11#error "Never use <avxintrin.h> directly; include <immintrin.h> instead."
19typedef long long __v4di
__attribute__ ((__vector_size__ (32)));
25typedef unsigned long long __v4du
__attribute__ ((__vector_size__ (32)));
26typedef unsigned int __v8su
__attribute__ ((__vector_size__ (32)));
27typedef unsigned short __v16hu
__attribute__ ((__vector_size__ (32)));
28typedef unsigned char __v32qu
__attribute__ ((__vector_size__ (32)));
32typedef signed char __v32qs
__attribute__((__vector_size__(32)));
34typedef float __m256
__attribute__ ((__vector_size__ (32), __aligned__(32)));
35typedef double __m256d
__attribute__((__vector_size__(32), __aligned__(32)));
36typedef long long __m256i
__attribute__((__vector_size__(32), __aligned__(32)));
38typedef float __m256_u
__attribute__ ((__vector_size__ (32), __aligned__(1)));
39typedef double __m256d_u
__attribute__((__vector_size__(32), __aligned__(1)));
40typedef long long __m256i_u
__attribute__((__vector_size__(32), __aligned__(1)));
48typedef __bf16 __v16bf
__attribute__((__vector_size__(32), __aligned__(32)));
49typedef __bf16 __m256bh
__attribute__((__vector_size__(32), __aligned__(32)));
53#if defined(__EVEX512__) && !defined(__AVX10_1_512__)
54#define __DEFAULT_FN_ATTRS \
55 __attribute__((__always_inline__, __nodebug__, __target__("avx,no-evex512"), \
56 __min_vector_width__(256)))
57#define __DEFAULT_FN_ATTRS128 \
58 __attribute__((__always_inline__, __nodebug__, __target__("avx,no-evex512"), \
59 __min_vector_width__(128)))
61#define __DEFAULT_FN_ATTRS \
62 __attribute__((__always_inline__, __nodebug__, __target__("avx"), \
63 __min_vector_width__(256)))
64#define __DEFAULT_FN_ATTRS128 \
65 __attribute__((__always_inline__, __nodebug__, __target__("avx"), \
66 __min_vector_width__(128)))
69#if defined(__cplusplus) && (__cplusplus >= 201103L)
70#define __DEFAULT_FN_ATTRS_CONSTEXPR __DEFAULT_FN_ATTRS constexpr
71#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS128 constexpr
73#define __DEFAULT_FN_ATTRS_CONSTEXPR __DEFAULT_FN_ATTRS128
74#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS
93 return (__m256d)((__v4df)
__a+(__v4df)
__b);
111 return (__m256)((__v8sf)
__a+(__v8sf)
__b);
129 return (__m256d)((__v4df)
__a-(__v4df)
__b);
147 return (__m256)((__v8sf)
__a-(__v8sf)
__b);
166 return (__m256d)__builtin_ia32_addsubpd256((__v4df)
__a, (__v4df)
__b);
185 return (__m256)__builtin_ia32_addsubps256((__v8sf)
__a, (__v8sf)
__b);
203 return (__m256d)((__v4df)
__a/(__v4df)
__b);
221 return (__m256)((__v8sf)
__a/(__v8sf)
__b);
242 return (__m256d)__builtin_ia32_maxpd256((__v4df)
__a, (__v4df)
__b);
263 return (__m256)__builtin_ia32_maxps256((__v8sf)
