
3518 (__v8di)(__m512i)(A), \
3519 ((int)(I) & 0x7) + 0, \
3520 ((int)(I) & 0x7) + 1, \
3521 ((int)(I) & 0x7) + 2, \
3522 ((int)(I) & 0x7) + 3, \
3523 ((int)(I) & 0x7) + 4, \
3524 ((int)(I) & 0x7) + 5, \
3525 ((int)(I) & 0x7) + 6, \
3526 ((int)(I) & 0x7) + 7); })
3528 #define _mm512_mask_alignr_epi64(W, U, A, B, imm) __extension__({\
3529 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3530 (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3531 (__v8di)(__m512i)(W)); })
3533 #define _mm512_maskz_alignr_epi64(U, A, B, imm) __extension__({\
3534 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3535 (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3536 (__v8di)_mm512_setzero_si512()); })
3538 #define _mm512_alignr_epi32(A, B, I) __extension__ ({ \
3539 (__m512i)__builtin_shufflevector((__v16si)(__m512i)(B), \
3540 (__v16si)(__m512i)(A), \
3541 ((int)(I) & 0xf) + 0, \
3542 ((int)(I) & 0xf) + 1, \
3543 ((int)(I) & 0xf) + 2, \
3544 ((int)(I) & 0xf) + 3, \
3545 ((int)(I) & 0xf) + 4, \
3546 ((int)(I) & 0xf) + 5, \
3547 ((int)(I) & 0xf) + 6, \
3548 ((int)(I) & 0xf) + 7, \
3549 ((int)(I) & 0xf) + 8, \
3550 ((int)(I) & 0xf) + 9, \
3551 ((int)(I) & 0xf) + 10, \
3552 ((int)(I) & 0xf) + 11, \
3553 ((int)(I) & 0xf) + 12, \
3554 ((int)(I) & 0xf) + 13, \
3555 ((int)(I) & 0xf) + 14, \
3556 ((int)(I) & 0xf) + 15); })
3558 #define _mm512_mask_alignr_epi32(W, U, A, B, imm) __extension__ ({\
3559 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3560 (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3561 (__v16si)(__m512i)(W)); })
3563 #define _mm512_maskz_alignr_epi32(U, A, B, imm) __extension__({\
3564 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3565 (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3566 (__v16si)_mm512_setzero_si512()); })
3569 #define _mm512_extractf64x4_pd(A, I) __extension__ ({ \
3570 (__m256d)__builtin_shufflevector((__v8df)(__m512d)(A), \
3571 (__v8df)_mm512_undefined_pd(), \
3572 ((I) & 1) ? 4 : 0, \
3573 ((I) & 1) ? 5 : 1, \
3574 ((I) & 1) ? 6 : 2, \
3575 ((I) & 1) ? 7 : 3); })
3577 #define _mm512_mask_extractf64x4_pd(W, U, A, imm) __extension__ ({\
3578 (__m256d)__builtin_ia32_selectpd_256((__mmask8)(U), \
3579 (__v4df)_mm512_extractf64x4_pd((A), (imm)), \
3582 #define _mm512_maskz_extractf64x4_pd(U, A, imm) __extension__ ({\
3583 (__m256d)__builtin_ia32_selectpd_256((__mmask8)(U), \
3584 (__v4df)_mm512_extractf64x4_pd((A), (imm)), \
3585 (__v4df)_mm256_setzero_pd()); })
3587 #define _mm512_extractf32x4_ps(A, I) __extension__ ({ \
3588 (__m128)__builtin_shufflevector((__v16sf)(__m512)(A), \
3589 (__v16sf)_mm512_undefined_ps(), \
3590 0 + ((I) & 0x3) * 4, \
3591 1 + ((I) & 0x3) * 4, \
3592 2 + ((I) & 0x3) * 4, \
3593 3 + ((I) & 0x3) * 4); })
3595 #define _mm512_mask_extractf32x4_ps(W, U, A, imm) __extension__ ({\
3596 (__m128)__builtin_ia32_selectps_128((__mmask8)(U), \
3597 (__v4sf)_mm512_extractf32x4_ps((A), (imm)), \
3600 #define _mm512_maskz_extractf32x4_ps(U, A, imm) __extension__ ({\
3601 (__m128)__builtin_ia32_selectps_128((__mmask8)(U), \
3602 (__v4sf)_mm512_extractf32x4_ps((A), (imm)), \
3603 (__v4sf)_mm_setzero_ps()); })
3610 return (__m512d) __builtin_ia32_selectpd_512 ((__mmask8) __U,
3618 return (__m512) __builtin_ia32_selectps_512 ((__mmask16) __U,
3626 return (__m512i) __builtin_ia32_selectq_512 ((__mmask8) __U,
3634 return (__m512i) __builtin_ia32_selectd_512 ((__mmask16) __U,
3641 #define _mm512_cmp_round_ps_mask(A, B, P, R) __extension__ ({ \
3642 (__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3643 (__v16sf)(__m512)(B), (int)(P), \
3644 (__mmask16)-1, (int)(R)); })
3646 #define _mm512_mask_cmp_round_ps_mask(U, A, B, P, R) __extension__ ({ \
3647 (__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3648 (__v16sf)(__m512)(B), (int)(P), \
3649 (__mmask16)(U), (int)(R)); })
3651 #define _mm512_cmp_ps_mask(A, B, P) \
3652 _mm512_cmp_round_ps_mask((A), (B), (P), _MM_FROUND_CUR_DIRECTION)
3653 #define _mm512_mask_cmp_ps_mask(U, A, B, P) \
3654 _mm512_mask_cmp_round_ps_mask((U), (A), (B), (P), _MM_FROUND_CUR_DIRECTION)
3656 #define _mm512_cmpeq_ps_mask(A, B) \
3657 _mm512_cmp_ps_mask((A), (B), _CMP_EQ_OQ)
3658 #define _mm512_mask_cmpeq_ps_mask(k, A, B) \
3659 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_EQ_OQ)
3661 #define _mm512_cmplt_ps_mask(A, B) \
3662 _mm512_cmp_ps_mask((A), (B), _CMP_LT_OS)
3663 #define _mm512_mask_cmplt_ps_mask(k, A, B) \
3664 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_LT_OS)
3666 #define _mm512_cmple_ps_mask(A, B) \
3667 _mm512_cmp_ps_mask((A), (B), _CMP_LE_OS)
3668 #define _mm512_mask_cmple_ps_mask(k, A, B) \
3669 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_LE_OS)
3671 #define _mm512_cmpunord_ps_mask(A, B) \
3672 _mm512_cmp_ps_mask((A), (B), _CMP_UNORD_Q)
3673 #define _mm512_mask_cmpunord_ps_mask(k, A, B) \
3674 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_UNORD_Q)
3676 #define _mm512_cmpneq_ps_mask(A, B) \
3677 _mm512_cmp_ps_mask((A), (B), _CMP_NEQ_UQ)
3678 #define _mm512_mask_cmpneq_ps_mask(k, A, B) \
3679 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_NEQ_UQ)
3681 #define _mm512_cmpnlt_ps_mask(A, B) \
3682 _mm512_cmp_ps_mask((A), (B), _CMP_NLT_US)
3683 #define _mm512_mask_cmpnlt_ps_mask(k, A, B) \
3684 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_NLT_US)
3686 #define _mm512_cmpnle_ps_mask(A, B) \
3687 _mm512_cmp_ps_mask((A), (B), _CMP_NLE_US)
3688 #define _mm512_mask_cmpnle_ps_mask(k, A, B) \
3689 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_NLE_US)
3691 #define _mm512_cmpord_ps_mask(A, B) \
3692 _mm512_cmp_ps_mask((A), (B), _CMP_ORD_Q)
3693 #define _mm512_mask_cmpord_ps_mask(k, A, B) \
3694 _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_ORD_Q)
3696 #define _mm512_cmp_round_pd_mask(A, B, P, R) __extension__ ({ \
3697 (__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3698 (__v8df)(__m512d)(B), (int)(P), \
3699 (__mmask8)-1, (int)(R)); })
3701 #define _mm512_mask_cmp_round_pd_mask(U, A, B, P, R) __extension__ ({ \
3702 (__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3703 (__v8df)(__m512d)(B), (int)(P), \
3704 (__mmask8)(U), (int)(R)); })
3706 #define _mm512_cmp_pd_mask(A, B, P) \
3707 _mm512_cmp_round_pd_mask((A), (B), (P), _MM_FROUND_CUR_DIRECTION)
3708 #define _mm512_mask_cmp_pd_mask(U, A, B, P) \
3709 _mm512_mask_cmp_round_pd_mask((U), (A), (B), (P), _MM_FROUND_CUR_DIRECTION)
3711 #define _mm512_cmpeq_pd_mask(A, B) \
3712 _mm512_cmp_pd_mask((A), (B), _CMP_EQ_OQ)
3713 #define _mm512_mask_cmpeq_pd_mask(k, A, B) \
3714 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_EQ_OQ)
3716 #define _mm512_cmplt_pd_mask(A, B) \
3717 _mm512_cmp_pd_mask((A), (B), _CMP_LT_OS)
3718 #define _mm512_mask_cmplt_pd_mask(k, A, B) \
3719 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_LT_OS)
3721 #define _mm512_cmple_pd_mask(A, B) \
3722 _mm512_cmp_pd_mask((A), (B), _CMP_LE_OS)
3723 #define _mm512_mask_cmple_pd_mask(k, A, B) \
3724 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_LE_OS)
3726 #define _mm512_cmpunord_pd_mask(A, B) \
3727 _mm512_cmp_pd_mask((A), (B), _CMP_UNORD_Q)
3728 #define _mm512_mask_cmpunord_pd_mask(k, A, B) \
3729 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_UNORD_Q)
3731 #define _mm512_cmpneq_pd_mask(A, B) \
3732 _mm512_cmp_pd_mask((A), (B), _CMP_NEQ_UQ)
3733 #define _mm512_mask_cmpneq_pd_mask(k, A, B) \
3734 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_NEQ_UQ)
3736 #define _mm512_cmpnlt_pd_mask(A, B) \
3737 _mm512_cmp_pd_mask((A), (B), _CMP_NLT_US)
3738 #define _mm512_mask_cmpnlt_pd_mask(k, A, B) \
3739 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_NLT_US)
3741 #define _mm512_cmpnle_pd_mask(A, B) \
3742 _mm512_cmp_pd_mask((A), (B), _CMP_NLE_US)
3743 #define _mm512_mask_cmpnle_pd_mask(k, A, B) \
3744 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_NLE_US)
3746 #define _mm512_cmpord_pd_mask(A, B) \
3747 _mm512_cmp_pd_mask((A), (B), _CMP_ORD_Q)
3748 #define _mm512_mask_cmpord_pd_mask(k, A, B) \
3749 _mm512_mask_cmp_pd_mask((k), (A), (B), _CMP_ORD_Q)
3753 #define _mm512_cvtt_roundps_epu32(A, R) __extension__ ({ \
3754 (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3755 (__v16si)_mm512_undefined_epi32(), \
3756 (__mmask16)-1, (int)(R)); })
3758 #define _mm512_mask_cvtt_roundps_epu32(W, U, A, R) __extension__ ({ \
3759 (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3760 (__v16si)(__m512i)(W), \
3761 (__mmask16)(U), (int)(R)); })
3763 #define _mm512_maskz_cvtt_roundps_epu32(U, A, R) __extension__ ({ \
3764 (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3765 (__v16si)_mm512_setzero_si512(), \
3766 (__mmask16)(U), (int)(R)); })
3772 return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
3782 return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
3791 return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
3797 #define _mm512_cvt_roundepi32_ps(A, R) __extension__ ({ \
3798 (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3799 (__v16sf)_mm512_setzero_ps(), \
3800 (__mmask16)-1, (int)(R)); })
3802 #define _mm512_mask_cvt_roundepi32_ps(W, U, A, R) __extension__ ({ \
3803 (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3804 (__v16sf)(__m512)(W), \
3805 (__mmask16)(U), (int)(R)); })
3807 #define _mm512_maskz_cvt_roundepi32_ps(U, A, R) __extension__ ({ \
3808 (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3809 (__v16sf)_mm512_setzero_ps(), \
3810 (__mmask16)(U), (int)(R)); })
3812 #define _mm512_cvt_roundepu32_ps(A, R) __extension__ ({ \
3813 (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3814 (__v16sf)_mm512_setzero_ps(), \
3815 (__mmask16)-1, (int)(R)); })
3817 #define _mm512_mask_cvt_roundepu32_ps(W, U, A, R) __extension__ ({ \
3818 (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3819 (__v16sf)(__m512)(W), \
3820 (__mmask16)(U), (int)(R)); })
3822 #define _mm512_maskz_cvt_roundepu32_ps(U, A, R) __extension__ ({ \
3823 (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3824 (__v16sf)_mm512_setzero_ps(), \
3825 (__mmask16)(U), (int)(R)); })
3830 return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
3839 return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
3848 return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
3857 return (__m512d)__builtin_convertvector((__v8si)__A, __v8df);
3863 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
3871 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
3891 return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
3900 return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
3909 return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
3918 return (__m512d)__builtin_convertvector((__v8su)__A, __v8df);
3924 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
3932 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
3949 #define _mm512_cvt_roundpd_ps(A, R) __extension__ ({ \
3950 (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3951 (__v8sf)_mm256_setzero_ps(), \
3952 (__mmask8)-1, (int)(R)); })
3954 #define _mm512_mask_cvt_roundpd_ps(W, U, A, R) __extension__ ({ \
3955 (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3956 (__v8sf)(__m256)(W), (__mmask8)(U), \
3959 #define _mm512_maskz_cvt_roundpd_ps(U, A, R) __extension__ ({ \
3960 (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3961 (__v8sf)_mm256_setzero_ps(), \
3962 (__mmask8)(U), (int)(R)); })
3967 return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
3976 return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
3985 return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
3994 return (__m512) __builtin_shufflevector((__v8sf)
_mm512_cvtpd_ps(__A),
3996 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
4002 return (__m512) __builtin_shufflevector (
4006 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
4009 #define _mm512_cvt_roundps_ph(A, I) __extension__ ({ \
4010 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4011 (__v16hi)_mm256_undefined_si256(), \
4014 #define _mm512_mask_cvt_roundps_ph(U, W, A, I) __extension__ ({ \
4015 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4016 (__v16hi)(__m256i)(U), \
4019 #define _mm512_maskz_cvt_roundps_ph(W, A, I) __extension__ ({ \
4020 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4021 (__v16hi)_mm256_setzero_si256(), \
4024 #define _mm512_cvtps_ph(A, I) __extension__ ({ \
4025 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4026 (__v16hi)_mm256_setzero_si256(), \
4029 #define _mm512_mask_cvtps_ph(U, W, A, I) __extension__ ({ \
4030 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4031 (__v16hi)(__m256i)(U), \
4034 #define _mm512_maskz_cvtps_ph(W, A, I) __extension__ ({\
4035 (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
4036 (__v16hi)_mm256_setzero_si256(), \
4039 #define _mm512_cvt_roundph_ps(A, R) __extension__ ({ \
4040 (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
4041 (__v16sf)_mm512_undefined_ps(), \
4042 (__mmask16)-1, (int)(R)); })
4044 #define _mm512_mask_cvt_roundph_ps(W, U, A, R) __extension__ ({ \
4045 (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
4046 (__v16sf)(__m512)(W), \
4047 (__mmask16)(U), (int)(R)); })
4049 #define _mm512_maskz_cvt_roundph_ps(U, A, R) __extension__ ({ \
4050 (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
4051 (__v16sf)_mm512_setzero_ps(), \
4052 (__mmask16)(U), (int)(R)); })
4058 return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
4068 return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
4077 return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
4083 #define _mm512_cvtt_roundpd_epi32(A, R) __extension__ ({ \
4084 (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
4085 (__v8si)_mm256_setzero_si256(), \
4086 (__mmask8)-1, (int)(R)); })
4088 #define _mm512_mask_cvtt_roundpd_epi32(W, U, A, R) __extension__ ({ \
4089 (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
4090 (__v8si)(__m256i)(W), \
4091 (__mmask8)(U), (int)(R)); })
4093 #define _mm512_maskz_cvtt_roundpd_epi32(U, A, R) __extension__ ({ \
4094 (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
4095 (__v8si)_mm256_setzero_si256(), \
4096 (__mmask8)(U), (int)(R)); })
4101 return (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df) __a,
4110 return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
4119 return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
4125 #define _mm512_cvtt_roundps_epi32(A, R) __extension__ ({ \
4126 (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
4127 (__v16si)_mm512_setzero_si512(), \
4128 (__mmask16)-1, (int)(R)); })
4130 #define _mm512_mask_cvtt_roundps_epi32(W, U, A, R) __extension__ ({ \
4131 (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
4132 (__v16si)(__m512i)(W), \
4133 (__mmask16)(U), (int)(R)); })
4135 #define _mm512_maskz_cvtt_roundps_epi32(U, A, R) __extension__ ({ \
4136 (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
4137 (__v16si)_mm512_setzero_si512(), \
4138 (__mmask16)(U), (int)(R)); })
4144 __builtin_ia32_cvttps2dq512_mask((__v16sf) __a,
4152 return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
4161 return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
4167 #define _mm512_cvt_roundps_epi32(A, R) __extension__ ({ \
4168 (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
4169 (__v16si)_mm512_setzero_si512(), \
4170 (__mmask16)-1, (int)(R)); })
4172 #define _mm512_mask_cvt_roundps_epi32(W, U, A, R) __extension__ ({ \
4173 (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
4174 (__v16si)(__m512i)(W), \
4175 (__mmask16)(U), (int)(R)); })
4177 #define _mm512_maskz_cvt_roundps_epi32(U, A, R) __extension__ ({ \
4178 (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
4179 (__v16si)_mm512_setzero_si512(), \
4180 (__mmask16)(U), (int)(R)); })
4185 return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
4194 return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
4203 return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
4210 #define _mm512_cvt_roundpd_epi32(A, R) __extension__ ({ \
4211 (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
4212 (__v8si)_mm256_setzero_si256(), \
4213 (__mmask8)-1, (int)(R)); })
4215 #define _mm512_mask_cvt_roundpd_epi32(W, U, A, R) __extension__ ({ \
4216 (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
4217 (__v8si)(__m256i)(W), \
4218 (__mmask8)(U), (int)(R)); })
4220 #define _mm512_maskz_cvt_roundpd_epi32(U, A, R) __extension__ ({ \
4221 (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
4222 (__v8si)_mm256_setzero_si256(), \
4223 (__mmask8)(U), (int)(R)); })
4228 return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
4238 return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
4247 return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
4254 #define _mm512_cvt_roundps_epu32(A, R) __extension__ ({ \
4255 (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4256 (__v16si)_mm512_setzero_si512(), \
4257 (__mmask16)-1, (int)(R)); })
4259 #define _mm512_mask_cvt_roundps_epu32(W, U, A, R) __extension__ ({ \
4260 (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4261 (__v16si)(__m512i)(W), \
4262 (__mmask16)(U), (int)(R)); })
4264 #define _mm512_maskz_cvt_roundps_epu32(U, A, R) __extension__ ({ \
4265 (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4266 (__v16si)_mm512_setzero_si512(), \
4267 (__mmask16)(U), (int)(R)); })
4272 return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,\
4282 return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
4291 return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
4298 #define _mm512_cvt_roundpd_epu32(A, R) __extension__ ({ \
4299 (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4300 (__v8si)_mm256_setzero_si256(), \
4301 (__mmask8)-1, (int)(R)); })
4303 #define _mm512_mask_cvt_roundpd_epu32(W, U, A, R) __extension__ ({ \
4304 (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4306 (__mmask8)(U), (int)(R)); })
4308 #define _mm512_maskz_cvt_roundpd_epu32(U, A, R) __extension__ ({ \
4309 (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4310 (__v8si)_mm256_setzero_si256(), \
4311 (__mmask8)(U), (int)(R)); })
4316 return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
4326 return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
4335 return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
4359 return (__m512d)__builtin_shufflevector((__v8df)__a, (__v8df)__b,
4360 1, 9, 1+2, 9+2, 1+4, 9+4, 1+6, 9+6);
4366 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
4374 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
4382 return (__m512d)__builtin_shufflevector((__v8df)__a, (__v8df)__b,
4383 0, 8, 0+2, 8+2, 0+4, 8+4, 0+6, 8+6);
4389 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
4397 return (__m512d)__builtin_ia32_selectpd_512((__mmask8) __U,
4405 return (__m512)__builtin_shufflevector((__v16sf)__a, (__v16sf)__b,
4407 2+4, 18+4, 3+4, 19+4,
4408 2+8, 18+8, 3+8, 19+8,
4409 2+12, 18+12, 3+12, 19+12);
4415 return (__m512)__builtin_ia32_selectps_512((__mmask16) __U,
4423 return (__m512)__builtin_ia32_selectps_512((__mmask16) __U,
4431 return (__m512)__builtin_shufflevector((__v16sf)__a, (__v16sf)__b,
4433 0+4, 16+4, 1+4, 17+4,
4434 0+8, 16+8, 1+8, 17+8,
4435 0+12, 16+12, 1+12, 17+12);
4441 return (__m512)__builtin_ia32_selectps_512((__mmask16) __U,
4449 return (__m512)__builtin_ia32_selectps_512((__mmask16) __U,
4457 return (__m512i)__builtin_shufflevector((__v16si)__A, (__v16si)__B,
4459 2+4, 18+4, 3+4, 19+4,
4460 2+8, 18+8, 3+8, 19+8,
4461 2+12, 18+12, 3+12, 19+12);
4467 return (__m512i)__builtin_ia32_selectd_512((__mmask16) __U,
4475 return (__m512i)__builtin_ia32_selectd_512((__mmask16) __U,
4483 return (__m512i)__builtin_shufflevector((__v16si)__A, (__v16si)__B,
4485 0+4, 16+4, 1+4, 17+4,
4486 0+8, 16+8, 1+8, 17+8,
4487 0+12, 16+12, 1+12, 17+12);
4493 return (__m512i)__builtin_ia32_selectd_512((__mmask16) __U,
4501 return (__m512i)__builtin_ia32_selectd_512((__mmask16) __U,
4509 return (__m512i)__builtin_shufflevector((__v8di)__A, (__v8di)__B,
4510 1, 9, 1+2, 9+2, 1+4, 9+4, 1+6, 9+6);
4516 return (__m512i)__builtin_ia32_selectq_512((__mmask8) __U,
4524 return (__m512i)__builtin_ia32_selectq_512((__mmask8) __U,
4532 return (__m512i)__builtin_shufflevector((__v8di)__A, (__v8di)__B,
4533 0, 8, 0+2, 8+2, 0+4, 8+4, 0+6, 8+6);
4539 return (__m512i)__builtin_ia32_selectq_512((__mmask8) __U,
4547 return (__m512i)__builtin_ia32_selectq_512((__mmask8) __U,
4557 return (__mmask16) __builtin_ia32_ptestmd512 ((__v16si) __A,
4565 return (__mmask16) __builtin_ia32_ptestmd512 ((__v16si) __A,
4566 (__v16si) __B, __U);
4572 return (__mmask8) __builtin_ia32_ptestmq512 ((__v8di) __A,
4580 return (__mmask8) __builtin_ia32_ptestmq512 ((__v8di) __A, (__v8di) __B, __U);
4589 return (__m512i) __builtin_ia32_loaddqusi512_mask ((
const int *) __P,
4598 return (__m512i) __builtin_ia32_loaddqusi512_mask ((
const int *) __P,
4607 return (__m512i) __builtin_ia32_loaddqusi512_mask ((
const int *)__P,
4616 return (__m512i) __builtin_ia32_loaddqudi512_mask ((
const long long *) __P,
4624 return (__m512i) __builtin_ia32_loaddqudi512_mask ((
const long long *)__P,
4633 return (__m512) __builtin_ia32_loadups512_mask ((
const float *) __P,
4641 return (__m512) __builtin_ia32_loadups512_mask ((
const float *)__P,
4650 return (__m512d) __builtin_ia32_loadupd512_mask ((
const double *) __P,
4658 return (__m512d) __builtin_ia32_loadupd512_mask ((
const double *)__P,
4670 return ((
struct __loadu_pd*)__p)->__v;
4679 return ((
struct __loadu_ps*)__p)->__v;
4685 return (__m512) __builtin_ia32_loadaps512_mask ((
const __v16sf *)__p,
4694 return (__m512) __builtin_ia32_loadaps512_mask ((
const __v16sf *) __P,
4702 return (__m512) __builtin_ia32_loadaps512_mask ((
const __v16sf *)__P,
4711 return (__m512d) __builtin_ia32_loadapd512_mask ((
const __v8df *)__p,
4720 return (__m512d) __builtin_ia32_loadapd512_mask ((
const __v8df *) __P,
4728 return (__m512d) __builtin_ia32_loadapd512_mask ((
const __v8df *)__P,
4737 return *(__m512i *) __P;
4743 return *(__m512i *) __P;
4749 return *(__m512i *) __P;
4757 __builtin_ia32_storedqudi512_mask ((
long long *)__P, (__v8di) __A,
4764 __builtin_ia32_storedqusi512_mask ((
int *) __P, (__v16si) __A,
4771 __builtin_ia32_storedqusi512_mask ((
int *)__P, (__v16si) __A,
4778 __builtin_ia32_storeupd512_mask ((
double *)__P, (__v8df) __A, (__mmask8) __U);
4784 __builtin_ia32_storeupd512_mask((
double *)__P, (__v8df)__A, (__mmask8)-1);
4790 __builtin_ia32_storeups512_mask ((
float *)__P, (__v16sf) __A,
4797 __builtin_ia32_storeups512_mask((
float *)__P, (__v16sf)__A, (__mmask16)-1);
4803 __builtin_ia32_storeapd512_mask ((__v8df *)__P, (__v8df) __A, (__mmask8) __U);
4809 *(__m512d*)__P = __A;
4815 __builtin_ia32_storeaps512_mask ((__v16sf *)__P, (__v16sf) __A,
4822 *(__m512*)__P = __A;
4828 *(__m512i *) __P = __A;
4834 *(__m512i *) __P = __A;
4840 *(__m512i *) __P = __A;
4848 return __builtin_ia32_knothi(__M);
4855 return (__mmask16)__builtin_ia32_pcmpeqd512_mask((__v16si)__a, (__v16si)__b,
4861 return (__mmask16)__builtin_ia32_pcmpeqd512_mask((__v16si)__a, (__v16si)__b,
4867 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 0,
4873 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 0,
4879 return (__mmask8)__builtin_ia32_pcmpeqq512_mask((__v8di)__a, (__v8di)__b,
4885 return (__mmask8)__builtin_ia32_pcmpeqq512_mask((__v8di)__a, (__v8di)__b,
4891 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 0,
4897 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 0,
4903 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 5,
4909 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 5,
4915 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 5,
4921 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 5,
4927 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 5,
4933 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 5,
4939 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 5,
4945 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 5,
4951 return (__mmask16)__builtin_ia32_pcmpgtd512_mask((__v16si)__a, (__v16si)__b,
4957 return (__mmask16)__builtin_ia32_pcmpgtd512_mask((__v16si)__a, (__v16si)__b,
4963 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 6,
4969 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 6,
4975 return (__mmask8)__builtin_ia32_pcmpgtq512_mask((__v8di)__a, (__v8di)__b,
4981 return (__mmask8)__builtin_ia32_pcmpgtq512_mask((__v8di)__a, (__v8di)__b,
4987 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 6,
4993 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 6,
4999 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 2,
5005 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 2,
5011 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 2,
5017 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 2,
5023 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 2,
5029 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 2,
5035 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 2,
5041 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 2,
5047 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 1,
5053 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 1,
5059 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 1,
5065 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 1,
5071 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 1,
5077 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 1,
5083 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 1,
5089 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 1,
5095 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 4,
5101 return (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)__a, (__v16si)__b, 4,
5107 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 4,
5113 return (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)__a, (__v16si)__b, 4,
5119 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 4,
5125 return (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)__a, (__v8di)__b, 4,
5131 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 4,
5137 return (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)__a, (__v8di)__b, 4,
5146 return (__m512i)__builtin_convertvector((__v16qs)__A, __v16si);
5152 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5160 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5170 return (__m512i)__builtin_convertvector(__builtin_shufflevector((__v16qs)__A, (__v16qs)__A, 0, 1, 2, 3, 4, 5, 6, 7), __v8di);
5176 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5184 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5192 return (__m512i)__builtin_convertvector((__v8si)__X, __v8di);
5198 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5206 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5214 return (__m512i)__builtin_convertvector((__v16hi)__A, __v16si);
5220 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5228 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5236 return (__m512i)__builtin_convertvector((__v8hi)__A, __v8di);
5242 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5250 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5258 return (__m512i)__builtin_convertvector((__v16qu)__A, __v16si);
5264 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5272 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5280 return (__m512i)__builtin_convertvector(__builtin_shufflevector((__v16qu)__A, (__v16qu)__A, 0, 1, 2, 3, 4, 5, 6, 7), __v8di);
5286 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5294 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5302 return (__m512i)__builtin_convertvector((__v8su)__X, __v8di);
5308 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5316 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5324 return (__m512i)__builtin_convertvector((__v16hu)__A, __v16si);
5330 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5338 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5346 return (__m512i)__builtin_convertvector((__v8hu)__A, __v8di);
5352 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5360 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5368 return (__m512i) __builtin_ia32_prorvd512_mask ((__v16si) __A,
5378 return (__m512i) __builtin_ia32_prorvd512_mask ((__v16si) __A,
5387 return (__m512i) __builtin_ia32_prorvd512_mask ((__v16si) __A,
5397 return (__m512i) __builtin_ia32_prorvq512_mask ((__v8di) __A,
5407 return (__m512i) __builtin_ia32_prorvq512_mask ((__v8di) __A,
5416 return (__m512i) __builtin_ia32_prorvq512_mask ((__v8di) __A,
5425 #define _mm512_cmp_epi32_mask(a, b, p) __extension__ ({ \
5426 (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
5427 (__v16si)(__m512i)(b), (int)(p), \
5430 #define _mm512_cmp_epu32_mask(a, b, p) __extension__ ({ \
5431 (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
5432 (__v16si)(__m512i)(b), (int)(p), \
5435 #define _mm512_cmp_epi64_mask(a, b, p) __extension__ ({ \
5436 (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
5437 (__v8di)(__m512i)(b), (int)(p), \
5440 #define _mm512_cmp_epu64_mask(a, b, p) __extension__ ({ \
5441 (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
5442 (__v8di)(__m512i)(b), (int)(p), \
5445 #define _mm512_mask_cmp_epi32_mask(m, a, b, p) __extension__ ({ \
5446 (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
5447 (__v16si)(__m512i)(b), (int)(p), \
5450 #define _mm512_mask_cmp_epu32_mask(m, a, b, p) __extension__ ({ \
5451 (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
5452 (__v16si)(__m512i)(b), (int)(p), \
5455 #define _mm512_mask_cmp_epi64_mask(m, a, b, p) __extension__ ({ \
5456 (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
5457 (__v8di)(__m512i)(b), (int)(p), \
5460 #define _mm512_mask_cmp_epu64_mask(m, a, b, p) __extension__ ({ \
5461 (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
5462 (__v8di)(__m512i)(b), (int)(p), \
5465 #define _mm512_rol_epi32(a, b) __extension__ ({ \
5466 (__m512i)__builtin_ia32_prold512_mask((__v16si)(__m512i)(a), (int)(b), \
5467 (__v16si)_mm512_setzero_si512(), \
5470 #define _mm512_mask_rol_epi32(W, U, a, b) __extension__ ({ \
5471 (__m512i)__builtin_ia32_prold512_mask((__v16si)(__m512i)(a), (int)(b), \
5472 (__v16si)(__m512i)(W), \
5475 #define _mm512_maskz_rol_epi32(U, a, b) __extension__ ({ \
5476 (__m512i)__builtin_ia32_prold512_mask((__v16si)(__m512i)(a), (int)(b), \
5477 (__v16si)_mm512_setzero_si512(), \
5480 #define _mm512_rol_epi64(a, b) __extension__ ({ \
5481 (__m512i)__builtin_ia32_prolq512_mask((__v8di)(__m512i)(a), (int)(b), \
5482 (__v8di)_mm512_setzero_si512(), \
5485 #define _mm512_mask_rol_epi64(W, U, a, b) __extension__ ({ \
5486 (__m512i)__builtin_ia32_prolq512_mask((__v8di)(__m512i)(a), (int)(b), \
5487 (__v8di)(__m512i)(W), (__mmask8)(U)); })
5489 #define _mm512_maskz_rol_epi64(U, a, b) __extension__ ({ \
5490 (__m512i)__builtin_ia32_prolq512_mask((__v8di)(__m512i)(a), (int)(b), \
5491 (__v8di)_mm512_setzero_si512(), \
5496 return (__m512i) __builtin_ia32_prolvd512_mask ((__v16si) __A,
5506 return (__m512i) __builtin_ia32_prolvd512_mask ((__v16si) __A,
5515 return (__m512i) __builtin_ia32_prolvd512_mask ((__v16si) __A,
5525 return (__m512i) __builtin_ia32_prolvq512_mask ((__v8di) __A,
5535 return (__m512i) __builtin_ia32_prolvq512_mask ((__v8di) __A,
5544 return (__m512i) __builtin_ia32_prolvq512_mask ((__v8di) __A,
5551 #define _mm512_ror_epi32(A, B) __extension__ ({ \
5552 (__m512i)__builtin_ia32_prord512_mask((__v16si)(__m512i)(A), (int)(B), \
5553 (__v16si)_mm512_setzero_si512(), \
5556 #define _mm512_mask_ror_epi32(W, U, A, B) __extension__ ({ \
5557 (__m512i)__builtin_ia32_prord512_mask((__v16si)(__m512i)(A), (int)(B), \
5558 (__v16si)(__m512i)(W), \
5561 #define _mm512_maskz_ror_epi32(U, A, B) __extension__ ({ \
5562 (__m512i)__builtin_ia32_prord512_mask((__v16si)(__m512i)(A), (int)(B), \
5563 (__v16si)_mm512_setzero_si512(), \
5566 #define _mm512_ror_epi64(A, B) __extension__ ({ \
5567 (__m512i)__builtin_ia32_prorq512_mask((__v8di)(__m512i)(A), (int)(B), \
5568 (__v8di)_mm512_setzero_si512(), \
5571 #define _mm512_mask_ror_epi64(W, U, A, B) __extension__ ({ \
5572 (__m512i)__builtin_ia32_prorq512_mask((__v8di)(__m512i)(A), (int)(B), \
5573 (__v8di)(__m512i)(W), (__mmask8)(U)); })
5575 #define _mm512_maskz_ror_epi64(U, A, B) __extension__ ({ \
5576 (__m512i)__builtin_ia32_prorq512_mask((__v8di)(__m512i)(A), (int)(B), \
5577 (__v8di)_mm512_setzero_si512(), \
5583 return (__m512i)__builtin_ia32_pslldi512((__v16si)__A, __B);
5589 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5596 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5604 return (__m512i)__builtin_ia32_psllqi512((__v8di)__A, __B);
5610 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5618 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5626 return (__m512i)__builtin_ia32_psrldi512((__v16si)__A, __B);
5632 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5639 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
5647 return (__m512i)__builtin_ia32_psrlqi512((__v8di)__A, __B);
5653 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5661 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
5669 return (__m512i) __builtin_ia32_movdqa32load512_mask ((
const __v16si *) __P,
5677 return (__m512i) __builtin_ia32_movdqa32load512_mask ((
const __v16si *) __P,
5686 __builtin_ia32_movdqa32store512_mask ((__v16si *) __P, (__v16si) __A,
5693 return (__m512i) __builtin_ia32_selectd_512 ((__mmask16) __U,
5701 return (__m512i) __builtin_ia32_selectd_512 ((__mmask16) __U,
5709 return (__m512i) __builtin_ia32_selectq_512 ((__mmask8) __U,
5717 return (__m512i) __builtin_ia32_selectq_512 ((__mmask8) __U,
5725 return (__m512i) __builtin_ia32_movdqa64load512_mask ((
const __v8di *) __P,
5733 return (__m512i) __builtin_ia32_movdqa64load512_mask ((
const __v8di *) __P,
5742 __builtin_ia32_movdqa64store512_mask ((__v8di *) __P, (__v8di) __A,
5749 return (__m512d)__builtin_shufflevector((__v8df)__A, (__v8df)__A,
5750 0, 0, 2, 2, 4, 4, 6, 6);
5756 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
5764 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
5769 #define _mm512_fixupimm_round_pd(A, B, C, imm, R) __extension__ ({ \
5770 (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5771 (__v8df)(__m512d)(B), \
5772 (__v8di)(__m512i)(C), (int)(imm), \
5773 (__mmask8)-1, (int)(R)); })
5775 #define _mm512_mask_fixupimm_round_pd(A, U, B, C, imm, R) __extension__ ({ \
5776 (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5777 (__v8df)(__m512d)(B), \
5778 (__v8di)(__m512i)(C), (int)(imm), \
5779 (__mmask8)(U), (int)(R)); })
5781 #define _mm512_fixupimm_pd(A, B, C, imm) __extension__ ({ \
5782 (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5783 (__v8df)(__m512d)(B), \
5784 (__v8di)(__m512i)(C), (int)(imm), \
5786 _MM_FROUND_CUR_DIRECTION); })
5788 #define _mm512_mask_fixupimm_pd(A, U, B, C, imm) __extension__ ({ \
5789 (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5790 (__v8df)(__m512d)(B), \
5791 (__v8di)(__m512i)(C), (int)(imm), \
5793 _MM_FROUND_CUR_DIRECTION); })
5795 #define _mm512_maskz_fixupimm_round_pd(U, A, B, C, imm, R) __extension__ ({ \
5796 (__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5797 (__v8df)(__m512d)(B), \
5798 (__v8di)(__m512i)(C), \
5799 (int)(imm), (__mmask8)(U), \
5802 #define _mm512_maskz_fixupimm_pd(U, A, B, C, imm) __extension__ ({ \
5803 (__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5804 (__v8df)(__m512d)(B), \
5805 (__v8di)(__m512i)(C), \
5806 (int)(imm), (__mmask8)(U), \
5807 _MM_FROUND_CUR_DIRECTION); })
5809 #define _mm512_fixupimm_round_ps(A, B, C, imm, R) __extension__ ({ \
5810 (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5811 (__v16sf)(__m512)(B), \
5812 (__v16si)(__m512i)(C), (int)(imm), \
5813 (__mmask16)-1, (int)(R)); })
5815 #define _mm512_mask_fixupimm_round_ps(A, U, B, C, imm, R) __extension__ ({ \
5816 (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5817 (__v16sf)(__m512)(B), \
5818 (__v16si)(__m512i)(C), (int)(imm), \
5819 (__mmask16)(U), (int)(R)); })
5821 #define _mm512_fixupimm_ps(A, B, C, imm) __extension__ ({ \
5822 (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5823 (__v16sf)(__m512)(B), \
5824 (__v16si)(__m512i)(C), (int)(imm), \
5826 _MM_FROUND_CUR_DIRECTION); })
5828 #define _mm512_mask_fixupimm_ps(A, U, B, C, imm) __extension__ ({ \
5829 (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5830 (__v16sf)(__m512)(B), \
5831 (__v16si)(__m512i)(C), (int)(imm), \
5833 _MM_FROUND_CUR_DIRECTION); })
5835 #define _mm512_maskz_fixupimm_round_ps(U, A, B, C, imm, R) __extension__ ({ \
5836 (__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5837 (__v16sf)(__m512)(B), \
5838 (__v16si)(__m512i)(C), \
5839 (int)(imm), (__mmask16)(U), \
5842 #define _mm512_maskz_fixupimm_ps(U, A, B, C, imm) __extension__ ({ \
5843 (__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5844 (__v16sf)(__m512)(B), \
5845 (__v16si)(__m512i)(C), \
5846 (int)(imm), (__mmask16)(U), \
5847 _MM_FROUND_CUR_DIRECTION); })
5849 #define _mm_fixupimm_round_sd(A, B, C, imm, R) __extension__ ({ \
5850 (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5851 (__v2df)(__m128d)(B), \
5852 (__v2di)(__m128i)(C), (int)(imm), \
5853 (__mmask8)-1, (int)(R)); })
5855 #define _mm_mask_fixupimm_round_sd(A, U, B, C, imm, R) __extension__ ({ \
5856 (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5857 (__v2df)(__m128d)(B), \
5858 (__v2di)(__m128i)(C), (int)(imm), \
5859 (__mmask8)(U), (int)(R)); })
5861 #define _mm_fixupimm_sd(A, B, C, imm) __extension__ ({ \
5862 (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5863 (__v2df)(__m128d)(B), \
5864 (__v2di)(__m128i)(C), (int)(imm), \
5866 _MM_FROUND_CUR_DIRECTION); })
5868 #define _mm_mask_fixupimm_sd(A, U, B, C, imm) __extension__ ({ \
5869 (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5870 (__v2df)(__m128d)(B), \
5871 (__v2di)(__m128i)(C), (int)(imm), \
5873 _MM_FROUND_CUR_DIRECTION); })
5875 #define _mm_maskz_fixupimm_round_sd(U, A, B, C, imm, R) __extension__ ({ \
5876 (__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5877 (__v2df)(__m128d)(B), \
5878 (__v2di)(__m128i)(C), (int)(imm), \
5879 (__mmask8)(U), (int)(R)); })
5881 #define _mm_maskz_fixupimm_sd(U, A, B, C, imm) __extension__ ({ \
5882 (__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5883 (__v2df)(__m128d)(B), \
5884 (__v2di)(__m128i)(C), (int)(imm), \
5886 _MM_FROUND_CUR_DIRECTION); })
5888 #define _mm_fixupimm_round_ss(A, B, C, imm, R) __extension__ ({ \
5889 (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5890 (__v4sf)(__m128)(B), \
5891 (__v4si)(__m128i)(C), (int)(imm), \
5892 (__mmask8)-1, (int)(R)); })
5894 #define _mm_mask_fixupimm_round_ss(A, U, B, C, imm, R) __extension__ ({ \
5895 (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5896 (__v4sf)(__m128)(B), \
5897 (__v4si)(__m128i)(C), (int)(imm), \
5898 (__mmask8)(U), (int)(R)); })
5900 #define _mm_fixupimm_ss(A, B, C, imm) __extension__ ({ \
5901 (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5902 (__v4sf)(__m128)(B), \
5903 (__v4si)(__m128i)(C), (int)(imm), \
5905 _MM_FROUND_CUR_DIRECTION); })
5907 #define _mm_mask_fixupimm_ss(A, U, B, C, imm) __extension__ ({ \
5908 (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5909 (__v4sf)(__m128)(B), \
5910 (__v4si)(__m128i)(C), (int)(imm), \
5912 _MM_FROUND_CUR_DIRECTION); })
5914 #define _mm_maskz_fixupimm_round_ss(U, A, B, C, imm, R) __extension__ ({ \
5915 (__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5916 (__v4sf)(__m128)(B), \
5917 (__v4si)(__m128i)(C), (int)(imm), \
5918 (__mmask8)(U), (int)(R)); })
5920 #define _mm_maskz_fixupimm_ss(U, A, B, C, imm) __extension__ ({ \
5921 (__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5922 (__v4sf)(__m128)(B), \
5923 (__v4si)(__m128i)(C), (int)(imm), \
5925 _MM_FROUND_CUR_DIRECTION); })
5927 #define _mm_getexp_round_sd(A, B, R) __extension__ ({ \
5928 (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5929 (__v2df)(__m128d)(B), \
5930 (__v2df)_mm_setzero_pd(), \
5931 (__mmask8)-1, (int)(R)); })
5937 return (__m128d) __builtin_ia32_getexpsd128_round_mask ((__v2df) __A,
5944 return (__m128d) __builtin_ia32_getexpsd128_round_mask ( (__v2df) __A,
5951 #define _mm_mask_getexp_round_sd(W, U, A, B, R) __extension__ ({\
5952 (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5953 (__v2df)(__m128d)(B), \
5954 (__v2df)(__m128d)(W), \
5955 (__mmask8)(U), (int)(R)); })
5960 return (__m128d) __builtin_ia32_getexpsd128_round_mask ( (__v2df) __A,
5967 #define _mm_maskz_getexp_round_sd(U, A, B, R) __extension__ ({\
5968 (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5969 (__v2df)(__m128d)(B), \
5970 (__v2df)_mm_setzero_pd(), \
5971 (__mmask8)(U), (int)(R)); })
5973 #define _mm_getexp_round_ss(A, B, R) __extension__ ({ \
5974 (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5975 (__v4sf)(__m128)(B), \
5976 (__v4sf)_mm_setzero_ps(), \
5977 (__mmask8)-1, (int)(R)); })
5982 return (__m128) __builtin_ia32_getexpss128_round_mask ((__v4sf) __A,
5989 return (__m128) __builtin_ia32_getexpss128_round_mask ((__v4sf) __A,
5996 #define _mm_mask_getexp_round_ss(W, U, A, B, R) __extension__ ({\
5997 (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5998 (__v4sf)(__m128)(B), \
5999 (__v4sf)(__m128)(W), \
6000 (__mmask8)(U), (int)(R)); })
6005 return (__m128) __builtin_ia32_getexpss128_round_mask ((__v4sf) __A,
6012 #define _mm_maskz_getexp_round_ss(U, A, B, R) __extension__ ({\
6013 (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
6014 (__v4sf)(__m128)(B), \
6015 (__v4sf)_mm_setzero_ps(), \
6016 (__mmask8)(U), (int)(R)); })
6018 #define _mm_getmant_round_sd(A, B, C, D, R) __extension__ ({ \
6019 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6020 (__v2df)(__m128d)(B), \
6021 (int)(((D)<<2) | (C)), \
6022 (__v2df)_mm_setzero_pd(), \
6023 (__mmask8)-1, (int)(R)); })
6025 #define _mm_getmant_sd(A, B, C, D) __extension__ ({ \
6026 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6027 (__v2df)(__m128d)(B), \
6028 (int)(((D)<<2) | (C)), \
6029 (__v2df)_mm_setzero_pd(), \
6031 _MM_FROUND_CUR_DIRECTION); })
6033 #define _mm_mask_getmant_sd(W, U, A, B, C, D) __extension__ ({\
6034 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6035 (__v2df)(__m128d)(B), \
6036 (int)(((D)<<2) | (C)), \
6037 (__v2df)(__m128d)(W), \
6039 _MM_FROUND_CUR_DIRECTION); })
6041 #define _mm_mask_getmant_round_sd(W, U, A, B, C, D, R)({\
6042 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6043 (__v2df)(__m128d)(B), \
6044 (int)(((D)<<2) | (C)), \
6045 (__v2df)(__m128d)(W), \
6046 (__mmask8)(U), (int)(R)); })
6048 #define _mm_maskz_getmant_sd(U, A, B, C, D) __extension__ ({\
6049 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6050 (__v2df)(__m128d)(B), \
6051 (int)(((D)<<2) | (C)), \
6052 (__v2df)_mm_setzero_pd(), \
6054 _MM_FROUND_CUR_DIRECTION); })
6056 #define _mm_maskz_getmant_round_sd(U, A, B, C, D, R) __extension__ ({\
6057 (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
6058 (__v2df)(__m128d)(B), \
6059 (int)(((D)<<2) | (C)), \
6060 (__v2df)_mm_setzero_pd(), \
6061 (__mmask8)(U), (int)(R)); })
6063 #define _mm_getmant_round_ss(A, B, C, D, R) __extension__ ({ \
6064 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6065 (__v4sf)(__m128)(B), \
6066 (int)(((D)<<2) | (C)), \
6067 (__v4sf)_mm_setzero_ps(), \
6068 (__mmask8)-1, (int)(R)); })
6070 #define _mm_getmant_ss(A, B, C, D) __extension__ ({ \
6071 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6072 (__v4sf)(__m128)(B), \
6073 (int)(((D)<<2) | (C)), \
6074 (__v4sf)_mm_setzero_ps(), \
6076 _MM_FROUND_CUR_DIRECTION); })
6078 #define _mm_mask_getmant_ss(W, U, A, B, C, D) __extension__ ({\
6079 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6080 (__v4sf)(__m128)(B), \
6081 (int)(((D)<<2) | (C)), \
6082 (__v4sf)(__m128)(W), \
6084 _MM_FROUND_CUR_DIRECTION); })
6086 #define _mm_mask_getmant_round_ss(W, U, A, B, C, D, R)({\
6087 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6088 (__v4sf)(__m128)(B), \
6089 (int)(((D)<<2) | (C)), \
6090 (__v4sf)(__m128)(W), \
6091 (__mmask8)(U), (int)(R)); })
6093 #define _mm_maskz_getmant_ss(U, A, B, C, D) __extension__ ({\
6094 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6095 (__v4sf)(__m128)(B), \
6096 (int)(((D)<<2) | (C)), \
6097 (__v4sf)_mm_setzero_pd(), \
6099 _MM_FROUND_CUR_DIRECTION); })
6101 #define _mm_maskz_getmant_round_ss(U, A, B, C, D, R) __extension__ ({\
6102 (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
6103 (__v4sf)(__m128)(B), \
6104 (int)(((D)<<2) | (C)), \
6105 (__v4sf)_mm_setzero_ps(), \
6106 (__mmask8)(U), (int)(R)); })
6114 #define _mm_comi_round_sd(A, B, P, R) __extension__ ({\
6115 (int)__builtin_ia32_vcomisd((__v2df)(__m128d)(A), (__v2df)(__m128d)(B), \
6116 (int)(P), (int)(R)); })
6118 #define _mm_comi_round_ss(A, B, P, R) __extension__ ({\
6119 (int)__builtin_ia32_vcomiss((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), \