__a, (__v8sf)
__b);
284 return (__m256d)__builtin_ia32_minpd256((__v4df)
__a, (__v4df)
__b);
305 return (__m256)__builtin_ia32_minps256((__v8sf)
__a, (__v8sf)
__b);
323 return (__m256d)((__v4df)
__a * (__v4df)
__b);
341 return (__m256)((__v8sf)
__a * (__v8sf)
__b);
358 return (__m256d)__builtin_ia32_sqrtpd256((__v4df)
__a);
375 return (__m256)__builtin_ia32_sqrtps256((__v8sf)
__a);
392 return (__m256)__builtin_ia32_rsqrtps256((__v8sf)
__a);
409 return (__m256)__builtin_ia32_rcpps256((__v8sf)
__a);
441#define _mm256_round_pd(V, M) \
442 ((__m256d)__builtin_ia32_roundpd256((__v4df)(__m256d)(V), (M)))
473#define _mm256_round_ps(V, M) \
474 ((__m256)__builtin_ia32_roundps256((__v8sf)(__m256)(V), (M)))
491#define _mm256_ceil_pd(V) _mm256_round_pd((V), _MM_FROUND_CEIL)
509#define _mm256_floor_pd(V) _mm256_round_pd((V), _MM_FROUND_FLOOR)
526#define _mm256_ceil_ps(V) _mm256_round_ps((V), _MM_FROUND_CEIL)
543#define _mm256_floor_ps(V) _mm256_round_ps((V), _MM_FROUND_FLOOR)
561 return (__m256d)((__v4du)
__a & (__v4du)
__b);
579 return (__m256)((__v8su)
__a & (__v8su)
__b);
600 return (__m256d)(~(__v4du)
__a & (__v4du)
__b);
621 return (__m256)(~(__v8su)
__a & (__v8su)
__b);
639 return (__m256d)((__v4du)
__a | (__v4du)
__b);
657 return (__m256)((__v8su)
__a | (__v8su)
__b);
675 return (__m256d)((__v4du)
__a ^ (__v4du)
__b);
693 return (__m256)((__v8su)
__a ^ (__v8su)
__b);
717 return (__m256d)__builtin_ia32_haddpd256((__v4df)
__a, (__v4df)
__b);
740 return (__m256)__builtin_ia32_haddps256((__v8sf)
__a, (__v8sf)
__b);
763 return (__m256d)__builtin_ia32_hsubpd256((__v4df)
__a, (__v4df)
__b);
786 return (__m256)__builtin_ia32_hsubps256((__v8sf)
__a, (__v8sf)
__b);
816 return (__m128d)__builtin_ia32_vpermilvarpd((__v2df)
__a, (__v2di)
__c);
855 return (__m256d)__builtin_ia32_vpermilvarpd256((__v4df)
__a, (__v4di)
__c);
910 return (__m128)__builtin_ia32_vpermilvarps((__v4sf)
__a, (__v4si)
__c);
1001 return (__m256)__builtin_ia32_vpermilvarps256((__v8sf)
__a, (__v8si)
__c);
1031#define _mm_permute_pd(A, C) \
1032 ((__m128d)__builtin_ia32_vpermilpd((__v2df)(__m128d)(A), (int)(C)))
1071#define _mm256_permute_pd(A, C) \
1072 ((__m256d)__builtin_ia32_vpermilpd256((__v4df)(__m256d)(A), (int)(C)))
1127#define _mm_permute_ps(A, C) \
1128 ((__m128)__builtin_ia32_vpermilps((__v4sf)(__m128)(A), (int)(C)))
1219#define _mm256_permute_ps(A, C) \
1220 ((__m256)__builtin_ia32_vpermilps256((__v8sf)(__m256)(A), (int)(C)))
1259#define _mm256_permute2f128_pd(V1, V2, M) \
1260 ((__m256d)__builtin_ia32_vperm2f128_pd256((__v4df)(__m256d)(V1), \
1261 (__v4df)(__m256d)(V2), (int)(M)))
1300#define _mm256_permute2f128_ps(V1, V2, M) \
1301 ((__m256)__builtin_ia32_vperm2f128_ps256((__v8sf)(__m256)(V1), \
1302 (__v8sf)(__m256)(V2), (int)(M)))
1340#define _mm256_permute2f128_si256(V1, V2, M) \
1341 ((__m256i)__builtin_ia32_vperm2f128_si256((__v8si)(__m256i)(V1), \
1342 (__v8si)(__m256i)(V2), (int)(M)))
1369#define _mm256_blend_pd(V1, V2, M) \
1370 ((__m256d)__builtin_ia32_blendpd256((__v4df)(__m256d)(V1), \
1371 (__v4df)(__m256d)(V2), (int)(M)))
1397#define _mm256_blend_ps(V1, V2, M) \
1398 ((__m256)__builtin_ia32_blendps256((__v8sf)(__m256)(V1), \
1399 (__v8sf)(__m256)(V2), (int)(M)))
1425 return (__m256d)__builtin_ia32_blendvpd256(
1453 return (__m256)__builtin_ia32_blendvps256(
1495#define _mm256_dp_ps(V1, V2, M) \
1496 ((__m256)__builtin_ia32_dpps256((__v8sf)(__m256)(V1), \
1497 (__v8sf)(__m256)(V2), (M)))
1552#define _mm256_shuffle_ps(a, b, mask) \
1553 ((__m256)__builtin_ia32_shufps256((__v8sf)(__m256)(a), \
1554 (__v8sf)(__m256)(b), (int)(mask)))
1598#define _mm256_shuffle_pd(a, b, mask) \
1599 ((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
1600 (__v4df)(__m256d)(b), (int)(mask)))
1603#define _CMP_EQ_UQ 0x08
1604#define _CMP_NGE_US 0x09
1605#define _CMP_NGT_US 0x0a
1606#define _CMP_FALSE_OQ 0x0b
1607#define _CMP_NEQ_OQ 0x0c
1608#define _CMP_GE_OS 0x0d
1609#define _CMP_GT_OS 0x0e
1610#define _CMP_TRUE_UQ 0x0f
1611#define _CMP_EQ_OS 0x10
1612#define _CMP_LT_OQ 0x11
1613#define _CMP_LE_OQ 0x12
1614#define _CMP_UNORD_S 0x13
1615#define _CMP_NEQ_US 0x14
1616#define _CMP_NLT_UQ 0x15
1617#define _CMP_NLE_UQ 0x16
1618#define _CMP_ORD_S 0x17
1619#define _CMP_EQ_US 0x18
1620#define _CMP_NGE_UQ 0x19
1621#define _CMP_NGT_UQ 0x1a
1622#define _CMP_FALSE_OS 0x1b
1623#define _CMP_NEQ_OS 0x1c
1624#define _CMP_GE_OQ 0x1d
1625#define _CMP_GT_OQ 0x1e
1626#define _CMP_TRUE_US 0x1f
1802#define _mm256_cmp_pd(a, b, c) \
1803 ((__m256d)__builtin_ia32_cmppd256((__v4df)(__m256d)(a), \
1804 (__v4df)(__m256d)(b), (c)))
1862#define _mm256_cmp_ps(a, b, c) \
1863 ((__m256)__builtin_ia32_cmpps256((__v8sf)(__m256)(a), \
1864 (__v8sf)(__m256)(b), (c)))
2003#define _mm256_extract_epi32(X, N) \
2004 ((int)__builtin_ia32_vec_ext_v8si((__v8si)(__m256i)(X), (int)(N)))
2025#define _mm256_extract_epi16(X, N) \
2026 ((int)(unsigned short)__builtin_ia32_vec_ext_v16hi((__v16hi)(__m256i)(X), \
2048#define _mm256_extract_epi8(X, N) \
2049 ((int)(unsigned char)__builtin_ia32_vec_ext_v32qi((__v32qi)(__m256i)(X), \
2072#define _mm256_extract_epi64(X, N) \
2073 ((long long)__builtin_ia32_vec_ext_v4di((__v4di)(__m256i)(X), (int)(N)))
2098#define _mm256_insert_epi32(X, I, N) \
2099 ((__m256i)__builtin_ia32_vec_set_v8si((__v8si)(__m256i)(X), \