6120 (int)(P), (int)(R)); })
6123 #define _mm_cvt_roundsd_si64(A, R) __extension__ ({ \
6124 (long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R)); })
6129 __mmask16 __U, __m512i __B)
6131 return (__m512i) __builtin_ia32_vpermi2vard512_mask ((__v16si) __A,
6141 return (__m512i)__builtin_ia32_pslld512((__v16si) __A, (__v4si)__B);
6147 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6155 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6163 return (__m512i)__builtin_ia32_psllq512((__v8di)__A, (__v2di)__B);
6169 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6177 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6185 return (__m512i)__builtin_ia32_psllv16si((__v16si)__X, (__v16si)__Y);
6191 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6199 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6207 return (__m512i)__builtin_ia32_psllv8di((__v8di)__X, (__v8di)__Y);
6213 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6221 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6229 return (__m512i)__builtin_ia32_psrad512((__v16si) __A, (__v4si)__B);
6235 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6243 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6251 return (__m512i)__builtin_ia32_psraq512((__v8di)__A, (__v2di)__B);
6257 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6265 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6273 return (__m512i)__builtin_ia32_psrav16si((__v16si)__X, (__v16si)__Y);
6279 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6287 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6295 return (__m512i)__builtin_ia32_psrav8di((__v8di)__X, (__v8di)__Y);
6301 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6309 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6317 return (__m512i)__builtin_ia32_psrld512((__v16si) __A, (__v4si)__B);
6323 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6331 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6339 return (__m512i)__builtin_ia32_psrlq512((__v8di)__A, (__v2di)__B);
6345 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6353 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6361 return (__m512i)__builtin_ia32_psrlv16si((__v16si)__X, (__v16si)__Y);
6367 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6375 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U,
6383 return (__m512i)__builtin_ia32_psrlv8di((__v8di)__X, (__v8di)__Y);
6389 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6397 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U,
6402 #define _mm512_ternarylogic_epi32(A, B, C, imm) __extension__ ({ \
6403 (__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
6404 (__v16si)(__m512i)(B), \
6405 (__v16si)(__m512i)(C), (int)(imm), \
6408 #define _mm512_mask_ternarylogic_epi32(A, U, B, C, imm) __extension__ ({ \
6409 (__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
6410 (__v16si)(__m512i)(B), \
6411 (__v16si)(__m512i)(C), (int)(imm), \
6414 #define _mm512_maskz_ternarylogic_epi32(U, A, B, C, imm) __extension__ ({ \
6415 (__m512i)__builtin_ia32_pternlogd512_maskz((__v16si)(__m512i)(A), \
6416 (__v16si)(__m512i)(B), \
6417 (__v16si)(__m512i)(C), \
6418 (int)(imm), (__mmask16)(U)); })
6420 #define _mm512_ternarylogic_epi64(A, B, C, imm) __extension__ ({ \
6421 (__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
6422 (__v8di)(__m512i)(B), \
6423 (__v8di)(__m512i)(C), (int)(imm), \
6426 #define _mm512_mask_ternarylogic_epi64(A, U, B, C, imm) __extension__ ({ \
6427 (__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
6428 (__v8di)(__m512i)(B), \
6429 (__v8di)(__m512i)(C), (int)(imm), \
6432 #define _mm512_maskz_ternarylogic_epi64(U, A, B, C, imm) __extension__ ({ \
6433 (__m512i)__builtin_ia32_pternlogq512_maskz((__v8di)(__m512i)(A), \
6434 (__v8di)(__m512i)(B), \
6435 (__v8di)(__m512i)(C), (int)(imm), \
6439 #define _mm_cvt_roundsd_i64(A, R) __extension__ ({ \
6440 (long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R)); })
6443 #define _mm_cvt_roundsd_si32(A, R) __extension__ ({ \
6444 (int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R)); })
6446 #define _mm_cvt_roundsd_i32(A, R) __extension__ ({ \
6447 (int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R)); })
6449 #define _mm_cvt_roundsd_u32(A, R) __extension__ ({ \
6450 (unsigned int)__builtin_ia32_vcvtsd2usi32((__v2df)(__m128d)(A), (int)(R)); })
6455 return (
unsigned) __builtin_ia32_vcvtsd2usi32 ((__v2df) __A,
6460 #define _mm_cvt_roundsd_u64(A, R) __extension__ ({ \
6461 (unsigned long long)__builtin_ia32_vcvtsd2usi64((__v2df)(__m128d)(A), \
6465 _mm_cvtsd_u64 (__m128d __A)
6467 return (
unsigned long long) __builtin_ia32_vcvtsd2usi64 ((__v2df)
6473 #define _mm_cvt_roundss_si32(A, R) __extension__ ({ \
6474 (int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R)); })
6476 #define _mm_cvt_roundss_i32(A, R) __extension__ ({ \
6477 (int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R)); })
6480 #define _mm_cvt_roundss_si64(A, R) __extension__ ({ \
6481 (long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R)); })
6483 #define _mm_cvt_roundss_i64(A, R) __extension__ ({ \
6484 (long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R)); })
6487 #define _mm_cvt_roundss_u32(A, R) __extension__ ({ \
6488 (unsigned int)__builtin_ia32_vcvtss2usi32((__v4sf)(__m128)(A), (int)(R)); })
6493 return (
unsigned) __builtin_ia32_vcvtss2usi32 ((__v4sf) __A,
6498 #define _mm_cvt_roundss_u64(A, R) __extension__ ({ \
6499 (unsigned long long)__builtin_ia32_vcvtss2usi64((__v4sf)(__m128)(A), \
6503 _mm_cvtss_u64 (__m128 __A)
6505 return (
unsigned long long) __builtin_ia32_vcvtss2usi64 ((__v4sf)
6511 #define _mm_cvtt_roundsd_i32(A, R) __extension__ ({ \
6512 (int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R)); })
6514 #define _mm_cvtt_roundsd_si32(A, R) __extension__ ({ \
6515 (int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R)); })
6520 return (
int) __builtin_ia32_vcvttsd2si32 ((__v2df) __A,
6525 #define _mm_cvtt_roundsd_si64(A, R) __extension__ ({ \
6526 (long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R)); })
6528 #define _mm_cvtt_roundsd_i64(A, R) __extension__ ({ \
6529 (long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R)); })
6532 _mm_cvttsd_i64 (__m128d __A)
6534 return (
long long) __builtin_ia32_vcvttsd2si64 ((__v2df) __A,
6539 #define _mm_cvtt_roundsd_u32(A, R) __extension__ ({ \
6540 (unsigned int)__builtin_ia32_vcvttsd2usi32((__v2df)(__m128d)(A), (int)(R)); })
6545 return (
unsigned) __builtin_ia32_vcvttsd2usi32 ((__v2df) __A,
6550 #define _mm_cvtt_roundsd_u64(A, R) __extension__ ({ \
6551 (unsigned long long)__builtin_ia32_vcvttsd2usi64((__v2df)(__m128d)(A), \
6555 _mm_cvttsd_u64 (__m128d __A)
6557 return (
unsigned long long) __builtin_ia32_vcvttsd2usi64 ((__v2df)
6563 #define _mm_cvtt_roundss_i32(A, R) __extension__ ({ \
6564 (int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R)); })
6566 #define _mm_cvtt_roundss_si32(A, R) __extension__ ({ \
6567 (int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R)); })
6572 return (
int) __builtin_ia32_vcvttss2si32 ((__v4sf) __A,
6577 #define _mm_cvtt_roundss_i64(A, R) __extension__ ({ \
6578 (long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R)); })
6580 #define _mm_cvtt_roundss_si64(A, R) __extension__ ({ \
6581 (long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R)); })
6584 _mm_cvttss_i64 (__m128 __A)
6586 return (
long long) __builtin_ia32_vcvttss2si64 ((__v4sf) __A,
6591 #define _mm_cvtt_roundss_u32(A, R) __extension__ ({ \
6592 (unsigned int)__builtin_ia32_vcvttss2usi32((__v4sf)(__m128)(A), (int)(R)); })
6597 return (
unsigned) __builtin_ia32_vcvttss2usi32 ((__v4sf) __A,
6602 #define _mm_cvtt_roundss_u64(A, R) __extension__ ({ \
6603 (unsigned long long)__builtin_ia32_vcvttss2usi64((__v4sf)(__m128)(A), \
6607 _mm_cvttss_u64 (__m128 __A)
6609 return (
unsigned long long) __builtin_ia32_vcvttss2usi64 ((__v4sf)
6619 return (__m512d) __builtin_ia32_vpermi2varpd512_mask ((__v8df) __A,
6630 return (__m512) __builtin_ia32_vpermi2varps512_mask ((__v16sf) __A,
6639 __mmask8 __U, __m512i __B)
6641 return (__m512i) __builtin_ia32_vpermi2varq512_mask ((__v8di) __A,
6648 #define _mm512_permute_pd(X, C) __extension__ ({ \
6649 (__m512d)__builtin_shufflevector((__v8df)(__m512d)(X), \
6650 (__v8df)_mm512_undefined_pd(), \
6651 0 + (((C) >> 0) & 0x1), \
6652 0 + (((C) >> 1) & 0x1), \
6653 2 + (((C) >> 2) & 0x1), \
6654 2 + (((C) >> 3) & 0x1), \
6655 4 + (((C) >> 4) & 0x1), \
6656 4 + (((C) >> 5) & 0x1), \
6657 6 + (((C) >> 6) & 0x1), \
6658 6 + (((C) >> 7) & 0x1)); })
6660 #define _mm512_mask_permute_pd(W, U, X, C) __extension__ ({ \
6661 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6662 (__v8df)_mm512_permute_pd((X), (C)), \
6663 (__v8df)(__m512d)(W)); })
6665 #define _mm512_maskz_permute_pd(U, X, C) __extension__ ({ \
6666 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6667 (__v8df)_mm512_permute_pd((X), (C)), \
6668 (__v8df)_mm512_setzero_pd()); })
6670 #define _mm512_permute_ps(X, C) __extension__ ({ \
6671 (__m512)__builtin_shufflevector((__v16sf)(__m512)(X), \
6672 (__v16sf)_mm512_undefined_ps(), \
6673 0 + (((C) >> 0) & 0x3), \
6674 0 + (((C) >> 2) & 0x3), \
6675 0 + (((C) >> 4) & 0x3), \
6676 0 + (((C) >> 6) & 0x3), \
6677 4 + (((C) >> 0) & 0x3), \
6678 4 + (((C) >> 2) & 0x3), \
6679 4 + (((C) >> 4) & 0x3), \
6680 4 + (((C) >> 6) & 0x3), \
6681 8 + (((C) >> 0) & 0x3), \
6682 8 + (((C) >> 2) & 0x3), \
6683 8 + (((C) >> 4) & 0x3), \
6684 8 + (((C) >> 6) & 0x3), \
6685 12 + (((C) >> 0) & 0x3), \
6686 12 + (((C) >> 2) & 0x3), \
6687 12 + (((C) >> 4) & 0x3), \
6688 12 + (((C) >> 6) & 0x3)); })
6690 #define _mm512_mask_permute_ps(W, U, X, C) __extension__ ({ \
6691 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6692 (__v16sf)_mm512_permute_ps((X), (C)), \
6693 (__v16sf)(__m512)(W)); })
6695 #define _mm512_maskz_permute_ps(U, X, C) __extension__ ({ \
6696 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6697 (__v16sf)_mm512_permute_ps((X), (C)), \
6698 (__v16sf)_mm512_setzero_ps()); })
6703 return (__m512d)__builtin_ia32_vpermilvarpd512((__v8df)__A, (__v8di)__C);
6709 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
6717 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__U,
6725 return (__m512)__builtin_ia32_vpermilvarps512((__v16sf)__A, (__v16si)__C);
6731 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
6739 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
6747 return (__m512d) __builtin_ia32_vpermt2varpd512_mask ((__v8di) __I
6757 return (__m512d) __builtin_ia32_vpermt2varpd512_mask ((__v8di) __I
6768 return (__m512d) __builtin_ia32_vpermt2varpd512_maskz ((__v8di) __I
6778 return (__m512) __builtin_ia32_vpermt2varps512_mask ((__v16si) __I
6788 return (__m512) __builtin_ia32_vpermt2varps512_mask ((__v16si) __I
6799 return (__m512) __builtin_ia32_vpermt2varps512_maskz ((__v16si) __I
6809 return (__mmask16) __builtin_ia32_ptestnmd512 ((__v16si) __A,
6817 return (__mmask16) __builtin_ia32_ptestnmd512 ((__v16si) __A,
6818 (__v16si) __B, __U);
6824 return (__mmask8) __builtin_ia32_ptestnmq512 ((__v8di) __A,
6832 return (__mmask8) __builtin_ia32_ptestnmq512 ((__v8di) __A,
6836 #define _mm512_cvtt_roundpd_epu32(A, R) __extension__ ({ \
6837 (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6838 (__v8si)_mm256_undefined_si256(), \
6839 (__mmask8)-1, (int)(R)); })
6841 #define _mm512_mask_cvtt_roundpd_epu32(W, U, A, R) __extension__ ({ \
6842 (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6843 (__v8si)(__m256i)(W), \
6844 (__mmask8)(U), (int)(R)); })
6846 #define _mm512_maskz_cvtt_roundpd_epu32(U, A, R) __extension__ ({ \
6847 (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6848 (__v8si)_mm256_setzero_si256(), \
6849 (__mmask8)(U), (int)(R)); })
6854 return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
6864 return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
6873 return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
6880 #define _mm_roundscale_round_sd(A, B, imm, R) __extension__ ({ \
6881 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6882 (__v2df)(__m128d)(B), \
6883 (__v2df)_mm_setzero_pd(), \
6884 (__mmask8)-1, (int)(imm), \
6887 #define _mm_roundscale_sd(A, B, imm) __extension__ ({ \
6888 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6889 (__v2df)(__m128d)(B), \
6890 (__v2df)_mm_setzero_pd(), \
6891 (__mmask8)-1, (int)(imm), \
6892 _MM_FROUND_CUR_DIRECTION); })
6894 #define _mm_mask_roundscale_sd(W, U, A, B, imm) __extension__ ({ \
6895 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6896 (__v2df)(__m128d)(B), \
6897 (__v2df)(__m128d)(W), \
6898 (__mmask8)(U), (int)(imm), \
6899 _MM_FROUND_CUR_DIRECTION); })
6901 #define _mm_mask_roundscale_round_sd(W, U, A, B, I, R) __extension__ ({ \
6902 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6903 (__v2df)(__m128d)(B), \
6904 (__v2df)(__m128d)(W), \
6905 (__mmask8)(U), (int)(I), \
6908 #define _mm_maskz_roundscale_sd(U, A, B, I) __extension__ ({ \
6909 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6910 (__v2df)(__m128d)(B), \
6911 (__v2df)_mm_setzero_pd(), \
6912 (__mmask8)(U), (int)(I), \
6913 _MM_FROUND_CUR_DIRECTION); })
6915 #define _mm_maskz_roundscale_round_sd(U, A, B, I, R) __extension__ ({ \
6916 (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6917 (__v2df)(__m128d)(B), \
6918 (__v2df)_mm_setzero_pd(), \
6919 (__mmask8)(U), (int)(I), \
6922 #define _mm_roundscale_round_ss(A, B, imm, R) __extension__ ({ \
6923 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6924 (__v4sf)(__m128)(B), \
6925 (__v4sf)_mm_setzero_ps(), \
6926 (__mmask8)-1, (int)(imm), \
6929 #define _mm_roundscale_ss(A, B, imm) __extension__ ({ \
6930 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6931 (__v4sf)(__m128)(B), \
6932 (__v4sf)_mm_setzero_ps(), \
6933 (__mmask8)-1, (int)(imm), \
6934 _MM_FROUND_CUR_DIRECTION); })
6936 #define _mm_mask_roundscale_ss(W, U, A, B, I) __extension__ ({ \
6937 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6938 (__v4sf)(__m128)(B), \
6939 (__v4sf)(__m128)(W), \
6940 (__mmask8)(U), (int)(I), \
6941 _MM_FROUND_CUR_DIRECTION); })
6943 #define _mm_mask_roundscale_round_ss(W, U, A, B, I, R) __extension__ ({ \
6944 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6945 (__v4sf)(__m128)(B), \
6946 (__v4sf)(__m128)(W), \
6947 (__mmask8)(U), (int)(I), \
6950 #define _mm_maskz_roundscale_ss(U, A, B, I) __extension__ ({ \
6951 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6952 (__v4sf)(__m128)(B), \
6953 (__v4sf)_mm_setzero_ps(), \
6954 (__mmask8)(U), (int)(I), \
6955 _MM_FROUND_CUR_DIRECTION); })
6957 #define _mm_maskz_roundscale_round_ss(U, A, B, I, R) __extension__ ({ \
6958 (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6959 (__v4sf)(__m128)(B), \
6960 (__v4sf)_mm_setzero_ps(), \
6961 (__mmask8)(U), (int)(I), \
6964 #define _mm512_scalef_round_pd(A, B, R) __extension__ ({ \
6965 (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6966 (__v8df)(__m512d)(B), \
6967 (__v8df)_mm512_undefined_pd(), \
6968 (__mmask8)-1, (int)(R)); })
6970 #define _mm512_mask_scalef_round_pd(W, U, A, B, R) __extension__ ({ \
6971 (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6972 (__v8df)(__m512d)(B), \
6973 (__v8df)(__m512d)(W), \
6974 (__mmask8)(U), (int)(R)); })
6976 #define _mm512_maskz_scalef_round_pd(U, A, B, R) __extension__ ({ \
6977 (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6978 (__v8df)(__m512d)(B), \
6979 (__v8df)_mm512_setzero_pd(), \
6980 (__mmask8)(U), (int)(R)); })
6985 return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
6996 return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
7006 return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
7014 #define _mm512_scalef_round_ps(A, B, R) __extension__ ({ \
7015 (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
7016 (__v16sf)(__m512)(B), \
7017 (__v16sf)_mm512_undefined_ps(), \
7018 (__mmask16)-1, (int)(R)); })
7020 #define _mm512_mask_scalef_round_ps(W, U, A, B, R) __extension__ ({ \
7021 (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
7022 (__v16sf)(__m512)(B), \
7023 (__v16sf)(__m512)(W), \
7024 (__mmask16)(U), (int)(R)); })
7026 #define _mm512_maskz_scalef_round_ps(U, A, B, R) __extension__ ({ \
7027 (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
7028 (__v16sf)(__m512)(B), \
7029 (__v16sf)_mm512_setzero_ps(), \
7030 (__mmask16)(U), (int)(R)); })
7035 return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
7046 return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
7056 return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
7064 #define _mm_scalef_round_sd(A, B, R) __extension__ ({ \
7065 (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
7066 (__v2df)(__m128d)(B), \
7067 (__v2df)_mm_setzero_pd(), \
7068 (__mmask8)-1, (int)(R)); })
7073 return (__m128d) __builtin_ia32_scalefsd_round_mask ((__v2df) __A,
7082 return (__m128d) __builtin_ia32_scalefsd_round_mask ( (__v2df) __A,
7089 #define _mm_mask_scalef_round_sd(W, U, A, B, R) __extension__ ({ \
7090 (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
7091 (__v2df)(__m128d)(B), \
7092 (__v2df)(__m128d)(W), \
7093 (__mmask8)(U), (int)(R)); })
7098 return (__m128d) __builtin_ia32_scalefsd_round_mask ( (__v2df) __A,
7105 #define _mm_maskz_scalef_round_sd(U, A, B, R) __extension__ ({ \
7106 (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
7107 (__v2df)(__m128d)(B), \
7108 (__v2df)_mm_setzero_pd(), \
7109 (__mmask8)(U), (int)(R)); })
7111 #define _mm_scalef_round_ss(A, B, R) __extension__ ({ \
7112 (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
7113 (__v4sf)(__m128)(B), \
7114 (__v4sf)_mm_setzero_ps(), \
7115 (__mmask8)-1, (int)(R)); })
7120 return (__m128) __builtin_ia32_scalefss_round_mask ((__v4sf) __A,
7129 return (__m128) __builtin_ia32_scalefss_round_mask ( (__v4sf) __A,
7136 #define _mm_mask_scalef_round_ss(W, U, A, B, R) __extension__ ({ \
7137 (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
7138 (__v4sf)(__m128)(B), \
7139 (__v4sf)(__m128)(W), \
7140 (__mmask8)(U), (int)(R)); })
7145 return (__m128) __builtin_ia32_scalefss_round_mask ( (__v4sf) __A,
7152 #define _mm_maskz_scalef_round_ss(U, A, B, R) __extension__ ({ \
7153 (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
7154 (__v4sf)(__m128)(B), \
7155 (__v4sf)_mm_setzero_ps(), \
7157 _MM_FROUND_CUR_DIRECTION); })
7162 return (__m512i)__builtin_ia32_psradi512((__v16si)__A, __B);
7168 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U, \
7175 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__U, \
7183 return (__m512i)__builtin_ia32_psraqi512((__v8di)__A, __B);
7189 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U, \
7197 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__U, \
7202 #define _mm512_shuffle_f32x4(A, B, imm) __extension__ ({ \
7203 (__m512)__builtin_ia32_shuf_f32x4_mask((__v16sf)(__m512)(A), \
7204 (__v16sf)(__m512)(B), (int)(imm), \
7205 (__v16sf)_mm512_undefined_ps(), \
7208 #define _mm512_mask_shuffle_f32x4(W, U, A, B, imm) __extension__ ({ \
7209 (__m512)__builtin_ia32_shuf_f32x4_mask((__v16sf)(__m512)(A), \
7210 (__v16sf)(__m512)(B), (int)(imm), \
7211 (__v16sf)(__m512)(W), \
7214 #define _mm512_maskz_shuffle_f32x4(U, A, B, imm) __extension__ ({ \
7215 (__m512)__builtin_ia32_shuf_f32x4_mask((__v16sf)(__m512)(A), \
7216 (__v16sf)(__m512)(B), (int)(imm), \
7217 (__v16sf)_mm512_setzero_ps(), \
7220 #define _mm512_shuffle_f64x2(A, B, imm) __extension__ ({ \
7221 (__m512d)__builtin_ia32_shuf_f64x2_mask((__v8df)(__m512d)(A), \
7222 (__v8df)(__m512d)(B), (int)(imm), \
7223 (__v8df)_mm512_undefined_pd(), \
7226 #define _mm512_mask_shuffle_f64x2(W, U, A, B, imm) __extension__ ({ \
7227 (__m512d)__builtin_ia32_shuf_f64x2_mask((__v8df)(__m512d)(A), \
7228 (__v8df)(__m512d)(B), (int)(imm), \
7229 (__v8df)(__m512d)(W), \
7232 #define _mm512_maskz_shuffle_f64x2(U, A, B, imm) __extension__ ({ \
7233 (__m512d)__builtin_ia32_shuf_f64x2_mask((__v8df)(__m512d)(A), \
7234 (__v8df)(__m512d)(B), (int)(imm), \
7235 (__v8df)_mm512_setzero_pd(), \
7238 #define _mm512_shuffle_i32x4(A, B, imm) __extension__ ({ \
7239 (__m512i)__builtin_ia32_shuf_i32x4_mask((__v16si)(__m512i)(A), \
7240 (__v16si)(__m512i)(B), (int)(imm), \
7241 (__v16si)_mm512_setzero_si512(), \
7244 #define _mm512_mask_shuffle_i32x4(W, U, A, B, imm) __extension__ ({ \
7245 (__m512i)__builtin_ia32_shuf_i32x4_mask((__v16si)(__m512i)(A), \
7246 (__v16si)(__m512i)(B), (int)(imm), \
7247 (__v16si)(__m512i)(W), \
7250 #define _mm512_maskz_shuffle_i32x4(U, A, B, imm) __extension__ ({ \
7251 (__m512i)__builtin_ia32_shuf_i32x4_mask((__v16si)(__m512i)(A), \
7252 (__v16si)(__m512i)(B), (int)(imm), \
7253 (__v16si)_mm512_setzero_si512(), \
7256 #define _mm512_shuffle_i64x2(A, B, imm) __extension__ ({ \
7257 (__m512i)__builtin_ia32_shuf_i64x2_mask((__v8di)(__m512i)(A), \
7258 (__v8di)(__m512i)(B), (int)(imm), \
7259 (__v8di)_mm512_setzero_si512(), \
7262 #define _mm512_mask_shuffle_i64x2(W, U, A, B, imm) __extension__ ({ \
7263 (__m512i)__builtin_ia32_shuf_i64x2_mask((__v8di)(__m512i)(A), \
7264 (__v8di)(__m512i)(B), (int)(imm), \
7265 (__v8di)(__m512i)(W), \
7268 #define _mm512_maskz_shuffle_i64x2(U, A, B, imm) __extension__ ({ \
7269 (__m512i)__builtin_ia32_shuf_i64x2_mask((__v8di)(__m512i)(A), \
7270 (__v8di)(__m512i)(B), (int)(imm), \
7271 (__v8di)_mm512_setzero_si512(), \
7274 #define _mm512_shuffle_pd(A, B, M) __extension__ ({ \
7275 (__m512d)__builtin_shufflevector((__v8df)(__m512d)(A), \
7276 (__v8df)(__m512d)(B), \
7277 0 + (((M) >> 0) & 0x1), \
7278 8 + (((M) >> 1) & 0x1), \
7279 2 + (((M) >> 2) & 0x1), \
7280 10 + (((M) >> 3) & 0x1), \
7281 4 + (((M) >> 4) & 0x1), \
7282 12 + (((M) >> 5) & 0x1), \
7283 6 + (((M) >> 6) & 0x1), \
7284 14 + (((M) >> 7) & 0x1)); })
7286 #define _mm512_mask_shuffle_pd(W, U, A, B, M) __extension__ ({ \
7287 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7288 (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
7289 (__v8df)(__m512d)(W)); })
7291 #define _mm512_maskz_shuffle_pd(U, A, B, M) __extension__ ({ \
7292 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7293 (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
7294 (__v8df)_mm512_setzero_pd()); })
7296 #define _mm512_shuffle_ps(A, B, M) __extension__ ({ \
7297 (__m512d)__builtin_shufflevector((__v16sf)(__m512)(A), \
7298 (__v16sf)(__m512)(B), \
7299 0 + (((M) >> 0) & 0x3), \
7300 0 + (((M) >> 2) & 0x3), \
7301 16 + (((M) >> 4) & 0x3), \
7302 16 + (((M) >> 6) & 0x3), \
7303 4 + (((M) >> 0) & 0x3), \
7304 4 + (((M) >> 2) & 0x3), \
7305 20 + (((M) >> 4) & 0x3), \
7306 20 + (((M) >> 6) & 0x3), \
7307 8 + (((M) >> 0) & 0x3), \
7308 8 + (((M) >> 2) & 0x3), \
7309 24 + (((M) >> 4) & 0x3), \
7310 24 + (((M) >> 6) & 0x3), \
7311 12 + (((M) >> 0) & 0x3), \
7312 12 + (((M) >> 2) & 0x3), \
7313 28 + (((M) >> 4) & 0x3), \
7314 28 + (((M) >> 6) & 0x3)); })
7316 #define _mm512_mask_shuffle_ps(W, U, A, B, M) __extension__ ({ \
7317 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7318 (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
7319 (__v16sf)(__m512)(W)); })
7321 #define _mm512_maskz_shuffle_ps(U, A, B, M) __extension__ ({ \
7322 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7323 (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
7324 (__v16sf)_mm512_setzero_ps()); })
7326 #define _mm_sqrt_round_sd(A, B, R) __extension__ ({ \
7327 (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
7328 (__v2df)(__m128d)(B), \
7329 (__v2df)_mm_setzero_pd(), \
7330 (__mmask8)-1, (int)(R)); })
7335 return (__m128d) __builtin_ia32_sqrtsd_round_mask ( (__v2df) __A,
7342 #define _mm_mask_sqrt_round_sd(W, U, A, B, R) __extension__ ({ \
7343 (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
7344 (__v2df)(__m128d)(B), \
7345 (__v2df)(__m128d)(W), \
7346 (__mmask8)(U), (int)(R)); })
7351 return (__m128d) __builtin_ia32_sqrtsd_round_mask ( (__v2df) __A,
7358 #define _mm_maskz_sqrt_round_sd(U, A, B, R) __extension__ ({ \
7359 (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
7360 (__v2df)(__m128d)(B), \
7361 (__v2df)_mm_setzero_pd(), \
7362 (__mmask8)(U), (int)(R)); })
7364 #define _mm_sqrt_round_ss(A, B, R) __extension__ ({ \
7365 (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
7366 (__v4sf)(__m128)(B), \
7367 (__v4sf)_mm_setzero_ps(), \
7368 (__mmask8)-1, (int)(R)); })
7373 return (__m128) __builtin_ia32_sqrtss_round_mask ( (__v4sf) __A,
7380 #define _mm_mask_sqrt_round_ss(W, U, A, B, R) __extension__ ({ \
7381 (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
7382 (__v4sf)(__m128)(B), \
7383 (__v4sf)(__m128)(W), (__mmask8)(U), \
7389 return (__m128) __builtin_ia32_sqrtss_round_mask ( (__v4sf) __A,
7396 #define _mm_maskz_sqrt_round_ss(U, A, B, R) __extension__ ({ \
7397 (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
7398 (__v4sf)(__m128)(B), \
7399 (__v4sf)_mm_setzero_ps(), \
7400 (__mmask8)(U), (int)(R)); })
7405 return (__m512)__builtin_shufflevector((__v4sf)__A, (__v4sf)__A,
7406 0, 1, 2, 3, 0, 1, 2, 3,
7407 0, 1, 2, 3, 0, 1, 2, 3);
7413 return (__m512)__builtin_ia32_selectps_512((__mmask16)__M,
7421 return (__m512)__builtin_ia32_selectps_512((__mmask16)__M,
7429 return (__m512d)__builtin_shufflevector((__v4df)__A, (__v4df)__A,
7430 0, 1, 2, 3, 0, 1, 2, 3);
7436 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__M,
7444 return (__m512d)__builtin_ia32_selectpd_512((__mmask8)__M,
7452 return (__m512i)__builtin_shufflevector((__v4si)__A, (__v4si)__A,
7453 0, 1, 2, 3, 0, 1, 2, 3,
7454 0, 1, 2, 3, 0, 1, 2, 3);
7460 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__M,
7468 return (__m512i)__builtin_ia32_selectd_512((__mmask16)__M,
7476 return (__m512i)__builtin_shufflevector((__v4di)__A, (__v4di)__A,
7477 0, 1, 2, 3, 0, 1, 2, 3);
7483 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
7491 return (__m512i)__builtin_ia32_selectq_512((__mmask8)__M,
7499 return (__m512d)__builtin_ia32_selectpd_512(__M,
7507 return (__m512d)__builtin_ia32_selectpd_512(__M,
7515 return (__m512)__builtin_ia32_selectps_512(__M,
7523 return (__m512)__builtin_ia32_selectps_512(__M,
7531 return (__m128i) __builtin_ia32_pmovsdb512_mask ((__v16si) __A,
7539 return (__m128i) __builtin_ia32_pmovsdb512_mask ((__v16si) __A,
7540 (__v16qi) __O, __M);
7546 return (__m128i) __builtin_ia32_pmovsdb512_mask ((__v16si) __A,
7554 __builtin_ia32_pmovsdb512mem_mask ((__v16qi *) __P, (__v16si) __A, __M);
7560 return (__m256i) __builtin_ia32_pmovsdw512_mask ((__v16si) __A,
7568 return (__m256i) __builtin_ia32_pmovsdw512_mask ((__v16si) __A,
7569 (__v16hi) __O, __M);
7575 return (__m256i) __builtin_ia32_pmovsdw512_mask ((__v16si) __A,
7583 __builtin_ia32_pmovsdw512mem_mask ((__v16hi*) __P, (__v16si) __A, __M);
7589 return (__m128i) __builtin_ia32_pmovsqb512_mask ((__v8di) __A,
7597 return (__m128i) __builtin_ia32_pmovsqb512_mask ((__v8di) __A,
7598 (__v16qi) __O, __M);
7604 return (__m128i) __builtin_ia32_pmovsqb512_mask ((__v8di) __A,
7612 __builtin_ia32_pmovsqb512mem_mask ((__v16qi *) __P, (__v8di) __A, __M);
7618 return (__m256i) __builtin_ia32_pmovsqd512_mask ((__v8di) __A,
7626 return (__m256i) __builtin_ia32_pmovsqd512_mask ((__v8di) __A,
7633 return (__m256i) __builtin_ia32_pmovsqd512_mask ((__v8di) __A,
7641 __builtin_ia32_pmovsqd512mem_mask ((__v8si *) __P, (__v8di) __A, __M);
7647 return (__m128i) __builtin_ia32_pmovsqw512_mask ((__v8di) __A,
7655 return (__m128i) __builtin_ia32_pmovsqw512_mask ((__v8di) __A,
7662 return (__m128i) __builtin_ia32_pmovsqw512_mask ((__v8di) __A,
7670 __builtin_ia32_pmovsqw512mem_mask ((__v8hi *) __P, (__v8di) __A, __M);
7676 return (__m128i) __builtin_ia32_pmovusdb512_mask ((__v16si) __A,
7684 return (__m128i) __builtin_ia32_pmovusdb512_mask ((__v16si) __A,
7692 return (__m128i) __builtin_ia32_pmovusdb512_mask ((__v16si) __A,
7700 __builtin_ia32_pmovusdb512mem_mask ((__v16qi *) __P, (__v16si) __A, __M);
7706 return (__m256i) __builtin_ia32_pmovusdw512_mask ((__v16si) __A,
7714 return (__m256i) __builtin_ia32_pmovusdw512_mask ((__v16si) __A,
7722 return (__m256i) __builtin_ia32_pmovusdw512_mask ((__v16si) __A,
7730 __builtin_ia32_pmovusdw512mem_mask ((__v16hi*) __P, (__v16si) __A, __M);
7736 return (__m128i) __builtin_ia32_pmovusqb512_mask ((__v8di) __A,
7744 return (__m128i) __builtin_ia32_pmovusqb512_mask ((__v8di) __A,
7752 return (__m128i) __builtin_ia32_pmovusqb512_mask ((__v8di) __A,
7760 __builtin_ia32_pmovusqb512mem_mask ((__v16qi *) __P, (__v8di) __A, __M);
7766 return (__m256i) __builtin_ia32_pmovusqd512_mask ((__v8di) __A,
7774 return (__m256i) __builtin_ia32_pmovusqd512_mask ((__v8di) __A,
7781 return (__m256i) __builtin_ia32_pmovusqd512_mask ((__v8di) __A,
7789 __builtin_ia32_pmovusqd512mem_mask ((__v8si*) __P, (__v8di) __A, __M);
7795 return (__m128i) __builtin_ia32_pmovusqw512_mask ((__v8di) __A,
7803 return (__m128i) __builtin_ia32_pmovusqw512_mask ((__v8di) __A,
7810 return (__m128i) __builtin_ia32_pmovusqw512_mask ((__v8di) __A,
7818 __builtin_ia32_pmovusqw512mem_mask ((__v8hi*) __P, (__v8di) __A, __M);
7824 return (__m128i) __builtin_ia32_pmovdb512_mask ((__v16si) __A,
7832 return (__m128i) __builtin_ia32_pmovdb512_mask ((__v16si) __A,
7833 (__v16qi) __O, __M);
7839 return (__m128i) __builtin_ia32_pmovdb512_mask ((__v16si) __A,
7847 __builtin_ia32_pmovdb512mem_mask ((__v16qi *) __P, (__v16si) __A, __M);
7853 return (__m256i) __builtin_ia32_pmovdw512_mask ((__v16si) __A,
7861 return (__m256i) __builtin_ia32_pmovdw512_mask ((__v16si) __A,
7862 (__v16hi) __O, __M);
7868 return (__m256i) __builtin_ia32_pmovdw512_mask ((__v16si) __A,
7876 __builtin_ia32_pmovdw512mem_mask ((__v16hi *) __P, (__v16si) __A, __M);
7882 return (__m128i) __builtin_ia32_pmovqb512_mask ((__v8di) __A,
7890 return (__m128i) __builtin_ia32_pmovqb512_mask ((__v8di) __A,
7891 (__v16qi) __O, __M);
7897 return (__m128i) __builtin_ia32_pmovqb512_mask ((__v8di) __A,
7905 __builtin_ia32_pmovqb512mem_mask ((__v16qi *) __P, (__v8di) __A, __M);
7911 return (__m256i) __builtin_ia32_pmovqd512_mask ((__v8di) __A,
7919 return (__m256i) __builtin_ia32_pmovqd512_mask ((__v8di) __A,
7926 return (__m256i) __builtin_ia32_pmovqd512_mask ((__v8di) __A,
7934 __builtin_ia32_pmovqd512mem_mask ((__v8si *) __P, (__v8di) __A, __M);
7940 return (__m128i) __builtin_ia32_pmovqw512_mask ((__v8di) __A,
7948 return (__m128i) __builtin_ia32_pmovqw512_mask ((__v8di) __A,
7955 return (__m128i) __builtin_ia32_pmovqw512_mask ((__v8di) __A,
7963 __builtin_ia32_pmovqw512mem_mask ((__v8hi *) __P, (__v8di) __A, __M);
7966 #define _mm512_extracti32x4_epi32(A, imm) __extension__ ({ \
7967 (__m128i)__builtin_shufflevector((__v16si)(__m512i)(A), \
7968 (__v16si)_mm512_undefined_epi32(), \
7969 0 + ((imm) & 0x3) * 4, \
7970 1 + ((imm) & 0x3) * 4, \
7971 2 + ((imm) & 0x3) * 4, \
7972 3 + ((imm) & 0x3) * 4); })
7974 #define _mm512_mask_extracti32x4_epi32(W, U, A, imm) __extension__ ({ \
7975 (__m128i)__builtin_ia32_selectd_128((__mmask8)(U), \
7976 (__v4si)_mm512_extracti32x4_epi32((A), (imm)), \
7979 #define _mm512_maskz_extracti32x4_epi32(U, A, imm) __extension__ ({ \
7980 (__m128i)__builtin_ia32_selectd_128((__mmask8)(U), \
7981 (__v4si)_mm512_extracti32x4_epi32((A), (imm)), \
7982 (__v4si)_mm_setzero_si128()); })
7984 #define _mm512_extracti64x4_epi64(A, imm) __extension__ ({ \
7985 (__m256i)__builtin_shufflevector((__v8di)(__m512i)(A), \
7986 (__v8di)_mm512_undefined_epi32(), \
7987 ((imm) & 1) ? 4 : 0, \
7988 ((imm) & 1) ? 5 : 1, \
7989 ((imm) & 1) ? 6 : 2, \
7990 ((imm) & 1) ? 7 : 3); })
7992 #define _mm512_mask_extracti64x4_epi64(W, U, A, imm) __extension__ ({ \
7993 (__m256i)__builtin_ia32_selectq_256((__mmask8)(U), \
7994 (__v4di)_mm512_extracti64x4_epi64((A), (imm)), \
7997 #define _mm512_maskz_extracti64x4_epi64(U, A, imm) __extension__ ({ \
7998 (__m256i)__builtin_ia32_selectq_256((__mmask8)(U), \
7999 (__v4di)_mm512_extracti64x4_epi64((A), (imm)), \
8000 (__v4di)_mm256_setzero_si256()); })
8002 #define _mm512_insertf64x4(A, B, imm) __extension__ ({ \
8003 (__m512d)__builtin_shufflevector((__v8df)(__m512d)(A), \
8004 (__v8df)_mm512_castpd256_pd512((__m256d)(B)), \
8005 ((imm) & 0x1) ? 0 : 8, \
8006 ((imm) & 0x1) ? 1 : 9, \
8007 ((imm) & 0x1) ? 2 : 10, \
8008 ((imm) & 0x1) ? 3 : 11, \
8009 ((imm) & 0x1) ? 8 : 4, \
8010 ((imm) & 0x1) ? 9 : 5, \
8011 ((imm) & 0x1) ? 10 : 6, \
8012 ((imm) & 0x1) ? 11 : 7); })
8014 #define _mm512_mask_insertf64x4(W, U, A, B, imm) __extension__ ({ \
8015 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8016 (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
8019 #define _mm512_maskz_insertf64x4(U, A, B, imm) __extension__ ({ \
8020 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8021 (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
8022 (__v8df)_mm512_setzero_pd()); })
8024 #define _mm512_inserti64x4(A, B, imm) __extension__ ({ \
8025 (__m512i)__builtin_shufflevector((__v8di)(__m512i)(A), \
8026 (__v8di)_mm512_castsi256_si512((__m256i)(B)), \
8027 ((imm) & 0x1) ? 0 : 8, \
8028 ((imm) & 0x1) ? 1 : 9, \
8029 ((imm) & 0x1) ? 2 : 10, \
8030 ((imm) & 0x1) ? 3 : 11, \
8031 ((imm) & 0x1) ? 8 : 4, \
8032 ((imm) & 0x1) ? 9 : 5, \
8033 ((imm) & 0x1) ? 10 : 6, \
8034 ((imm) & 0x1) ? 11 : 7); })
8036 #define _mm512_mask_inserti64x4(W, U, A, B, imm) __extension__ ({ \
8037 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8038 (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
8041 #define _mm512_maskz_inserti64x4(U, A, B, imm) __extension__ ({ \
8042 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8043 (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
8044 (__v8di)_mm512_setzero_si512()); })
8046 #define _mm512_insertf32x4(A, B, imm) __extension__ ({ \
8047 (__m512)__builtin_shufflevector((__v16sf)(__m512)(A), \
8048 (__v16sf)_mm512_castps128_ps512((__m128)(B)),\
8049 (((imm) & 0x3) == 0) ? 16 : 0, \
8050 (((imm) & 0x3) == 0) ? 17 : 1, \
8051 (((imm) & 0x3) == 0) ? 18 : 2, \
8052 (((imm) & 0x3) == 0) ? 19 : 3, \
8053 (((imm) & 0x3) == 1) ? 16 : 4, \
8054 (((imm) & 0x3) == 1) ? 17 : 5, \
8055 (((imm) & 0x3) == 1) ? 18 : 6, \
8056 (((imm) & 0x3) == 1) ? 19 : 7, \
8057 (((imm) & 0x3) == 2) ? 16 : 8, \
8058 (((imm) & 0x3) == 2) ? 17 : 9, \
8059 (((imm) & 0x3) == 2) ? 18 : 10, \
8060 (((imm) & 0x3) == 2) ? 19 : 11, \
8061 (((imm) & 0x3) == 3) ? 16 : 12, \
8062 (((imm) & 0x3) == 3) ? 17 : 13, \
8063 (((imm) & 0x3) == 3) ? 18 : 14, \
8064 (((imm) & 0x3) == 3) ? 19 : 15); })
8066 #define _mm512_mask_insertf32x4(W, U, A, B, imm) __extension__ ({ \
8067 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
8068 (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
8071 #define _mm512_maskz_insertf32x4(U, A, B, imm) __extension__ ({ \
8072 (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
8073 (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
8074 (__v16sf)_mm512_setzero_ps()); })
8076 #define _mm512_inserti32x4(A, B, imm) __extension__ ({ \
8077 (__m512i)__builtin_shufflevector((__v16si)(__m512i)(A), \
8078 (__v16si)_mm512_castsi128_si512((__m128i)(B)),\
8079 (((imm) & 0x3) == 0) ? 16 : 0, \
8080 (((imm) & 0x3) == 0) ? 17 : 1, \
8081 (((imm) & 0x3) == 0) ? 18 : 2, \
8082 (((imm) & 0x3) == 0) ? 19 : 3, \
8083 (((imm) & 0x3) == 1) ? 16 : 4, \
8084 (((imm) & 0x3) == 1) ? 17 : 5, \
8085 (((imm) & 0x3) == 1) ? 18 : 6, \
8086 (((imm) & 0x3) == 1) ? 19 : 7, \
8087 (((imm) & 0x3) == 2) ? 16 : 8, \
8088 (((imm) & 0x3) == 2) ? 17 : 9, \
8089 (((imm) & 0x3) == 2) ? 18 : 10, \
8090 (((imm) & 0x3) == 2) ? 19 : 11, \
8091 (((imm) & 0x3) == 3) ? 16 : 12, \
8092 (((imm) & 0x3) == 3) ? 17 : 13, \
8093 (((imm) & 0x3) == 3) ? 18 : 14, \
8094 (((imm) & 0x3) == 3) ? 19 : 15); })
8096 #define _mm512_mask_inserti32x4(W, U, A, B, imm) __extension__ ({ \
8097 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8098 (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
8101 #define _mm512_maskz_inserti32x4(U, A, B, imm) __extension__ ({ \
8102 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8103 (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
8104 (__v16si)_mm512_setzero_si512()); })
8106 #define _mm512_getmant_round_pd(A, B, C, R) __extension__ ({ \
8107 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8108 (int)(((C)<<2) | (B)), \
8109 (__v8df)_mm512_undefined_pd(), \
8110 (__mmask8)-1, (int)(R)); })
8112 #define _mm512_mask_getmant_round_pd(W, U, A, B, C, R) __extension__ ({ \
8113 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8114 (int)(((C)<<2) | (B)), \
8115 (__v8df)(__m512d)(W), \
8116 (__mmask8)(U), (int)(R)); })
8118 #define _mm512_maskz_getmant_round_pd(U, A, B, C, R) __extension__ ({ \
8119 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8120 (int)(((C)<<2) | (B)), \
8121 (__v8df)_mm512_setzero_pd(), \
8122 (__mmask8)(U), (int)(R)); })
8124 #define _mm512_getmant_pd(A, B, C) __extension__ ({ \
8125 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8126 (int)(((C)<<2) | (B)), \
8127 (__v8df)_mm512_setzero_pd(), \
8129 _MM_FROUND_CUR_DIRECTION); })
8131 #define _mm512_mask_getmant_pd(W, U, A, B, C) __extension__ ({ \
8132 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8133 (int)(((C)<<2) | (B)), \
8134 (__v8df)(__m512d)(W), \
8136 _MM_FROUND_CUR_DIRECTION); })
8138 #define _mm512_maskz_getmant_pd(U, A, B, C) __extension__ ({ \
8139 (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
8140 (int)(((C)<<2) | (B)), \
8141 (__v8df)_mm512_setzero_pd(), \
8143 _MM_FROUND_CUR_DIRECTION); })
8145 #define _mm512_getmant_round_ps(A, B, C, R) __extension__ ({ \
8146 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8147 (int)(((C)<<2) | (B)), \
8148 (__v16sf)_mm512_undefined_ps(), \
8149 (__mmask16)-1, (int)(R)); })
8151 #define _mm512_mask_getmant_round_ps(W, U, A, B, C, R) __extension__ ({ \
8152 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8153 (int)(((C)<<2) | (B)), \
8154 (__v16sf)(__m512)(W), \
8155 (__mmask16)(U), (int)(R)); })
8157 #define _mm512_maskz_getmant_round_ps(U, A, B, C, R) __extension__ ({ \
8158 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8159 (int)(((C)<<2) | (B)), \
8160 (__v16sf)_mm512_setzero_ps(), \
8161 (__mmask16)(U), (int)(R)); })
8163 #define _mm512_getmant_ps(A, B, C) __extension__ ({ \
8164 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8165 (int)(((C)<<2)|(B)), \
8166 (__v16sf)_mm512_undefined_ps(), \
8168 _MM_FROUND_CUR_DIRECTION); })
8170 #define _mm512_mask_getmant_ps(W, U, A, B, C) __extension__ ({ \
8171 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8172 (int)(((C)<<2)|(B)), \
8173 (__v16sf)(__m512)(W), \
8175 _MM_FROUND_CUR_DIRECTION); })
8177 #define _mm512_maskz_getmant_ps(U, A, B, C) __extension__ ({ \
8178 (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
8179 (int)(((C)<<2)|(B)), \
8180 (__v16sf)_mm512_setzero_ps(), \
8182 _MM_FROUND_CUR_DIRECTION); })
8184 #define _mm512_getexp_round_pd(A, R) __extension__ ({ \
8185 (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
8186 (__v8df)_mm512_undefined_pd(), \
8187 (__mmask8)-1, (int)(R)); })
8189 #define _mm512_mask_getexp_round_pd(W, U, A, R) __extension__ ({ \
8190 (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
8191 (__v8df)(__m512d)(W), \
8192 (__mmask8)(U), (int)(R)); })
8194 #define _mm512_maskz_getexp_round_pd(U, A, R) __extension__ ({ \
8195 (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
8196 (__v8df)_mm512_setzero_pd(), \
8197 (__mmask8)(U), (int)(R)); })
8202 return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
8211 return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
8220 return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
8226 #define _mm512_getexp_round_ps(A, R) __extension__ ({ \
8227 (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
8228 (__v16sf)_mm512_undefined_ps(), \
8229 (__mmask16)-1, (int)(R)); })
8231 #define _mm512_mask_getexp_round_ps(W, U, A, R) __extension__ ({ \
8232 (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
8233 (__v16sf)(__m512)(W), \
8234 (__mmask16)(U), (int)(R)); })
8236 #define _mm512_maskz_getexp_round_ps(U, A, R) __extension__ ({ \
8237 (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
8238 (__v16sf)_mm512_setzero_ps(), \
8239 (__mmask16)(U), (int)(R)); })
8244 return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
8253 return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
8262 return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
8268 #define _mm512_i64gather_ps(index, addr, scale) __extension__ ({ \
8269 (__m256)__builtin_ia32_gatherdiv16sf((__v8sf)_mm256_undefined_ps(), \
8270 (float const *)(addr), \
8271 (__v8di)(__m512i)(index), (__mmask8)-1, \
8274 #define _mm512_mask_i64gather_ps(v1_old, mask, index, addr, scale) __extension__({\
8275 (__m256)__builtin_ia32_gatherdiv16sf((__v8sf)(__m256)(v1_old),\
8276 (float const *)(addr), \
8277 (__v8di)(__m512i)(index), \
8278 (__mmask8)(mask), (int)(scale)); })
8280 #define _mm512_i64gather_epi32(index, addr, scale) __extension__ ({\
8281 (__m256i)__builtin_ia32_gatherdiv16si((__v8si)_mm256_undefined_ps(), \
8282 (int const *)(addr), \
8283 (__v8di)(__m512i)(index), \
8284 (__mmask8)-1, (int)(scale)); })
8286 #define _mm512_mask_i64gather_epi32(v1_old, mask, index, addr, scale) __extension__ ({\
8287 (__m256i)__builtin_ia32_gatherdiv16si((__v8si)(__m256i)(v1_old), \
8288 (int const *)(addr), \
8289 (__v8di)(__m512i)(index), \
8290 (__mmask8)(mask), (int)(scale)); })
8292 #define _mm512_i64gather_pd(index, addr, scale) __extension__ ({\
8293 (__m512d)__builtin_ia32_gatherdiv8df((__v8df)_mm512_undefined_pd(), \
8294 (double const *)(addr), \
8295 (__v8di)(__m512i)(index), (__mmask8)-1, \
8298 #define _mm512_mask_i64gather_pd(v1_old, mask, index, addr, scale) __extension__ ({\
8299 (__m512d)__builtin_ia32_gatherdiv8df((__v8df)(__m512d)(v1_old), \
8300 (double const *)(addr), \
8301 (__v8di)(__m512i)(index), \
8302 (__mmask8)(mask), (int)(scale)); })
8304 #define _mm512_i64gather_epi64(index, addr, scale) __extension__ ({\
8305 (__m512i)__builtin_ia32_gatherdiv8di((__v8di)_mm512_undefined_pd(), \
8306 (long long const *)(addr), \
8307 (__v8di)(__m512i)(index), (__mmask8)-1, \
8310 #define _mm512_mask_i64gather_epi64(v1_old, mask, index, addr, scale) __extension__ ({\
8311 (__m512i)__builtin_ia32_gatherdiv8di((__v8di)(__m512i)(v1_old), \
8312 (long long const *)(addr), \
8313 (__v8di)(__m512i)(index), \
8314 (__mmask8)(mask), (int)(scale)); })
8316 #define _mm512_i32gather_ps(index, addr, scale) __extension__ ({\
8317 (__m512)__builtin_ia32_gathersiv16sf((__v16sf)_mm512_undefined_ps(), \
8318 (float const *)(addr), \
8319 (__v16sf)(__m512)(index), \
8320 (__mmask16)-1, (int)(scale)); })
8322 #define _mm512_mask_i32gather_ps(v1_old, mask, index, addr, scale) __extension__ ({\
8323 (__m512)__builtin_ia32_gathersiv16sf((__v16sf)(__m512)(v1_old), \
8324 (float const *)(addr), \
8325 (__v16sf)(__m512)(index), \
8326 (__mmask16)(mask), (int)(scale)); })
8328 #define _mm512_i32gather_epi32(index, addr, scale) __extension__ ({\
8329 (__m512i)__builtin_ia32_gathersiv16si((__v16si)_mm512_undefined_epi32(), \
8330 (int const *)(addr), \
8331 (__v16si)(__m512i)(index), \
8332 (__mmask16)-1, (int)(scale)); })
8334 #define _mm512_mask_i32gather_epi32(v1_old, mask, index, addr, scale) __extension__ ({\
8335 (__m512i)__builtin_ia32_gathersiv16si((__v16si)(__m512i)(v1_old), \
8336 (int const *)(addr), \
8337 (__v16si)(__m512i)(index), \
8338 (__mmask16)(mask), (int)(scale)); })
8340 #define _mm512_i32gather_pd(index, addr, scale) __extension__ ({\
8341 (__m512d)__builtin_ia32_gathersiv8df((__v8df)_mm512_undefined_pd(), \
8342 (double const *)(addr), \
8343 (__v8si)(__m256i)(index), (__mmask8)-1, \
8346 #define _mm512_mask_i32gather_pd(v1_old, mask, index, addr, scale) __extension__ ({\
8347 (__m512d)__builtin_ia32_gathersiv8df((__v8df)(__m512d)(v1_old), \
8348 (double const *)(addr), \
8349 (__v8si)(__m256i)(index), \
8350 (__mmask8)(mask), (int)(scale)); })
8352 #define _mm512_i32gather_epi64(index, addr, scale) __extension__ ({\
8353 (__m512i)__builtin_ia32_gathersiv8di((__v8di)_mm512_undefined_epi32(), \
8354 (long long const *)(addr), \
8355 (__v8si)(__m256i)(index), (__mmask8)-1, \
8358 #define _mm512_mask_i32gather_epi64(v1_old, mask, index, addr, scale) __extension__ ({\
8359 (__m512i)__builtin_ia32_gathersiv8di((__v8di)(__m512i)(v1_old), \
8360 (long long const *)(addr), \
8361 (__v8si)(__m256i)(index), \
8362 (__mmask8)(mask), (int)(scale)); })
8364 #define _mm512_i64scatter_ps(addr, index, v1, scale) __extension__ ({\
8365 __builtin_ia32_scatterdiv16sf((float *)(addr), (__mmask8)-1, \
8366 (__v8di)(__m512i)(index), \
8367 (__v8sf)(__m256)(v1), (int)(scale)); })
8369 #define _mm512_mask_i64scatter_ps(addr, mask, index, v1, scale) __extension__ ({\
8370 __builtin_ia32_scatterdiv16sf((float *)(addr), (__mmask8)(mask), \
8371 (__v8di)(__m512i)(index), \
8372 (__v8sf)(__m256)(v1), (int)(scale)); })
8374 #define _mm512_i64scatter_epi32(addr, index, v1, scale) __extension__ ({\
8375 __builtin_ia32_scatterdiv16si((int *)(addr), (__mmask8)-1, \
8376 (__v8di)(__m512i)(index), \
8377 (__v8si)(__m256i)(v1), (int)(scale)); })
8379 #define _mm512_mask_i64scatter_epi32(addr, mask, index, v1, scale) __extension__ ({\
8380 __builtin_ia32_scatterdiv16si((int *)(addr), (__mmask8)(mask), \
8381 (__v8di)(__m512i)(index), \
8382 (__v8si)(__m256i)(v1), (int)(scale)); })
8384 #define _mm512_i64scatter_pd(addr, index, v1, scale) __extension__ ({\
8385 __builtin_ia32_scatterdiv8df((double *)(addr), (__mmask8)-1, \
8386 (__v8di)(__m512i)(index), \
8387 (__v8df)(__m512d)(v1), (int)(scale)); })
8389 #define _mm512_mask_i64scatter_pd(addr, mask, index, v1, scale) __extension__ ({\
8390 __builtin_ia32_scatterdiv8df((double *)(addr), (__mmask8)(mask), \
8391 (__v8di)(__m512i)(index), \
8392 (__v8df)(__m512d)(v1), (int)(scale)); })
8394 #define _mm512_i64scatter_epi64(addr, index, v1, scale) __extension__ ({\
8395 __builtin_ia32_scatterdiv8di((long long *)(addr), (__mmask8)-1, \
8396 (__v8di)(__m512i)(index), \
8397 (__v8di)(__m512i)(v1), (int)(scale)); })
8399 #define _mm512_mask_i64scatter_epi64(addr, mask, index, v1, scale) __extension__ ({\
8400 __builtin_ia32_scatterdiv8di((long long *)(addr), (__mmask8)(mask), \
8401 (__v8di)(__m512i)(index), \
8402 (__v8di)(__m512i)(v1), (int)(scale)); })
8404 #define _mm512_i32scatter_ps(addr, index, v1, scale) __extension__ ({\
8405 __builtin_ia32_scattersiv16sf((float *)(addr), (__mmask16)-1, \
8406 (__v16si)(__m512i)(index), \
8407 (__v16sf)(__m512)(v1), (int)(scale)); })
8409 #define _mm512_mask_i32scatter_ps(addr, mask, index, v1, scale) __extension__ ({\
8410 __builtin_ia32_scattersiv16sf((float *)(addr), (__mmask16)(mask), \
8411 (__v16si)(__m512i)(index), \
8412 (__v16sf)(__m512)(v1), (int)(scale)); })
8414 #define _mm512_i32scatter_epi32(addr, index, v1, scale) __extension__ ({\
8415 __builtin_ia32_scattersiv16si((int *)(addr), (__mmask16)-1, \
8416 (__v16si)(__m512i)(index), \
8417 (__v16si)(__m512i)(v1), (int)(scale)); })
8419 #define _mm512_mask_i32scatter_epi32(addr, mask, index, v1, scale) __extension__ ({\
8420 __builtin_ia32_scattersiv16si((int *)(addr), (__mmask16)(mask), \
8421 (__v16si)(__m512i)(index), \
8422 (__v16si)(__m512i)(v1), (int)(scale)); })
8424 #define _mm512_i32scatter_pd(addr, index, v1, scale) __extension__ ({\
8425 __builtin_ia32_scattersiv8df((double *)(addr), (__mmask8)-1, \
8426 (__v8si)(__m256i)(index), \
8427 (__v8df)(__m512d)(v1), (int)(scale)); })
8429 #define _mm512_mask_i32scatter_pd(addr, mask, index, v1, scale) __extension__ ({\
8430 __builtin_ia32_scattersiv8df((double *)(addr), (__mmask8)(mask), \
8431 (__v8si)(__m256i)(index), \
8432 (__v8df)(__m512d)(v1), (int)(scale)); })
8434 #define _mm512_i32scatter_epi64(addr, index, v1, scale) __extension__ ({\
8435 __builtin_ia32_scattersiv8di((long long *)(addr), (__mmask8)-1, \
8436 (__v8si)(__m256i)(index), \
8437 (__v8di)(__m512i)(v1), (int)(scale)); })
8439 #define _mm512_mask_i32scatter_epi64(addr, mask, index, v1, scale) __extension__ ({\
8440 __builtin_ia32_scattersiv8di((long long *)(addr), (__mmask8)(mask), \
8441 (__v8si)(__m256i)(index), \
8442 (__v8di)(__m512i)(v1), (int)(scale)); })
8447 return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
8454 #define _mm_mask_fmadd_round_ss(W, U, A, B, R) __extension__({\
8455 (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
8456 (__v4sf)(__m128)(A), \
8457 (__v4sf)(__m128)(B), (__mmask8)(U), \
8463 return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __A,
8470 #define _mm_maskz_fmadd_round_ss(U, A, B, C, R) __extension__ ({\
8471 (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
8472 (__v4sf)(__m128)(B), \
8473 (__v4sf)(__m128)(C), (__mmask8)(U), \
8474 _MM_FROUND_CUR_DIRECTION); })
8479 return (__m128) __builtin_ia32_vfmaddss3_mask3 ((__v4sf) __W,
8486 #define _mm_mask3_fmadd_round_ss(W, X, Y, U, R) __extension__ ({\
8487 (__m128)__builtin_ia32_vfmaddss3_mask3((__v4sf)(__m128)(W), \
8488 (__v4sf)(__m128)(X), \
8489 (__v4sf)(__m128)(Y), (__mmask8)(U), \
8495 return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
8502 #define _mm_mask_fmsub_round_ss(W, U, A, B, R) __extension__ ({\
8503 (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
8504 (__v4sf)(__m128)(A), \
8505 (__v4sf)(__m128)(B), (__mmask8)(U), \
8511 return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __A,
8518 #define _mm_maskz_fmsub_round_ss(U, A, B, C, R) __extension__ ({\
8519 (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
8520 (__v4sf)(__m128)(B), \
8521 -(__v4sf)(__m128)(C), (__mmask8)(U), \
8527 return (__m128) __builtin_ia32_vfmsubss3_mask3 ((__v4sf) __W,
8534 #define _mm_mask3_fmsub_round_ss(W, X, Y, U, R) __extension__ ({\
8535 (__m128)__builtin_ia32_vfmsubss3_mask3((__v4sf)(__m128)(W), \
8536 (__v4sf)(__m128)(X), \
8537 (__v4sf)(__m128)(Y), (__mmask8)(U), \
8543 return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
8550 #define _mm_mask_fnmadd_round_ss(W, U, A, B, R) __extension__ ({\
8551 (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
8552 -(__v4sf)(__m128)(A), \
8553 (__v4sf)(__m128)(B), (__mmask8)(U), \
8559 return (__m128) __builtin_ia32_vfmaddss3_maskz (-(__v4sf) __A,
8566 #define _mm_maskz_fnmadd_round_ss(U, A, B, C, R) __extension__ ({\
8567 (__m128)__builtin_ia32_vfmaddss3_maskz(-(__v4sf)(__m128)(A), \
8568 (__v4sf)(__m128)(B), \
8569 (__v4sf)(__m128)(C), (__mmask8)(U), \
8575 return (__m128) __builtin_ia32_vfmaddss3_mask3 (-(__v4sf) __W,
8582 #define _mm_mask3_fnmadd_round_ss(W, X, Y, U, R) __extension__({\
8583 (__m128)__builtin_ia32_vfmaddss3_mask3(-(__v4sf)(__m128)(W), \
8584 (__v4sf)(__m128)(X), \
8585 (__v4sf)(__m128)(Y), (__mmask8)(U), \
8591 return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
8598 #define _mm_mask_fnmsub_round_ss(W, U, A, B, R) __extension__ ({\
8599 (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
8600 -(__v4sf)(__m128)(A), \
8601 -(__v4sf)(__m128)(B), (__mmask8)(U), \
8607 return (__m128) __builtin_ia32_vfmaddss3_maskz (-(__v4sf) __A,
8614 #define _mm_maskz_fnmsub_round_ss(U, A, B, C, R) __extension__ ({\
8615 (__m128)__builtin_ia32_vfmaddss3_maskz(-(__v4sf)(__m128)(A), \
8616 (__v4sf)(__m128)(B), \
8617 -(__v4sf)(__m128)(C), (__mmask8)(U), \
8618 _MM_FROUND_CUR_DIRECTION); })
8623 return (__m128) __builtin_ia32_vfnmsubss3_mask3 ((__v4sf) __W,
8630 #define _mm_mask3_fnmsub_round_ss(W, X, Y, U, R) __extension__({\
8631 (__m128)__builtin_ia32_vfnmsubss3_mask3((__v4sf)(__m128)(W), \
8632 (__v4sf)(__m128)(X), \
8633 (__v4sf)(__m128)(Y), (__mmask8)(U), \
8639 return (__m128d) __builtin_ia32_vfmaddsd3_mask ( (__v2df) __W,
8646 #define _mm_mask_fmadd_round_sd(W, U, A, B, R) __extension__({\
8647 (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8648 (__v2df)(__m128d)(A), \
8649 (__v2df)(__m128d)(B), (__mmask8)(U), \
8655 return (__m128d) __builtin_ia32_vfmaddsd3_maskz ( (__v2df) __A,
8662 #define _mm_maskz_fmadd_round_sd(U, A, B, C, R) __extension__ ({\
8663 (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8664 (__v2df)(__m128d)(B), \
8665 (__v2df)(__m128d)(C), (__mmask8)(U), \
8666 _MM_FROUND_CUR_DIRECTION); })
8671 return (__m128d) __builtin_ia32_vfmaddsd3_mask3 ((__v2df) __W,
8678 #define _mm_mask3_fmadd_round_sd(W, X, Y, U, R) __extension__ ({\
8679 (__m128d)__builtin_ia32_vfmaddsd3_mask3((__v2df)(__m128d)(W), \
8680 (__v2df)(__m128d)(X), \
8681 (__v2df)(__m128d)(Y), (__mmask8)(U), \
8687 return (__m128d) __builtin_ia32_vfmaddsd3_mask ( (__v2df) __W,
8694 #define _mm_mask_fmsub_round_sd(W, U, A, B, R) __extension__ ({\
8695 (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8696 (__v2df)(__m128d)(A), \
8697 -(__v2df)(__m128d)(B), (__mmask8)(U), \
8703 return (__m128d) __builtin_ia32_vfmaddsd3_maskz ( (__v2df) __A,
8710 #define _mm_maskz_fmsub_round_sd(U, A, B, C, R) __extension__ ({\
8711 (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8712 (__v2df)(__m128d)(B), \
8713 -(__v2df)(__m128d)(C), \
8714 (__mmask8)(U), (int)(R)); })
8719 return (__m128d) __builtin_ia32_vfmsubsd3_mask3 ((__v2df) __W,
8726 #define _mm_mask3_fmsub_round_sd(W, X, Y, U, R) __extension__ ({\
8727 (__m128d)__builtin_ia32_vfmsubsd3_mask3((__v2df)(__m128d)(W), \
8728 (__v2df)(__m128d)(X), \
8729 (__v2df)(__m128d)(Y), \
8730 (__mmask8)(U), (int)(R)); })
8735 return (__m128d) __builtin_ia32_vfmaddsd3_mask ( (__v2df) __W,
8742 #define _mm_mask_fnmadd_round_sd(W, U, A, B, R) __extension__ ({\
8743 (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8744 -(__v2df)(__m128d)(A), \
8745 (__v2df)(__m128d)(B), (__mmask8)(U), \
8751 return (__m128d) __builtin_ia32_vfmaddsd3_maskz ( -(__v2df) __A,
8758 #define _mm_maskz_fnmadd_round_sd(U, A, B, C, R) __extension__ ({\
8759 (__m128d)__builtin_ia32_vfmaddsd3_maskz(-(__v2df)(__m128d)(A), \
8760 (__v2df)(__m128d)(B), \
8761 (__v2df)(__m128d)(C), (__mmask8)(U), \
8767 return (__m128d) __builtin_ia32_vfmaddsd3_mask3 (-(__v2df) __W,
8774 #define _mm_mask3_fnmadd_round_sd(W, X, Y, U, R) __extension__({\
8775 (__m128d)__builtin_ia32_vfmaddsd3_mask3(-(__v2df)(__m128d)(W), \
8776 (__v2df)(__m128d)(X), \
8777 (__v2df)(__m128d)(Y), (__mmask8)(U), \
8783 return (__m128d) __builtin_ia32_vfmaddsd3_mask ( (__v2df) __W,
8790 #define _mm_mask_fnmsub_round_sd(W, U, A, B, R) __extension__ ({\
8791 (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8792 -(__v2df)(__m128d)(A), \
8793 -(__v2df)(__m128d)(B), (__mmask8)(U), \
8799 return (__m128d) __builtin_ia32_vfmaddsd3_maskz ( -(__v2df) __A,
8806 #define _mm_maskz_fnmsub_round_sd(U, A, B, C, R) __extension__ ({\
8807 (__m128d)__builtin_ia32_vfmaddsd3_maskz(-(__v2df)(__m128d)(A), \
8808 (__v2df)(__m128d)(B), \
8809 -(__v2df)(__m128d)(C), \
8811 _MM_FROUND_CUR_DIRECTION); })
8816 return (__m128d) __builtin_ia32_vfnmsubsd3_mask3 ((__v2df) (__W),
8823 #define _mm_mask3_fnmsub_round_sd(W, X, Y, U, R) __extension__({\
8824 (__m128d)__builtin_ia32_vfnmsubsd3_mask3((__v2df)(__m128d)(W), \
8825 (__v2df)(__m128d)(X), \
8826 (__v2df)(__m128d)(Y), \
8827 (__mmask8)(U), (int)(R)); })
8829 #define _mm512_permutex_pd(X, C) __extension__ ({ \
8830 (__m512d)__builtin_shufflevector((__v8df)(__m512d)(X), \
8831 (__v8df)_mm512_undefined_pd(), \
8832 0 + (((C) >> 0) & 0x3), \
8833 0 + (((C) >> 2) & 0x3), \
8834 0 + (((C) >> 4) & 0x3), \
8835 0 + (((C) >> 6) & 0x3), \
8836 4 + (((C) >> 0) & 0x3), \
8837 4 + (((C) >> 2) & 0x3), \
8838 4 + (((C) >> 4) & 0x3), \
8839 4 + (((C) >> 6) & 0x3)); })
8841 #define _mm512_mask_permutex_pd(W, U, X, C) __extension__ ({ \
8842 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8843 (__v8df)_mm512_permutex_pd((X), (C)), \
8844 (__v8df)(__m512d)(W)); })
8846 #define _mm512_maskz_permutex_pd(U, X, C) __extension__ ({ \
8847 (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8848 (__v8df)_mm512_permutex_pd((X), (C)), \
8849 (__v8df)_mm512_setzero_pd()); })
8851 #define _mm512_permutex_epi64(X, C) __extension__ ({ \
8852 (__m512i)__builtin_shufflevector((__v8di)(__m512i)(X), \
8853 (__v8di)_mm512_undefined_epi32(), \
8854 0 + (((C) >> 0) & 0x3), \
8855 0 + (((C) >> 2) & 0x3), \
8856 0 + (((C) >> 4) & 0x3), \
8857 0 + (((C) >> 6) & 0x3), \
8858 4 + (((C) >> 0) & 0x3), \
8859 4 + (((C) >> 2) & 0x3), \
8860 4 + (((C) >> 4) & 0x3), \
8861 4 + (((C) >> 6) & 0x3)); })
8863 #define _mm512_mask_permutex_epi64(W, U, X, C) __extension__ ({ \
8864 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8865 (__v8di)_mm512_permutex_epi64((X), (C)), \
8866 (__v8di)(__m512i)(W)); })
8868 #define _mm512_maskz_permutex_epi64(U, X, C) __extension__ ({ \
8869 (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8870 (__v8di)_mm512_permutex_epi64((X), (C)), \
8871 (__v8di)_mm512_setzero_si512()); })
8876 return (__m512d) __builtin_ia32_permvardf512_mask ((__v8df) __Y,
8885 return (__m512d) __builtin_ia32_permvardf512_mask ((__v8df) __Y,
8894 return (__m512d) __builtin_ia32_permvardf512_mask ((__v8df) __Y,
8903 return (__m512i) __builtin_ia32_permvardi512_mask ((__v8di) __Y,
8912 return (__m512i) __builtin_ia32_permvardi512_mask ((__v8di) __Y,
8922 return (__m512i) __builtin_ia32_permvardi512_mask ((__v8di) __Y,
8931 return (__m512) __builtin_ia32_permvarsf512_mask ((__v16sf) __Y,
8940 return (__m512) __builtin_ia32_permvarsf512_mask ((__v16sf) __Y,
8949 return (__m512) __builtin_ia32_permvarsf512_mask ((__v16sf) __Y,
8958 return (__m512i) __builtin_ia32_permvarsi512_mask ((__v16si) __Y,
8967 return (__m512i) __builtin_ia32_permvarsi512_mask ((__v16si) __Y,
8973 #define _mm512_permutevar_epi32 _mm512_permutexvar_epi32
8979 return (__m512i) __builtin_ia32_permvarsi512_mask ((__v16si) __Y,
8985 #define _mm512_mask_permutevar_epi32 _mm512_mask_permutexvar_epi32
8990 return (__mmask16) __builtin_ia32_kandhi ((__mmask16) __A, (__mmask16) __B);
8996 return (__mmask16) __builtin_ia32_kandnhi ((__mmask16) __A, (__mmask16) __B);
9002 return (__mmask16) __builtin_ia32_korhi ((__mmask16) __A, (__mmask16) __B);
9008 return __builtin_ia32_kortestchi ((__mmask16) __A, (__mmask16) __B);
9014 return __builtin_ia32_kortestzhi ((__mmask16) __A, (__mmask16) __B);
9020 return (__mmask16) __builtin_ia32_kunpckhi ((__mmask16) __A, (__mmask16) __B);
9026 return (__mmask16) __builtin_ia32_kxnorhi ((__mmask16) __A, (__mmask16) __B);
9032 return (__mmask16) __builtin_ia32_kxorhi ((__mmask16) __A, (__mmask16) __B);
9039 __builtin_nontemporal_store((__v8di_aligned)__A, (__v8di_aligned*)__P);
9046 return (__m512i) __builtin_nontemporal_load((
const __v8di_aligned *)__P);
9053 __builtin_nontemporal_store((__v8df_aligned)__A, (__v8df_aligned*)__P);
9059 typedef __v16sf __v16sf_aligned
__attribute__((aligned(64)));
9060 __builtin_nontemporal_store((__v16sf_aligned)__A, (__v16sf_aligned*)__P);
9066 return (__m512d) __builtin_ia32_compressdf512_mask ((__v8df) __A,
9074 return (__m512d) __builtin_ia32_compressdf512_mask ((__v8df) __A,
9083 return (__m512i) __builtin_ia32_compressdi512_mask ((__v8di) __A,
9091 return (__m512i) __builtin_ia32_compressdi512_mask ((__v8di) __A,
9100 return (__m512) __builtin_ia32_compresssf512_mask ((__v16sf) __A,
9108 return (__m512) __builtin_ia32_compresssf512_mask ((__v16sf) __A,
9117 return (__m512i) __builtin_ia32_compresssi512_mask ((__v16si) __A,
9125 return (__m512i) __builtin_ia32_compresssi512_mask ((__v16si) __A,
9131 #define _mm_cmp_round_ss_mask(X, Y, P, R) __extension__ ({ \
9132 (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
9133 (__v4sf)(__m128)(Y), (int)(P), \
9134 (__mmask8)-1, (int)(R)); })
9136 #define _mm_mask_cmp_round_ss_mask(M, X, Y, P, R) __extension__ ({ \
9137 (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
9138 (__v4sf)(__m128)(Y), (int)(P), \
9139 (__mmask8)(M), (int)(R)); })
9141 #define _mm_cmp_ss_mask(X, Y, P) __extension__ ({ \
9142 (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
9143 (__v4sf)(__m128)(Y), (int)(P), \
9145 _MM_FROUND_CUR_DIRECTION); })
9147 #define _mm_mask_cmp_ss_mask(M, X, Y, P) __extension__ ({ \
9148 (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
9149 (__v4sf)(__m128)(Y), (int)(P), \
9151 _MM_FROUND_CUR_DIRECTION); })
9153 #define _mm_cmp_round_sd_mask(X, Y, P, R) __extension__ ({ \
9154 (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
9155 (__v2df)(__m128d)(Y), (int)(P), \
9156 (__mmask8)-1, (int)(R)); })
9158 #define _mm_mask_cmp_round_sd_mask(M, X, Y, P, R) __extension__ ({ \
9159 (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
9160 (__v2df)(__m128d)(Y), (int)(P), \
9161 (__mmask8)(M), (int)(R)); })
9163 #define _mm_cmp_sd_mask(X, Y, P) __extension__ ({ \
9164 (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
9165 (__v2df)(__m128d)(Y), (int)(P), \
9167 _MM_FROUND_CUR_DIRECTION); })
9169 #define _mm_mask_cmp_sd_mask(M, X, Y, P) __extension__ ({ \
9170 (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
9171 (__v2df)(__m128d)(Y), (int)(P), \
9173 _MM_FROUND_CUR_DIRECTION); })
9178 return (__m512)__builtin_shufflevector((__v16sf)__A, (__v16sf)__A,
9179 1, 1, 3, 3, 5, 5, 7, 7, 9, 9, 11, 11, 13, 13, 15, 15);
9185 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
9193 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
9201 return (__m512)__builtin_shufflevector((__v16sf)__A, (__v16sf)__A,
9202 0, 0, 2, 2, 4, 4, 6, 6, 8, 8, 10, 10, 12, 12, 14, 14);
9208 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
9216 return (__m512)__builtin_ia32_selectps_512((__mmask16)__U,
9225 res[0] = (__U & 1) ? __B[0] : __W[0];
9233 res[0] = (__U & 1) ? __B[0] : 0;
9241 res[0] = (__U & 1) ? __B[0] : __W[0];
9249 res[0] = (__U & 1) ? __B[0] : 0;
9256 __builtin_ia32_storess128_mask ((__v16sf *)__W,
9258 (__mmask16) __U & (__mmask16)1);
9264 __builtin_ia32_storesd128_mask ((__v8df *)__W,
9266 (__mmask8) __U & 1);
9272 __m128 src = (__v4sf) __builtin_shufflevector((__v4sf) __W,
9273 (__v4sf) {0.0, 0.0, 0.0, 0.0},
9276 return (__m128) __builtin_shufflevector(
9277 __builtin_ia32_loadss128_mask ((__v16sf *) __A,
9279 (__mmask16) __U & 1),
9286 return (__m128) __builtin_shufflevector(
9287 __builtin_ia32_loadss128_mask ((__v16sf *) __A,
9289 (__mmask16) __U & 1),
9296 __m128d src = (__v2df) __builtin_shufflevector((__v2df) __W,
9297 (__v2df) {0.0, 0.0}, 0, 2);
9299 return (__m128d) __builtin_shufflevector(
9300 __builtin_ia32_loadsd128_mask ((__v8df *) __A,
9302 (__mmask8) __U & 1),
9309 return (__m128d) __builtin_shufflevector(
9310 __builtin_ia32_loadsd128_mask ((__v8df *) __A,
9312 (__mmask8) __U & 1),
9316 #define _mm512_shuffle_epi32(A, I) __extension__ ({ \
9317 (__m512i)__builtin_shufflevector((__v16si)(__m512i)(A), \
9318 (__v16si)_mm512_undefined_epi32(), \
9319 0 + (((I) >> 0) & 0x3), \
9320 0 + (((I) >> 2) & 0x3), \
9321 0 + (((I) >> 4) & 0x3), \
9322 0 + (((I) >> 6) & 0x3), \
9323 4 + (((I) >> 0) & 0x3), \
9324 4 + (((I) >> 2) & 0x3), \
9325 4 + (((I) >> 4) & 0x3), \
9326 4 + (((I) >> 6) & 0x3), \
9327 8 + (((I) >> 0) & 0x3), \
9328 8 + (((I) >> 2) & 0x3), \
9329 8 + (((I) >> 4) & 0x3), \
9330 8 + (((I) >> 6) & 0x3), \
9331 12 + (((I) >> 0) & 0x3), \
9332 12 + (((I) >> 2) & 0x3), \
9333 12 + (((I) >> 4) & 0x3), \
9334 12 + (((I) >> 6) & 0x3)); })
9336 #define _mm512_mask_shuffle_epi32(W, U, A, I) __extension__ ({ \
9337 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
9338 (__v16si)_mm512_shuffle_epi32((A), (I)), \
9339 (__v16si)(__m512i)(W)); })
9341 #define _mm512_maskz_shuffle_epi32(U, A, I) __extension__ ({ \
9342 (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
9343 (__v16si)_mm512_shuffle_epi32((A), (I)), \
9344 (__v16si)_mm512_setzero_si512()); })
9349 return (__m512d) __builtin_ia32_expanddf512_mask ((__v8df) __A,
9357 return (__m512d) __builtin_ia32_expanddf512_mask ((__v8df) __A,
9365 return (__m512i) __builtin_ia32_expanddi512_mask ((__v8di) __A,
9373 return (__m512i) __builtin_ia32_expanddi512_mask ((__v8di) __A,
9381 return (__m512d) __builtin_ia32_expandloaddf512_mask ((
const __v8df *)__P,
9389 return (__m512d) __builtin_ia32_expandloaddf512_mask ((
const __v8df *)__P,
9397 return (__m512i) __builtin_ia32_expandloaddi512_mask ((
const __v8di *)__P,
9405 return (__m512i) __builtin_ia32_expandloaddi512_mask ((
const __v8di *)__P,
9413 return (__m512) __builtin_ia32_expandloadsf512_mask ((
const __v16sf *)__P,
9421 return (__m512) __builtin_ia32_expandloadsf512_mask ((
const __v16sf *)__P,
9429 return (__m512i) __builtin_ia32_expandloadsi512_mask ((
const __v16si *)__P,
9437 return (__m512i) __builtin_ia32_expandloadsi512_mask ((
const __v16si *)__P,
9445 return (__m512) __builtin_ia32_expandsf512_mask ((__v16sf) __A,
9453 return (__m512) __builtin_ia32_expandsf512_mask ((__v16sf) __A,
9461 return (__m512i) __builtin_ia32_expandsi512_mask ((__v16si) __A,
9469 return (__m512i) __builtin_ia32_expandsi512_mask ((__v16si) __A,
9474 #define _mm512_cvt_roundps_pd(A, R) __extension__ ({ \
9475 (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
9476 (__v8df)_mm512_undefined_pd(), \
9477 (__mmask8)-1, (int)(R)); })
9479 #define _mm512_mask_cvt_roundps_pd(W, U, A, R) __extension__ ({ \
9480 (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
9481 (__v8df)(__m512d)(W), \
9482 (__mmask8)(U), (int)(R)); })
9484 #define _mm512_maskz_cvt_roundps_pd(U, A, R) __extension__ ({ \
9485 (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
9486 (__v8df)_mm512_setzero_pd(), \
9487 (__mmask8)(U), (int)(R)); })
9492 return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
9502 return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
9511 return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
9533 return (__m512d) __builtin_ia32_selectpd_512 ((__mmask8) __U,
9541 return (__m512d) __builtin_ia32_selectpd_512 ((__mmask8) __U,
9549 return (__m512) __builtin_ia32_selectps_512 ((__mmask16) __U,
9557 return (__m512) __builtin_ia32_selectps_512 ((__mmask16) __U,
9565 __builtin_ia32_compressstoredf512_mask ((__v8df *) __P, (__v8df) __A,
9572 __builtin_ia32_compressstoredi512_mask ((__v8di *) __P, (__v8di) __A,
9579 __builtin_ia32_compressstoresf512_mask ((__v16sf *) __P, (__v16sf) __A,
9586 __builtin_ia32_compressstoresi512_mask ((__v16si *) __P, (__v16si) __A,
9590 #define _mm_cvt_roundsd_ss(A, B, R) __extension__ ({ \
9591 (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9592 (__v2df)(__m128d)(B), \
9593 (__v4sf)_mm_undefined_ps(), \
9594 (__mmask8)-1, (int)(R)); })
9596 #define _mm_mask_cvt_roundsd_ss(W, U, A, B, R) __extension__ ({ \
9597 (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9598 (__v2df)(__m128d)(B), \
9599 (__v4sf)(__m128)(W), \
9600 (__mmask8)(U), (int)(R)); })
9602 #define _mm_maskz_cvt_roundsd_ss(U, A, B, R) __extension__ ({ \
9603 (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9604 (__v2df)(__m128d)(B), \
9605 (__v4sf)_mm_setzero_ps(), \
9606 (__mmask8)(U), (int)(R)); })
9611 return __builtin_ia32_cvtsd2ss_round_mask ((__v4sf)(__A),
9620 return __builtin_ia32_cvtsd2ss_round_mask ((__v4sf)(__A),
9626 #define _mm_cvtss_i32 _mm_cvtss_si32
9627 #define _mm_cvtsd_i32 _mm_cvtsd_si32
9628 #define _mm_cvti32_sd _mm_cvtsi32_sd
9629 #define _mm_cvti32_ss _mm_cvtsi32_ss
9631 #define _mm_cvtss_i64 _mm_cvtss_si64
9632 #define _mm_cvtsd_i64 _mm_cvtsd_si64
9633 #define _mm_cvti64_sd _mm_cvtsi64_sd
9634 #define _mm_cvti64_ss _mm_cvtsi64_ss
9638 #define _mm_cvt_roundi64_sd(A, B, R) __extension__ ({ \
9639 (__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9642 #define _mm_cvt_roundsi64_sd(A, B, R) __extension__ ({ \
9643 (__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9647 #define _mm_cvt_roundsi32_ss(A, B, R) __extension__ ({ \
9648 (__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R)); })
9650 #define _mm_cvt_roundi32_ss(A, B, R) __extension__ ({ \
9651 (__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R)); })
9654 #define _mm_cvt_roundsi64_ss(A, B, R) __extension__ ({ \
9655 (__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9658 #define _mm_cvt_roundi64_ss(A, B, R) __extension__ ({ \
9659 (__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9663 #define _mm_cvt_roundss_sd(A, B, R) __extension__ ({ \
9664 (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9665 (__v4sf)(__m128)(B), \
9666 (__v2df)_mm_undefined_pd(), \
9667 (__mmask8)-1, (int)(R)); })
9669 #define _mm_mask_cvt_roundss_sd(W, U, A, B, R) __extension__ ({ \
9670 (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9671 (__v4sf)(__m128)(B), \
9672 (__v2df)(__m128d)(W), \
9673 (__mmask8)(U), (int)(R)); })
9675 #define _mm_maskz_cvt_roundss_sd(U, A, B, R) __extension__ ({ \
9676 (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9677 (__v4sf)(__m128)(B), \
9678 (__v2df)_mm_setzero_pd(), \
9679 (__mmask8)(U), (int)(R)); })
9684 return __builtin_ia32_cvtss2sd_round_mask((__v2df)(__A),
9693 return __builtin_ia32_cvtss2sd_round_mask((__v2df)(__A),
9702 return (__m128d) __builtin_ia32_cvtusi2sd32 ((__v2df) __A, __B);
9706 #define _mm_cvt_roundu64_sd(A, B, R) __extension__ ({ \
9707 (__m128d)__builtin_ia32_cvtusi2sd64((__v2df)(__m128d)(A), \
9708 (unsigned long long)(B), (int)(R)); })
9711 _mm_cvtu64_sd (__m128d __A,
unsigned long long __B)
9713 return (__m128d) __builtin_ia32_cvtusi2sd64 ((__v2df) __A, __B,
9718 #define _mm_cvt_roundu32_ss(A, B, R) __extension__ ({ \
9719 (__m128)__builtin_ia32_cvtusi2ss32((__v4sf)(__m128)(A), (unsigned int)(B), \
9725 return (__m128) __builtin_ia32_cvtusi2ss32 ((__v4sf) __A, __B,
9730 #define _mm_cvt_roundu64_ss(A, B, R) __extension__ ({ \
9731 (__m128)__builtin_ia32_cvtusi2ss64((__v4sf)(__m128)(A), \
9732 (unsigned long long)(B), (int)(R)); })
9735 _mm_cvtu64_ss (__m128 __A,
unsigned long long __B)
9737 return (__m128) __builtin_ia32_cvtusi2ss64 ((__v4sf) __A, __B,
9745 return (__m512i) __builtin_ia32_pbroadcastd512_gpr_mask (__A, (__v16si) __O,
9751 _mm512_mask_set1_epi64 (__m512i __O, __mmask8 __M,
long long __A)
9753 return (__m512i) __builtin_ia32_pbroadcastq512_gpr_mask (__A, (__v8di) __O,
9760 char __e58,
char __e57,
char __e56,
char __e55,
char __e54,
char __e53,
9761 char __e52,
char __e51,
char __e50,
char __e49,
char __e48,
char __e47,
9762 char __e46,
char __e45,
char __e44,
char __e43,
char __e42,
char __e41,
9763 char __e40,
char __e39,
char __e38,
char __e37,
char __e36,
char __e35,
9764 char __e34,
char __e33,
char __e32,
char __e31,
char __e30,
char __e29,
9765 char __e28,
char __e27,
char __e26,
char __e25,
char __e24,
char __e23,
9766 char __e22,
char __e21,
char __e20,
char __e19,
char __e18,
char __e17,
9767 char __e16,
char __e15,
char __e14,
char __e13,
char __e12,
char __e11,
9768 char __e10,
char __e9,
char __e8,
char __e7,
char __e6,
char __e5,
9769 char __e4,
char __e3,
char __e2,
char __e1,
char __e0) {
9771 return __extension__ (__m512i)(__v64qi)
9772 {__e0, __e1, __e2, __e3, __e4, __e5, __e6, __e7,
9773 __e8, __e9, __e10, __e11, __e12, __e13, __e14, __e15,
9774 __e16, __e17, __e18, __e19, __e20, __e21, __e22, __e23,
9775 __e24, __e25, __e26, __e27, __e28, __e29, __e30, __e31,
9776 __e32, __e33, __e34, __e35, __e36, __e37, __e38, __e39,
9777 __e40, __e41, __e42, __e43, __e44, __e45, __e46, __e47,
9778 __e48, __e49, __e50, __e51, __e52, __e53, __e54, __e55,
9779 __e56, __e57, __e58, __e59, __e60, __e61, __e62, __e63};
9784 short __e27,
short __e26,
short __e25,
short __e24,
short __e23,
9785 short __e22,
short __e21,
short __e20,
short __e19,
short __e18,
9786 short __e17,
short __e16,
short __e15,
short __e14,
short __e13,
9787 short __e12,
short __e11,
short __e10,
short __e9,
short __e8,
9788 short __e7,
short __e6,
short __e5,
short __e4,
short __e3,
9789 short __e2,
short __e1,
short __e0) {
9790 return __extension__ (__m512i)(__v32hi)
9791 {__e0, __e1, __e2, __e3, __e4, __e5, __e6, __e7,
9792 __e8, __e9, __e10, __e11, __e12, __e13, __e14, __e15,
9793 __e16, __e17, __e18, __e19, __e20, __e21, __e22, __e23,
9794 __e24, __e25, __e26, __e27, __e28, __e29, __e30, __e31 };
9799 int __E,
int __F,
int __G,
int __H,
9800 int __I,
int __J,
int __K,
int __L,
9801 int __M,
int __N,
int __O,
int __P)
9803 return __extension__ (__m512i)(__v16si)
9804 { __P, __O, __N, __M, __L, __K, __J, __I,
9805 __H, __G, __F, __E, __D, __C, __B, __A };
9808 #define _mm512_setr_epi32(e0,e1,e2,e3,e4,e5,e6,e7, \
9809 e8,e9,e10,e11,e12,e13,e14,e15) \
9810 _mm512_set_epi32((e15),(e14),(e13),(e12),(e11),(e10),(e9),(e8),(e7),(e6), \
9811 (e5),(e4),(e3),(e2),(e1),(e0))
9815 long long __D,
long long __E,
long long __F,
9816 long long __G,
long long __H)
9818 return __extension__ (__m512i) (__v8di)
9819 { __H, __G, __F, __E, __D, __C, __B, __A };
9822 #define _mm512_setr_epi64(e0,e1,e2,e3,e4,e5,e6,e7) \
9823 _mm512_set_epi64((e7),(e6),(e5),(e4),(e3),(e2),(e1),(e0))
9827 double __E,
double __F,
double __G,
double __H)
9829 return __extension__ (__m512d)
9830 { __H, __G, __F, __E, __D, __C, __B, __A };
9833 #define _mm512_setr_pd(e0,e1,e2,e3,e4,e5,e6,e7) \
9834 _mm512_set_pd((e7),(e6),(e5),(e4),(e3),(e2),(e1),(e0))
9838 float __E,
float __F,
float __G,
float __H,
9839 float __I,
float __J,
float __K,
float __L,
9840 float __M,
float __N,
float __O,
float __P)
9842 return __extension__ (__m512)
9843 { __P, __O, __N, __M, __L, __K, __J, __I,
9844 __H, __G, __F, __E, __D, __C, __B, __A };
9847 #define _mm512_setr_ps(e0,e1,e2,e3,e4,e5,e6,e7,e8,e9,e10,e11,e12,e13,e14,e15) \
9848 _mm512_set_ps((e15),(e14),(e13),(e12),(e11),(e10),(e9),(e8),(e7),(e6),(e5), \
9849 (e4),(e3),(e2),(e1),(e0))
9889 #define _mm512_reduce_operator_64bit(Vec512, Operator, T2, T1) \
9891 __m256##T1 Vec256 = __builtin_shufflevector( \
9892 (__v8d##T2)Vec512, \
9893 (__v8d##T2)Vec512, \
9896 __builtin_shufflevector( \
9897 (__v8d##T2)Vec512, \
9898 (__v8d##T2)Vec512, \
9900 __m128##T1 Vec128 = __builtin_shufflevector( \
9901 (__v4d##T2)Vec256, \
9902 (__v4d##T2)Vec256, \
9905 __builtin_shufflevector( \
9906 (__v4d##T2)Vec256, \
9907 (__v4d##T2)Vec256, \
9909 Vec128 = __builtin_shufflevector((__v2d##T2)Vec128, \
9910 (__v2d##T2)Vec128, 0, -1) \
9912 __builtin_shufflevector((__v2d##T2)Vec128, \
9913 (__v2d##T2)Vec128, 1, -1); \
9950 #define _mm512_mask_reduce_operator_64bit(Vec512, Vec512Neutral, Operator, \
9953 Vec512 = __builtin_ia32_select##T3##_512( \
9955 (__v8d##T2)Vec512, \
9956 (__v8d##T2)Vec512Neutral); \
9957 _mm512_reduce_operator_64bit(Vec512, Operator, T2, T1); \
9999 #define _mm512_reduce_operator_32bit(Vec512, Operator, T2, T1) __extension__({ \
10000 __m256##T1 Vec256 = \
10001 (__m256##T1)(__builtin_shufflevector( \
10002 (__v16s##T2)Vec512, \
10003 (__v16s##T2)Vec512, \
10004 0, 1, 2, 3, 4, 5, 6, 7) \
10006 __builtin_shufflevector( \
10007 (__v16s##T2)Vec512, \
10008 (__v16s##T2)Vec512, \
10009 8, 9, 10, 11, 12, 13, 14, 15)); \
10010 __m128##T1 Vec128 = \
10011 (__m128##T1)(__builtin_shufflevector( \
10012 (__v8s##T2)Vec256, \
10013 (__v8s##T2)Vec256, \
10016 __builtin_shufflevector( \
10017 (__v8s##T2)Vec256, \
10018 (__v8s##T2)Vec256, \
10020 Vec128 = (__m128##T1)(__builtin_shufflevector( \
10021 (__v4s##T2)Vec128, \
10022 (__v4s##T2)Vec128, \
10025 __builtin_shufflevector( \
10026 (__v4s##T2)Vec128, \
10027 (__v4s##T2)Vec128, \
10029 Vec128 = (__m128##T1)(__builtin_shufflevector( \
10030 (__v4s##T2)Vec128, \
10031 (__v4s##T2)Vec128, \
10034 __builtin_shufflevector( \
10035 (__v4s##T2)Vec128, \
10036 (__v4s##T2)Vec128, \
10038 return Vec128[0]; \
10080 #define _mm512_mask_reduce_operator_32bit(Vec512, Vec512Neutral, Operator, \
10081 Mask, T2, T1, T3) \
10083 Vec512 = (__m512##T1)__builtin_ia32_select##T3##_512( \
10085 (__v16s##T2)Vec512, \
10086 (__v16s##T2)Vec512Neutral); \
10087 _mm512_reduce_operator_32bit(Vec512, Operator, T2, T1); \
10132 #define _mm512_reduce_maxMin_64bit(Vec512, IntrinName, T1, T2) __extension__({ \
10133 Vec512 = _mm512_##IntrinName( \
10134 (__m512##T1)__builtin_shufflevector( \
10135 (__v8d##T2)Vec512, \
10136 (__v8d##T2)Vec512, \
10137 0, 1, 2, 3, -1, -1, -1, -1), \
10138 (__m512##T1)__builtin_shufflevector( \
10139 (__v8d##T2)Vec512, \
10140 (__v8d##T2)Vec512, \
10141 4, 5, 6, 7, -1, -1, -1, -1)); \
10142 Vec512 = _mm512_##IntrinName( \
10143 (__m512##T1)__builtin_shufflevector( \
10144 (__v8d##T2)Vec512, \
10145 (__v8d##T2)Vec512, \
10146 0, 1, -1, -1, -1, -1, -1, -1),\
10147 (__m512##T1)__builtin_shufflevector( \
10148 (__v8d##T2)Vec512, \
10149 (__v8d##T2)Vec512, \
10150 2, 3, -1, -1, -1, -1, -1, \
10152 Vec512 = _mm512_##IntrinName( \
10153 (__m512##T1)__builtin_shufflevector( \
10154 (__v8d##T2)Vec512, \
10155 (__v8d##T2)Vec512, \
10156 0, -1, -1, -1, -1, -1, -1, -1),\
10157 (__m512##T1)__builtin_shufflevector( \
10158 (__v8d##T2)Vec512, \
10159 (__v8d##T2)Vec512, \
10160 1, -1, -1, -1, -1, -1, -1, -1))\
10162 return Vec512[0]; \
10212 #define _mm512_mask_reduce_maxMin_64bit(Vec512, Vec512Neutral, IntrinName, T1, \
10215 Vec512 = (__m512##T1)__builtin_ia32_select##T3##_512( \
10217 (__v8d##T2)Vec512, \
10218 (__v8d##T2)Vec512Neutral); \
10219 _mm512_reduce_maxMin_64bit(Vec512, IntrinName, T1, T2); \
10225 max_epi64, i, i, q, __M);
10231 max_epu64, i, i, q, __M);
10237 max_pd, d, f, pd, __M);
10243 min_epi64, i, i, q, __M);
10249 min_epu64, i, i, q, __M);
10255 min_pd, d, f, pd, __M);
10264 #define _mm512_reduce_maxMin_32bit(Vec512, IntrinName, T1, T2) __extension__({ \
10265 Vec512 = _mm512_##IntrinName( \
10266 (__m512##T1)__builtin_shufflevector( \
10267 (__v16s##T2)Vec512, \
10268 (__v16s##T2)Vec512, \
10269 0, 1, 2, 3, 4, 5, 6, 7, \
10270 -1, -1, -1, -1, -1, -1, -1, -1), \
10271 (__m512##T1)__builtin_shufflevector( \
10272 (__v16s##T2)Vec512, \
10273 (__v16s##T2)Vec512, \
10274 8, 9, 10, 11, 12, 13, 14, 15, \
10275 -1, -1, -1, -1, -1, -1, -1, -1)); \
10276 Vec512 = _mm512_##IntrinName( \
10277 (__m512##T1)__builtin_shufflevector( \
10278 (__v16s##T2)Vec512, \
10279 (__v16s##T2)Vec512, \
10280 0, 1, 2, 3, -1, -1, -1, -1, \
10281 -1, -1, -1, -1, -1, -1, -1, -1), \
10282 (__m512##T1)__builtin_shufflevector( \
10283 (__v16s##T2)Vec512, \
10284 (__v16s##T2)Vec512, \
10285 4, 5, 6, 7, -1, -1, -1, -1, \
10286 -1, -1, -1, -1, -1, -1, -1, -1)); \
10287 Vec512 = _mm512_##IntrinName( \
10288 (__m512##T1)__builtin_shufflevector( \
10289 (__v16s##T2)Vec512, \
10290 (__v16s##T2)Vec512, \
10291 0, 1, -1, -1, -1, -1, -1, -1, \
10292 -1, -1, -1, -1, -1, -1, -1, -1), \
10293 (__m512##T1)__builtin_shufflevector( \
10294 (__v16s##T2)Vec512, \
10295 (__v16s##T2)Vec512, \
10296 2, 3, -1, -1, -1, -1, -1, -1, \
10297 -1, -1, -1, -1, -1, -1, -1, -1)); \
10298 Vec512 = _mm512_##IntrinName( \
10299 (__m512##T1)__builtin_shufflevector( \
10300 (__v16s##T2)Vec512, \
10301 (__v16s##T2)Vec512, \
10302 0, -1, -1, -1, -1, -1, -1, -1, \
10303 -1, -1, -1, -1, -1, -1, -1, -1), \
10304 (__m512##T1)__builtin_shufflevector( \
10305 (__v16s##T2)Vec512, \
10306 (__v16s##T2)Vec512, \
10307 1, -1, -1, -1, -1, -1, -1, -1, \
10308 -1, -1, -1, -1, -1, -1, -1, -1)); \
10309 return Vec512[0]; \
10355 #define _mm512_mask_reduce_maxMin_32bit(Vec512, Vec512Neutral, IntrinName, T1, \
10358 Vec512 = (__m512##T1)__builtin_ia32_select##T3##_512( \
10360 (__v16s##T2)Vec512, \
10361 (__v16s##T2)Vec512Neutral); \
10362 _mm512_reduce_maxMin_32bit(Vec512, IntrinName, T1, T2); \
10401 #undef __DEFAULT_FN_ATTRS
10403 #endif // __AVX512FINTRIN_H
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_castsi128_si512(__m128i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_sqrt_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_cvtph_ps(__m256i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_cvtps_pd(__mmask8 __U, __m256 __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtepi64_epi8(__m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepu8_epi64(__m128i __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpeq_epi32_mask(__m512i __a, __m512i __b)
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_store_pd(void *__P, __mmask8 __U, __m512d __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpgt_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
#define _mm512_reduce_maxMin_64bit(Vec512, IntrinName, T1, T2)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_mullo_epi32(__mmask16 __M, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_permutexvar_epi32(__mmask16 __M, __m512i __X, __m512i __Y)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fnmsub_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_cvtps_pd(__m512d __W, __mmask8 __U, __m256 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_expand_epi32(__m512i __W, __mmask16 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_xor_si512(__m512i __a, __m512i __b)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_epi8(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_add_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_broadcast_i64x4(__m256i __A)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_reduce_mul_ps(__m512 __W)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fmadd_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_unpackhi_ps(__m512 __a, __m512 __b)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_reduce_min_pd(__m512d __V)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvttpd_epi32(__mmask8 __U, __m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_cvtepu32_ps(__m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_and_epi32(__m512i __src, __mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sub_epi32(__m512i __A, __m512i __B)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpeq_epi64_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_permutex2var_epi64(__m512i __A, __mmask8 __U, __m512i __I, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_blend_epi64(__mmask8 __U, __m512i __A, __m512i __W)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sll_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m128i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fmsub_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kor(__mmask16 __A, __mmask16 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_mul_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_broadcast_i32x4(__m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepi32_epi64(__m256i __X)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_storeu_epi8(void *__P, __mmask16 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_sub_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mul_epu32(__m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_sqrt_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_setzero_ps(void)
Constructs a 256-bit floating-point vector of [8 x float] with all vector elements initialized to zer...
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_sub_ps(__m512 __a, __m512 __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fmaddsub_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_rsqrt14_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_expandloadu_ps(__m512 __W, __mmask16 __U, void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_set_epi64(long long __A, long long __B, long long __C, long long __D, long long __E, long long __F, long long __G, long long __H)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_setzero_pd(void)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_rolv_epi64(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_add_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_storeu_pd(void *__P, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepu16_epi64(__m512i __W, __mmask8 __U, __m128i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_broadcastsd_pd(__m128d __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kunpackb(__mmask16 __A, __mmask16 __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_floor_pd(__m512d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fnmsub_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_scalef_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_fmadd_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fnmsub_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_div_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_mul_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvttpd_epu32(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_max_epu64(__mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_compress_epi32(__m512i __W, __mmask16 __U, __m512i __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_min_epu64(__m512i __A, __m512i __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_scalef_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sllv_epi32(__m512i __W, __mmask16 __U, __m512i __X, __m512i __Y)
static __inline __m128 __DEFAULT_FN_ATTRS _mm512_castps512_ps128(__m512 __a)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtusepi64_epi32(__m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi32_ps(__mmask16 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srl_epi32(__m512i __A, __m128i __B)
#define _mm512_mask_reduce_operator_64bit(Vec512, Vec512Neutral, Operator,Mask, T2, T1, T3)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fmsubadd_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_cvtepu32lo_pd(__m512d __W, __mmask8 __U, __m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_stream_si512(__m512i *__P, __m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fmsubadd_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_broadcast_i32x4(__mmask16 __M, __m128i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_unpacklo_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_broadcast_f32x4(__m512 __O, __mmask16 __M, __m128 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srai_epi64(__m512i __A, int __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kxnor(__mmask16 __A, __mmask16 __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_max_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_broadcast_f64x4(__m256d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_getexp_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask2_permutex2var_epi64(__m512i __A, __m512i __I, __mmask8 __U, __m512i __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_store_si512(void *__P, __m512i __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpgt_epi32_mask(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_broadcastsd_pd(__mmask8 __M, __m128d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sub_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_max_epu64(__m512i __W, __mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask3_fmsub_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_min_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_and_epi32(__m512i __W)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtsepi64_epi32(__m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_permutevar_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512i __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_floor_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_mov_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_rsqrt14_ps(__m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_expandloadu_epi64(__mmask8 __U, void const *__P)
static __inline __m256 __DEFAULT_FN_ATTRS _mm512_castps512_ps256(__m512 __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_zextpd256_pd512(__m256d __a)
Constructs a 512-bit floating-point vector of [8 x double] from a 256-bit floating-point vector of [4...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_load_sd(__mmask8 __U, const double *__A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_scalef_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask3_fmsub_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_and_epi64(__mmask8 __k, __m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set1_epi8(char __w)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu16_epi64(__mmask8 __U, __m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu8_epi64(__mmask8 __U, __m128i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmple_epu64_mask(__m512i __a, __m512i __b)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_epi8(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_cvtph_ps(__m512 __W, __mmask16 __U, __m256i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_stream_ps(float *__P, __m512 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fnmsub_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_div_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_min_epi64(__m512i __W, __mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_xor_epi32(__m512i __src, __mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtusepi64_epi8(__m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_storeu_epi8(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_or_epi64(__mmask8 __k, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srl_epi64(__mmask8 __U, __m512i __A, __m128i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_castps_si512(__m512 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_scalef_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_broadcast_f64x4(__mmask8 __M, __m256d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask2_permutex2var_pd(__m512d __A, __m512i __I, __mmask8 __U, __m512d __B)
static __inline __m256i __DEFAULT_FN_ATTRS _mm512_castsi512_si256(__m512i __A)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_cvtss_f32(__m512 __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_broadcastd_epi32(__mmask16 __M, __m128i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpeq_epu64_mask(__m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_rcp14_ps(__mmask16 __U, __m512 __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpgt_epu64_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_andnot_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_slli_epi64(__m512i __A, int __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_store_ps(void *__P, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_xor_epi64(__mmask8 __k, __m512i __a, __m512i __b)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_max_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fmadd_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_zextps128_ps512(__m128 __a)
Constructs a 512-bit floating-point vector of [16 x float] from a 128-bit floating-point vector of [4...
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_max_epi64(__m512i __V)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fmaddsub_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_maskz_load_ps(__mmask16 __U, void const *__P)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fmsub_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_move_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_expandloadu_epi64(__m512i __W, __mmask8 __U, void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sra_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m128i __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_fnmsub_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_loadu_epi64(__mmask8 __U, void const *__P)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_storeu_epi8(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_sqrt_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_and_epi32(__m512i __a, __m512i __b)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_setzero_ps(void)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_expand_ps(__mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srl_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m128i __B)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_unpacklo_ps(__m512 __a, __m512 __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sllv_epi32(__mmask16 __U, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_min_epu32(__m512i __W, __mmask16 __M, __m512i __A, __m512i __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_kortestz(__mmask16 __A, __mmask16 __B)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtsepi64_epi8(__m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_fmadd_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_blend_epi32(__mmask16 __U, __m512i __A, __m512i __W)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_min_epu32(__mmask16 __M, __m512i __A, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set1_epi32(int __s)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_min_ps(__m512 __A, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srlv_epi32(__m512i __X, __m512i __Y)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmplt_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set_epi16(short __e31, short __e30, short __e29, short __e28, short __e27, short __e26, short __e25, short __e24, short __e23, short __e22, short __e21, short __e20, short __e19, short __e18, short __e17, short __e16, short __e15, short __e14, short __e13, short __e12, short __e11, short __e10, short __e9, short __e8, short __e7, short __e6, short __e5, short __e4, short __e3, short __e2, short __e1, short __e0)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpneq_epu32_mask(__m512i __a, __m512i __b)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpge_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_max_epi64(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_rolv_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_mul_epi32(__m512i __W)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_test_epi64_mask(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_min_epi64(__m512i __V)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_min_epi32(__m512i __W, __mmask16 __M, __m512i __A, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_permutex2var_epi64(__m512i __A, __m512i __I, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpgt_epu32_mask(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_unpackhi_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_compress_pd(__mmask8 __U, __m512d __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_max_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_undefined_ps(void)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_permutexvar_ps(__m512i __X, __m512 __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi16_epi64(__mmask8 __U, __m128i __A)
static __inline void __DEFAULT_FN_ATTRS _mm512_store_epi64(void *__P, __m512i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_add_ps(__m512 __a, __m512 __b)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_set4_ps(float __A, float __B, float __C, float __D)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_castpd_ps(__m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sllv_epi64(__m512i __W, __mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_movehdup_ps(__m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sll_epi64(__m512i __A, __m128i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srav_epi32(__mmask16 __U, __m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_sqrt_pd(__m512d __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_permutexvar_epi32(__m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fnmsub_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_set1_epi64(__mmask8 __M, long long __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepi16_epi32(__m256i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_or_epi64(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srav_epi64(__m512i __W, __mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_unpackhi_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_movehdup_ps(__mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srav_epi32(__m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu32_epi64(__mmask8 __U, __m256i __X)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_load_ss(__m128 __W, __mmask8 __U, const float *__A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_fnmsub_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_max_epu32(__mmask16 __M, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sllv_epi32(__m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fmsub_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_abs_pd(__m512d __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpge_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_rcp14_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_cvtps_pd(__m256 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_max_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtsepi32_epi16(__m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_rcp14_pd(__m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sub_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_andnot_epi64(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_broadcastq_epi64(__m512i __O, __mmask8 __M, __m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_expand_epi64(__mmask8 __U, __m512i __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmple_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_undefined_ps(void)
Create a 128-bit vector of [4 x float] with undefined values.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask3_fnmadd_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sra_epi32(__mmask16 __U, __m512i __A, __m128i __B)
static __inline __m128i __DEFAULT_FN_ATTRS _mm512_castsi512_si128(__m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_rsqrt14_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_compressstoreu_ps(void *__P, __mmask16 __U, __m512 __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_min_epi64(__m512i __A, __m512i __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_cvtepu32_pd(__m256i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_ps(__m512 __W, __mmask16 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_rorv_epi64(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_rcp14_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_unpackhi_epi32(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_abs_epi64(__mmask8 __U, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_fnmadd_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpgt_epi64_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sllv_epi64(__mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpneq_epi64_mask(__m512i __a, __m512i __b)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_mask_load_pd(__m512d __W, __mmask8 __U, void const *__P)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvttpd_epi32(__m256i __W, __mmask8 __U, __m512d __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kand(__mmask16 __A, __mmask16 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sll_epi32(__m512i __A, __m128i __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_rcp14_sd(__m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_andnot_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_fmsub_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_slli_epi32(__mmask16 __U, __m512i __A, int __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask3_fnmsub_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_permutex2var_ps(__mmask16 __U, __m512 __A, __m512i __I, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_rorv_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_permutexvar_epi32(__m512i __W, __mmask16 __M, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srav_epi32(__m512i __W, __mmask16 __U, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srl_epi32(__mmask16 __U, __m512i __A, __m128i __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_set1_pd(double __w)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtpd_epu32(__m256i __W, __mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sra_epi64(__m512i __A, __m128i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_mul_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fnmsub_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_cvtepi32lo_pd(__m512i __A)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_add_epi32(__mmask16 __M, __m512i __W)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi32_epi8(__mmask16 __M, __m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_compressstoreu_epi32(void *__P, __mmask16 __U, __m512i __A)
static __inline__ unsigned long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_epu64(__mmask8 __M, __m512i __V)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_castsi256_si512(__m256i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_slli_epi32(__m512i __A, int __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtps_epi32(__mmask16 __U, __m512 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fnmsub_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srli_epi64(__mmask8 __U, __m512i __A, int __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_maskz_loadu_pd(__mmask8 __U, void const *__P)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi64_epi32(__mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fmadd_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_abs_pd(__m512d __W, __mmask8 __K, __m512d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_ceil_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtusepi64_epi16(__m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_sqrt_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_storeu_epi64(void *__P, __mmask8 __U, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fmadd_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_loadu_epi32(__mmask16 __U, void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_load_epi32(__mmask16 __U, void const *__P)
#define _mm512_reduce_operator_32bit(Vec512, Operator, T2, T1)
static __inline__ __m256 __DEFAULT_FN_ATTRS _mm512_maskz_cvtpd_ps(__mmask8 __U, __m512d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_expandloadu_pd(__mmask8 __U, void const *__P)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_expand_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_min_epu64(__mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_mul_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srai_epi64(__mmask8 __U, __m512i __A, int __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_andnot_epi32(__m512i __A, __m512i __B)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_epi8(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_min_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu32_pd(__mmask8 __U, __m256i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fmsubadd_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_unpacklo_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_cvtph_ps(__mmask16 __U, __m256i __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi64_epi8(__mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_rcp14_pd(__mmask8 __U, __m512d __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_min_epi32(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_add_epi64(__m512i __A, __m512i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_add_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srlv_epi64(__m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_abs_epi32(__mmask16 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_expand_epi64(__m512i __W, __mmask8 __U, __m512i __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtsepi64_epi16(__m512i __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtusepi32_epi8(__mmask16 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_slli_epi64(__m512i __W, __mmask8 __U, __m512i __A, int __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm256_undefined_si256(void)
Create a 256-bit integer vector with undefined values.
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmplt_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_div_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srai_epi32(__m512i __W, __mmask16 __U, __m512i __A, int __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_max_pd(__m512d __A, __m512d __B)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpgt_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_reduce_max_pd(__m512d __V)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_castpd128_pd512(__m128d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sra_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m128i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_mov_epi64(__mmask8 __U, __m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_storeu_epi32(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_movedup_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_unpackhi_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpeq_epu32_mask(__m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set_epi32(int __A, int __B, int __C, int __D, int __E, int __F, int __G, int __H, int __I, int __J, int __K, int __L, int __M, int __N, int __O, int __P)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi32_storeu_epi8(void *__P, __mmask16 __M, __m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_sub_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_rolv_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi32_pd(__mmask8 __U, __m256i __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi32_epi16(__mmask16 __M, __m512i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_load_ps(void const *__p)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_cvtpd_pslo(__m512d __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtusepi32_epi16(__m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fnmadd_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvttps_epu32(__mmask16 __U, __m512 __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_div_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_castsi512_pd(__m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_unpackhi_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_max_epi64(__m512i __W, __mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_div_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_add_epi64(__mmask8 __M, __m512i __W)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_permutex2var_pd(__m512d __A, __mmask8 __U, __m512i __I, __m512d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_rcp14_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_movehdup_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepu32_epi64(__m256i __X)
static __inline__ int __DEFAULT_FN_ATTRS _mm_cvttss_i32(__m128 __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_getexp_sd(__m128d __A, __m128d __B)
static __inline __mmask8 __DEFAULT_FN_ATTRS _mm512_test_epi64_mask(__m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_max_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_testn_epi64_mask(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_xor_epi32(__mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask3_fnmsub_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_cvtpslo_pd(__m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sll_epi32(__mmask16 __U, __m512i __A, __m128i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpgt_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_loadu_pd(void const *__p)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_permutevar_pd(__mmask8 __U, __m512d __A, __m512i __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_max_epi32(__m512i __W, __mmask16 __M, __m512i __A, __m512i __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_store_epi32(void *__P, __mmask16 __U, __m512i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_zextps256_ps512(__m256 __a)
Constructs a 512-bit floating-point vector of [16 x float] from a 256-bit floating-point vector of [8...
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepi16_epi64(__m128i __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtpd_epi32(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_or_si512(__m512i __a, __m512i __b)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_reduce_add_pd(__m512d __W)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_storeu_epi32(void *__P, __mmask8 __M, __m512i __A)
char __v64qi __attribute__((__vector_size__(64)))
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_broadcastss_ps(__mmask16 __M, __m128 __A)
static __inline __mmask16 __DEFAULT_FN_ATTRS _mm512_knot(__mmask16 __M)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_undefined_epi32(void)
#define _mm512_mask_reduce_maxMin_32bit(Vec512, Vec512Neutral, IntrinName, T1, T2, T3, Mask)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srlv_epi32(__m512i __W, __mmask16 __U, __m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_pd(__m512d __W, __mmask8 __U, __m256i __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_zextsi256_si512(__m256i __a)
Constructs a 512-bit integer vector from a 256-bit integer vector.
static __inline__ float __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_ps(__mmask16 __M, __m512 __V)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask3_fmadd_sd(__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_set1_epi32(__mmask16 __M, int __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_epi64(__mmask8 __M, __m512i __V)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_mul_epi32(__m512i __W, __mmask8 __M, __m512i __X, __m512i __Y)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_reduce_max_ps(__m512 a)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_loadu_epi32(__m512i __W, __mmask16 __U, void const *__P)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpge_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_rolv_epi32(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_unpacklo_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_unpackhi_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_scalef_ss(__m128 __A, __m128 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_scalef_pd(__m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtps_epu32(__m512 __A)
static __inline__ uint32_t volatile uint32_t * __p
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fmadd_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_unpackhi_pd(__m512d __a, __m512d __b)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_storeu_epi8(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_rsqrt14_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_scalef_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_unpackhi_epi64(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_scalef_ps(__m512 __A, __m512 __B)
static __inline__ unsigned int __DEFAULT_FN_ATTRS _mm512_reduce_max_epu32(__m512i a)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_floor_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_storeu_epi16(void *__P, __mmask16 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_setzero_ps(void)
Constructs a 128-bit floating-point vector of [4 x float] initialized to zero.
static __inline__ __m256 __DEFAULT_FN_ATTRS _mm512_cvtpd_ps(__m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_add_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepu32_epi64(__m512i __W, __mmask8 __U, __m256i __X)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtps_epu32(__m512i __W, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtps_epu32(__mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_abs_epi64(__m512i __W, __mmask8 __U, __m512i __A)
static __inline__ unsigned __DEFAULT_FN_ATTRS _mm_cvttss_u32(__m128 __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_loadu_epi64(__m512i __W, __mmask8 __U, void const *__P)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_loadu_ps(void const *__p)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srli_epi32(__mmask16 __U, __m512i __A, int __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_moveldup_ps(__m512 __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpge_epi32_mask(__m512i __a, __m512i __b)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmplt_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline void __DEFAULT_FN_ATTRS _mm512_storeu_ps(void *__P, __m512 __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_abs_epi64(__m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fmaddsub_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_fmadd_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_min_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_broadcastsd_pd(__m512d __O, __mmask8 __M, __m128d __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_mask_loadu_ps(__m512 __W, __mmask16 __U, void const *__P)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_setzero_pd(void)
Constructs a 128-bit floating-point vector of [2 x double] initialized to zero.
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_mul_epi32(__mmask16 __M, __m512i __W)
static __inline__ __m256 __DEFAULT_FN_ATTRS _mm256_undefined_ps(void)
Create a 256-bit vector of [8 x float] with undefined values.
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_storeu_ps(void *__P, __mmask16 __U, __m512 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fnmadd_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_set1_ps(float __w)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_permutexvar_ps(__mmask16 __U, __m512i __X, __m512 __Y)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_sqrt_ps(__m512 __a)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_fmadd_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_add_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_min_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kandn(__mmask16 __A, __mmask16 __B)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_ps(__mmask16 __M, __m512 __V)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_mov_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_mul_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_load_epi64(__mmask8 __U, void const *__P)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtepi64_epi16(__m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepu8_epi32(__m512i __W, __mmask16 __U, __m128i __A)
static __inline__ unsigned __DEFAULT_FN_ATTRS _mm_cvttsd_u32(__m128d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fmadd_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_cvtsd_ss(__mmask8 __U, __m128 __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_permutex2var_epi64(__mmask8 __U, __m512i __A, __m512i __I, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_move_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_unpacklo_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_mask_reduce_mul_pd(__mmask8 __M, __m512d __W)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fmaddsub_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_or_epi32(__mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_movedup_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_cvttps_epi32(__m512 __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sll_epi64(__mmask8 __U, __m512i __A, __m128i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_broadcastq_epi64(__mmask8 __M, __m128i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_moveldup_ps(__mmask16 __U, __m512 __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmplt_epu64_mask(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_sqrt_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fnmsub_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_cvtpslo_pd(__m512d __W, __mmask8 __U, __m512 __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_add_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_max_epu32(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fmaddsub_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_epi64(__m512i __W, __mmask8 __U, __m256i __X)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set_epi8(char __e63, char __e62, char __e61, char __e60, char __e59, char __e58, char __e57, char __e56, char __e55, char __e54, char __e53, char __e52, char __e51, char __e50, char __e49, char __e48, char __e47, char __e46, char __e45, char __e44, char __e43, char __e42, char __e41, char __e40, char __e39, char __e38, char __e37, char __e36, char __e35, char __e34, char __e33, char __e32, char __e31, char __e30, char __e29, char __e28, char __e27, char __e26, char __e25, char __e24, char __e23, char __e22, char __e21, char __e20, char __e19, char __e18, char __e17, char __e16, char __e15, char __e14, char __e13, char __e12, char __e11, char __e10, char __e9, char __e8, char __e7, char __e6, char __e5, char __e4, char __e3, char __e2, char __e1, char __e0)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_min_epu32(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fnmadd_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_or_epi32(__m512i __W)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_unpacklo_epi64(__m512i __A, __m512i __B)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_and_epi64(__m512i __W)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_cvtsd_f64(__m512d __a)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_cvtpd_pslo(__m512 __W, __mmask8 __U, __m512d __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmplt_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvttps_epi32(__m512i __W, __mmask16 __U, __m512 __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmple_epi32_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_or_epi32(__m512i __a, __m512i __b)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_mask_reduce_mul_ps(__mmask16 __M, __m512 __W)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32lo_pd(__m512d __W, __mmask8 __U, __m512i __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mullo_epi32(__m512i __A, __m512i __B)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_testn_epi64_mask(__m512i __A, __m512i __B)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_epi16(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_min_pd(__m512d __A, __m512d __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_getexp_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_unpackhi_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set1_epi16(short __w)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmple_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_rsqrt14_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_add_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_int2mask(int __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_mov_epi32(__m512i __W, __mmask16 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_and_epi64(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_permutexvar_pd(__m512d __W, __mmask8 __U, __m512i __X, __m512d __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_rorv_epi32(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_permutexvar_pd(__mmask8 __U, __m512i __X, __m512d __Y)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtsepi32_epi8(__mmask16 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_permutex2var_pd(__mmask8 __U, __m512d __A, __m512i __I, __m512d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_compress_ps(__mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_permutex2var_epi32(__mmask16 __U, __m512i __A, __m512i __I, __m512i __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtepi64_epi32(__m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_compressstoreu_epi64(void *__P, __mmask8 __U, __m512i __A)
static __inline__ __m256 __DEFAULT_FN_ATTRS _mm512_mask_cvtpd_ps(__m256 __W, __mmask8 __U, __m512d __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_cvtss_sd(__mmask8 __U, __m128d __A, __m128 __B)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_pd(__mmask8 __M, __m512d __V)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi8_epi64(__mmask8 __U, __m128i __A)
static __inline__ vector float vector float __b
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_cvtss_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128 __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvttpd_epu32(__m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_broadcastq_epi64(__m128i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpeq_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_mul_epu32(__m512i __W, __mmask8 __M, __m512i __X, __m512i __Y)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_ceil_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_add_epi32(__m512i __W)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_rsqrt14_ss(__m128 __A, __m128 __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_load_si512(void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_set1_epi32(__m512i __O, __mmask16 __M, int __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_permutevar_pd(__m512d __A, __m512i __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_unpackhi_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_max_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_min_epu64(__m512i __W, __mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_rsqrt14_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_and_epi64(__mmask8 __M, __m512i __W)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_compressstoreu_pd(void *__P, __mmask8 __U, __m512d __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_ceil_pd(__m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepi8_epi32(__m128i __A)
static __inline void __DEFAULT_FN_ATTRS _mm512_storeu_si512(void *__P, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_rsqrt14_pd(__m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_slli_epi64(__mmask8 __U, __m512i __A, int __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_sqrt_ps(__mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtps_epi32(__m512 __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmple_epu32_mask(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fnmadd_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_and_epi32(__mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_move_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_broadcastd_epi32(__m512i __O, __mmask16 __M, __m128i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_maskz_load_pd(__mmask8 __U, void const *__P)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kxor(__mmask16 __A, __mmask16 __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm_mask_store_sd(double *__W, __mmask8 __U, __m128d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_compress_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_stream_load_si512(void *__P)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_getexp_ps(__m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_expandloadu_epi32(__mmask16 __U, void const *__P)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_compress_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpge_epu64_mask(__m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_broadcastss_ps(__m512 __O, __mmask16 __M, __m128 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sra_epi32(__m512i __A, __m128i __B)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_castsi512_ps(__m512i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_mask_blend_pd(__mmask8 __U, __m512d __A, __m512d __W)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_rcp14_ss(__m128 __A, __m128 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_fnmsub_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_permutexvar_epi64(__m512i __W, __mmask8 __M, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_rorv_epi32(__m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_permutexvar_epi64(__m512i __X, __m512i __Y)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmplt_epi64_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_compress_epi64(__mmask8 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_andnot_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_broadcastd_epi32(__m128i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_fmsub_pd(__m512d __A, __m512d __B, __m512d __C)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_mul_epi32(__mmask8 __M, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_mov_epi64(__m512i __W, __mmask8 __U, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask3_fnmadd_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_epi32(__mmask16 __M, __m512i __V)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_sub_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_testn_epi32_mask(__m512i __A, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set4_epi64(long long __A, long long __B, long long __C, long long __D)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtusepi32_epi8(__m512i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpgt_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpneq_epu64_mask(__m512i __a, __m512i __b)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_mul_pd(__m512d __a, __m512d __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_unpackhi_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_cvtepi32_ps(__m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_mul_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sub_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_castps256_ps512(__m256 __a)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_mov_ps(__mmask16 __U, __m512 __A)
#define _mm512_reduce_maxMin_32bit(Vec512, IntrinName, T1, T2)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_xor_epi64(__m512i __a, __m512i __b)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_add_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_unpacklo_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi32_epi16(__m256i __O, __mmask16 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_scalef_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srai_epi32(__m512i __A, int __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_cvtepu32_ps(__m512 __W, __mmask16 __U, __m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_min_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srl_epi64(__m512i __A, __m128i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepu16_epi64(__m128i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_rsqrt14_ps(__mmask16 __U, __m512 __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_storeu_epi16(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_scalef_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_load_pd(void const *__p)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_fnmsub_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
#define _mm512_reduce_operator_64bit(Vec512, Operator, T2, T1)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_add_epi64(__m512i __W)
static __inline __m128d __DEFAULT_FN_ATTRS _mm512_castpd512_pd128(__m512d __a)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_castpd256_pd512(__m256d __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_add_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask2_permutex2var_epi32(__m512i __A, __m512i __I, __mmask16 __U, __m512i __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi32_storeu_epi16(void *__P, __mmask16 __M, __m512i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpeq_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_or_epi64(__m512i __src, __mmask8 __k, __m512i __a, __m512i __b)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpneq_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask2_permutex2var_ps(__m512 __A, __m512i __I, __mmask16 __U, __m512 __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask2int(__mmask16 __a)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_sqrt_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_load_epi32(void const *__P)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_castpd_si512(__m512d __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_storeu_epi32(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_expand_epi32(__mmask16 __U, __m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_broadcastss_ps(__m128 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepu16_epi32(__m256i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi16_epi64(__m512i __W, __mmask8 __U, __m128i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_load_sd(__m128d __W, __mmask8 __U, const double *__A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_div_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_sub_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_max_epu32(__m512i __W, __mmask16 __M, __m512i __A, __m512i __B)
static __inline __m256i __DEFAULT_FN_ATTRS _mm512_cvttpd_epi32(__m512d __a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sll_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m128i __B)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_cvtepi32_pd(__m256i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_mask_loadu_pd(__m512d __W, __mmask8 __U, void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepu8_epi64(__m512i __W, __mmask8 __U, __m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srli_epi64(__m512i __A, int __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_epi32(__mmask16 __M, __m512i __V)
static __inline__ unsigned long long __DEFAULT_FN_ATTRS _mm512_reduce_max_epu64(__m512i __V)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_add_pd(__m512d __a, __m512d __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srli_epi32(__m512i __A, int __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_rsqrt14_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepu16_epi32(__m512i __W, __mmask16 __U, __m256i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_div_pd(__m512d __a, __m512d __b)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtsepi32_epi16(__mmask16 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_min_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepi8_epi64(__m128i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_fnmadd_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_fmsub_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtpd_epi32(__m256i __W, __mmask8 __U, __m512d __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_epi16(__m256i __O, __mmask16 __M, __m512i __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtusepi64_epi32(__mmask8 __M, __m512i __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_mul_epi64(__m512i __W)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srav_epi64(__mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_cvtepu8_epi32(__m128i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_movedup_pd(__m512d __A)
static __inline__ unsigned int __DEFAULT_FN_ATTRS _mm512_reduce_min_epu32(__m512i a)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fmadd_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setzero_si128(void)
Creates a 128-bit integer vector initialized to zero.
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpeq_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_epi16(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_mov_epi32(__mmask16 __U, __m512i __A)
static __inline __m256d __DEFAULT_FN_ATTRS _mm512_castpd512_pd256(__m512d __A)
#define _MM_FROUND_CUR_DIRECTION
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_sub_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtpd_epi32(__m512d __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_rcp14_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi32_epi8(__m128i __O, __mmask16 __M, __m512i __A)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmple_epi64_mask(__m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fnmadd_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline __mmask16 __DEFAULT_FN_ATTRS _mm512_test_epi32_mask(__m512i __A, __m512i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_getexp_pd(__m512d __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_epi64(__mmask8 __M, __m512i __V)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvttpd_epu32(__m256i __W, __mmask8 __U, __m512d __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_epi32(__m256i __O, __mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srai_epi32(__mmask16 __U, __m512i __A, int __B)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_epi16(__m128i __O, __mmask8 __M, __m512i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_cvtu32_ss(__m128 __A, unsigned __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fmadd_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmple_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvttps_epi32(__mmask16 __U, __m512 __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_set4_pd(double __A, double __B, double __C, double __D)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_max_epi32(__mmask16 __M, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srlv_epi32(__mmask16 __U, __m512i __X, __m512i __Y)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpneq_epu64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_srlv_epi64(__mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_kmov(__mmask16 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_unpacklo_epi64(__mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sllv_epi64(__m512i __X, __m512i __Y)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_permutex2var_pd(__m512d __A, __m512i __I, __m512d __B)
#define _mm512_mask_reduce_maxMin_64bit(Vec512, Vec512Neutral, IntrinName, T1, T2, T3, Mask)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_scalef_sd(__m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_broadcast_i64x4(__m512i __O, __mmask8 __M, __m256i __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_fnmadd_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_mask_blend_ps(__mmask16 __U, __m512 __A, __m512 __W)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_cmpge_epi64_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_broadcast_i32x4(__m512i __O, __mmask16 __M, __m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_permutexvar_epi64(__mmask8 __M, __m512i __X, __m512i __Y)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_zextsi128_si512(__m128i __a)
Constructs a 512-bit integer vector from a 128-bit integer vector.
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_load_epi64(void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_srav_epi64(__m512i __X, __m512i __Y)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_expand_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_maskz_mul_epu32(__mmask8 __M, __m512i __X, __m512i __Y)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_div_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_reduce_mul_pd(__m512d __W)
static __inline void __DEFAULT_FN_ATTRS _mm512_store_pd(void *__P, __m512d __A)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_min_epi32(__m512i a)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_sub_epi64(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fmsubadd_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_fmsubadd_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi16_epi32(__m512i __W, __mmask16 __U, __m256i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu16_epi32(__mmask16 __U, __m256i __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtsepi64_epi32(__mmask8 __M, __m512i __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set1_epi64(long long __d)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_floor_ps(__m512 __A)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_reduce_min_ps(__m512 a)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_cvtepu32lo_pd(__m512i __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi32_storeu_epi8(void *__P, __mmask16 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_broadcast_f64x4(__m512d __O, __mmask8 __M, __m256d __A)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi32_epi16(__m256i __O, __mmask16 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fmsub_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline__ __mmask8 __DEFAULT_FN_ATTRS _mm512_mask_cmpneq_epi64_mask(__mmask8 __u, __m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_loadu_si512(void const *__P)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_testn_epi32_mask(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_min_epi64(__mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_max_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_set4_epi32(int __A, int __B, int __C, int __D)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_undefined_si128(void)
Generates a 128-bit vector of [4 x i32] with unspecified content.
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_mul_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fmaddsub_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtsepi32_epi8(__m512i __A)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_reduce_max_epi32(__m512i a)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtusepi64_epi16(__mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi8_epi32(__m512i __W, __mmask16 __U, __m128i __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtsepi64_epi8(__mmask8 __M, __m512i __A)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_reduce_add_ps(__m512 __W)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_castps_pd(__m512 __A)
static __inline__ unsigned long long __DEFAULT_FN_ATTRS _mm512_reduce_min_epu64(__m512i __V)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srai_epi64(__m512i __W, __mmask8 __U, __m512i __A, int __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi8_epi64(__m512i __W, __mmask8 __U, __m128i __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mask_mullo_epi32(__m512i __W, __mmask16 __M, __m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_getexp_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ unsigned int __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_epu32(__mmask16 __M, __m512i __V)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi32_epi8(__m128i __O, __mmask16 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_and_si512(__m512i __a, __m512i __b)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_store_epi64(void *__P, __mmask8 __U, __m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_rcp14_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_setzero_si512(void)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_rsqrt14_sd(__m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu32_ps(__mmask16 __U, __m512i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_getexp_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srl_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m128i __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtepi32_epi16(__m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_and_epi64(__m512i __src, __mmask8 __k, __m512i __a, __m512i __b)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_sub_pd(__m512d __a, __m512d __b)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_and_epi32(__mmask16 __M, __m512i __W)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_fmsub_ss(__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
static __inline__ unsigned int __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_epu32(__mmask16 __M, __m512i __V)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_expand_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_reduce_or_epi64(__m512i __W)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_unpacklo_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_add_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask3_fnmadd_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi16_epi32(__mmask16 __U, __m256i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_undefined_pd(void)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_min_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_andnot_si512(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fmaddsub_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_permutex2var_epi32(__m512i __A, __mmask16 __U, __m512i __I, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_mul_epi32(__m512i __X, __m512i __Y)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_storeu_epi16(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fmsubadd_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_set_ps(float __A, float __B, float __C, float __D, float __E, float __F, float __G, float __H, float __I, float __J, float __K, float __L, float __M, float __N, float __O, float __P)
#define __DEFAULT_FN_ATTRS
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_abs_epi32(__m512i __W, __mmask16 __U, __m512i __A)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_mul_epi64(__mmask8 __M, __m512i __W)
static __inline__ long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_or_epi64(__mmask8 __M, __m512i __W)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_load_epi32(__m512i __W, __mmask16 __U, void const *__P)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtsepi64_epi16(__mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepu8_epi32(__mmask16 __U, __m128i __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_castps128_ps512(__m128 __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_getexp_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fmsubadd_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_xor_epi32(__m512i __a, __m512i __b)
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setzero_si256(void)
Constructs a 256-bit integer vector initialized to zero.
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_abs_ps(__m512 __W, __mmask16 __K, __m512 __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_cvtepi32_epi8(__m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_fmsub_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_abs_ps(__m512 __A)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_stream_pd(double *__P, __m512d __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fmsubadd_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_ceil_ps(__m512 __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_sub_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_move_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_fmsub_ps(__m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_expandloadu_pd(__m512d __W, __mmask8 __U, void const *__P)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fmsub_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_rolv_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_cvttps_epu32(__m512 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_mul_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_store_epi32(void *__P, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srli_epi32(__m512i __W, __mmask16 __U, __m512i __A, int __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_div_sd(__mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_sra_epi64(__mmask8 __U, __m512i __A, __m128i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_load_ss(__mmask8 __U, const float *__A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_or_epi32(__m512i __src, __mmask16 __k, __m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_permutexvar_ps(__m512 __W, __mmask16 __U, __m512i __X, __m512 __Y)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_getexp_ss(__m128 __A, __m128 __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_sub_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_expandloadu_ps(__mmask16 __U, void const *__P)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvttps_epu32(__m512i __W, __mmask16 __U, __m512 __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_scalef_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi32_storeu_epi16(void *__P, __mmask16 __M, __m512i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_zextpd128_pd512(__m128d __a)
Constructs a 512-bit floating-point vector of [8 x double] from a 128-bit floating-point vector of [2...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_cvtsd_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128d __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpneq_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_rcp14_ps(__m512 __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_test_epi32_mask(__mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi8_epi32(__mmask16 __U, __m128i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_rorv_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_storeu_epi32(void *__P, __mmask16 __U, __m512i __A)
static __inline__ unsigned __DEFAULT_FN_ATTRS _mm_cvtsd_u32(__m128d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_slli_epi32(__m512i __W, __mmask16 __U, __m512i __A, int __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtpd_epu32(__mmask8 __U, __m512d __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_permutex2var_ps(__m512 __A, __m512i __I, __m512 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_fmsub_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_sub_epi32(__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpge_epu32_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srli_epi64(__m512i __W, __mmask8 __U, __m512i __A, int __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmpneq_epi32_mask(__m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_rolv_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_max_epu64(__m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_permutex2var_ps(__m512 __A, __mmask16 __U, __m512i __I, __m512 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_broadcast_f32x4(__mmask16 __M, __m128 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_getexp_pd(__mmask8 __U, __m512d __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_broadcast_i64x4(__mmask8 __M, __m256i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_mul_ps(__m512 __a, __m512 __b)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_add_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_cvtps_epi32(__m512i __W, __mmask16 __U, __m512 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_permutevar_ps(__m512 __A, __m512i __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_add_epi32(__m512i __A, __m512i __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtu32_sd(__m128d __A, unsigned __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask_rcp14_ss(__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_unpacklo_epi32(__m512i __A, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmplt_epu32_mask(__m512i __a, __m512i __b)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi64_epi16(__mmask8 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_andnot_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_max_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_set_pd(double __A, double __B, double __C, double __D, double __E, double __F, double __G, double __H)
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_setzero_pd(void)
Constructs a 256-bit floating-point vector of [4 x double] with all vector elements initialized to ze...
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_permutex2var_epi32(__m512i __A, __m512i __I, __m512i __B)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_rorv_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtepi64_epi32(__m256i __O, __mmask8 __M, __m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_getexp_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_mask_load_ps(__m512 __W, __mmask16 __U, void const *__P)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_fmaddsub_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpeq_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_max_ps(__m512 __A, __m512 __B)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mask_rsqrt14_sd(__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_mov_pd(__m512d __W, __mmask8 __U, __m512d __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_sqrt_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_undefined(void)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_broadcast_f32x4(__m128 __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi32_epi8(__m128i __O, __mmask16 __M, __m512i __A)
static __inline__ unsigned __DEFAULT_FN_ATTRS _mm_cvtss_u32(__m128 __A)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_mask3_fmadd_ss(__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_fnmadd_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_cvtepi32_epi64(__mmask8 __U, __m256i __X)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_mask_reduce_or_epi32(__mmask16 __M, __m512i __W)
#define _mm512_mask_reduce_operator_32bit(Vec512, Vec512Neutral, Operator,Mask, T2, T1, T3)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_getexp_ps(__mmask16 __U, __m512 __A)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask3_fnmadd_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_cvtepu32_pd(__m512d __W, __mmask8 __U, __m256i __A)
static __inline __m512d __DEFAULT_FN_ATTRS _mm512_unpacklo_pd(__m512d __a, __m512d __b)
static __inline__ int __DEFAULT_FN_ATTRS _mm_cvttsd_i32(__m128d __A)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_abs_epi32(__m512i __A)
static __inline void __DEFAULT_FN_ATTRS _mm512_mask_storeu_pd(void *__P, __mmask8 __U, __m512d __A)
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm512_maskz_cvtusepi64_epi8(__mmask8 __M, __m512i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_maskz_loadu_ps(__mmask16 __U, void const *__P)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_permutevar_ps(__m512 __W, __mmask16 __U, __m512 __A, __m512i __C)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_maskz_add_pd(__mmask8 __U, __m512d __A, __m512d __B)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_mask_unpacklo_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm512_mask_cvtusepi64_storeu_epi16(void *__P, __mmask8 __M, __m512i __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_undefined_pd(void)
Constructs a 128-bit floating-point vector of [2 x double] with unspecified content.
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_srlv_epi64(__m512i __W, __mmask8 __U, __m512i __X, __m512i __Y)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_load_epi64(__m512i __W, __mmask8 __U, void const *__P)
static __inline__ unsigned long long __DEFAULT_FN_ATTRS _mm512_mask_reduce_max_epu64(__mmask8 __M, __m512i __V)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_min_epi32(__mmask16 __M, __m512i __A, __m512i __B)
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_maskz_min_ss(__mmask8 __U, __m128 __A, __m128 __B)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_mask_moveldup_ps(__m512 __W, __mmask16 __U, __m512 __A)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_mask_reduce_min_pd(__mmask8 __M, __m512d __V)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_sub_ps(__mmask16 __U, __m512 __A, __m512 __B)
static __inline__ void __DEFAULT_FN_ATTRS _mm_mask_store_ss(float *__W, __mmask8 __U, __m128 __A)
static __inline__ float __DEFAULT_FN_ATTRS _mm512_mask_reduce_add_ps(__mmask16 __M, __m512 __W)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmpge_epi32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_xor_epi64(__m512i __src, __mmask8 __k, __m512i __a, __m512i __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_compress_epi64(__m512i __W, __mmask8 __U, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_mask_expandloadu_epi32(__m512i __W, __mmask16 __U, void const *__P)
static __inline__ __m512 __DEFAULT_FN_ATTRS _mm512_maskz_permutevar_ps(__mmask16 __U, __m512 __A, __m512i __C)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_mask_cvtsepi64_epi32(__m256i __O, __mmask8 __M, __m512i __A)
static __inline __m512 __DEFAULT_FN_ATTRS _mm512_div_ps(__m512 __a, __m512 __b)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_max_epi64(__mmask8 __M, __m512i __A, __m512i __B)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_mask_cmple_epu32_mask(__mmask16 __u, __m512i __a, __m512i __b)
static __inline__ __m512d __DEFAULT_FN_ATTRS _mm512_permutexvar_pd(__m512i __X, __m512d __Y)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_cvtpd_epu32(__m512d __A)
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_maskz_fnmadd_sd(__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm512_maskz_cvtusepi32_epi16(__mmask16 __M, __m512i __A)
static __inline__ __m512i __DEFAULT_FN_ATTRS _mm512_maskz_compress_epi32(__mmask16 __U, __m512i __A)
static __inline__ double __DEFAULT_FN_ATTRS _mm512_mask_reduce_add_pd(__mmask8 __M, __m512d __W)
static __inline void __DEFAULT_FN_ATTRS _mm512_store_ps(void *__P, __m512 __A)
static __inline__ __mmask16 __DEFAULT_FN_ATTRS _mm512_cmplt_epi32_mask(__m512i __a, __m512i __b)
static __inline __m512i __DEFAULT_FN_ATTRS _mm512_max_epi32(__m512i __A, __m512i __B)
static __inline__ int __DEFAULT_FN_ATTRS _mm512_kortestc(__mmask16 __A, __mmask16 __B)