2100 (int)(I), (int)(N)))
2125#define _mm256_insert_epi16(X, I, N) \
2126 ((__m256i)__builtin_ia32_vec_set_v16hi((__v16hi)(__m256i)(X), \
2127 (int)(I), (int)(N)))
2151#define _mm256_insert_epi8(X, I, N) \
2152 ((__m256i)__builtin_ia32_vec_set_v32qi((__v32qi)(__m256i)(X), \
2153 (int)(I), (int)(N)))
2178#define _mm256_insert_epi64(X, I, N) \
2179 ((__m256i)__builtin_ia32_vec_set_v4di((__v4di)(__m256i)(X), \
2180 (long long)(I), (int)(N)))
2196 return (__m256d)__builtin_convertvector((__v4si)
__a, __v4df);
2211 return (__m256)__builtin_convertvector((__v8si)
__a, __v8sf);
2227 return (__m128)__builtin_ia32_cvtpd2ps256((__v4df)
__a);
2246 return (__m256i)__builtin_ia32_cvtps2dq256((__v8sf)
__a);
2262 return (__m256d)__builtin_convertvector((__v4sf)
__a, __v4df);
2283 return (__m128i)__builtin_ia32_cvttpd2dq256((__v4df)
__a);
2303 return (__m128i)__builtin_ia32_cvtpd2dq256((__v4df)
__a);
2323 return (__m256i)__builtin_ia32_cvttps2dq256((__v8sf)
__a);
2355 __v8si
__b = (__v8si)
__a;
2398 return __builtin_shufflevector((__v8sf)
__a, (__v8sf)
__a, 1, 1, 3, 3, 5, 5, 7, 7);
2423 return __builtin_shufflevector((__v8sf)
__a, (__v8sf)
__a, 0, 0, 2, 2, 4, 4, 6, 6);
2445 return __builtin_shufflevector((__v4df)
__a, (__v4df)
__a, 0, 0, 2, 2);
2468 return __builtin_shufflevector((__v4df)
__a, (__v4df)
__b, 1, 5, 1+2, 5+2);
2490 return __builtin_shufflevector((__v4df)
__a, (__v4df)
__b, 0, 4, 0+2, 4+2);
2517 return __builtin_shufflevector((__v8sf)
__a, (__v8sf)
__b, 2, 10, 2+1, 10+1, 6, 14, 6+1, 14+1);
2544 return __builtin_shufflevector((__v8sf)
__a, (__v8sf)
__b, 0, 8, 0+1, 8+1, 4, 12, 4+1, 12+1);
2574 return __builtin_ia32_vtestzpd((__v2df)
__a, (__v2df)
__b);
2603 return __builtin_ia32_vtestcpd((__v2df)
__a, (__v2df)
__b);
2633 return __builtin_ia32_vtestnzcpd((__v2df)
__a, (__v2df)
__b);
2662 return __builtin_ia32_vtestzps((__v4sf)
__a, (__v4sf)
__b);
2691 return __builtin_ia32_vtestcps((__v4sf)
__a, (__v4sf)
__b);
2721 return __builtin_ia32_vtestnzcps((__v4sf)
__a, (__v4sf)
__b);
2750 return __builtin_ia32_vtestzpd256((__v4df)
__a, (__v4df)
__b);
2779 return __builtin_ia32_vtestcpd256((__v4df)
__a, (__v4df)
__b);
2809 return __builtin_ia32_vtestnzcpd256((__v4df)
__a, (__v4df)
__b);
2838 return __builtin_ia32_vtestzps256((__v8sf)
__a, (__v8sf)
__b);
2867 return __builtin_ia32_vtestcps256((__v8sf)
__a, (__v8sf)
__b);
2897 return __builtin_ia32_vtestnzcps256((__v8sf)
__a, (__v8sf)
__b);
2923 return __builtin_ia32_ptestz256((__v4di)
__a, (__v4di)
__b);
2949 return __builtin_ia32_ptestc256((__v4di)
__a, (__v4di)
__b);
2976 return __builtin_ia32_ptestnzc256((__v4di)
__a, (__v4di)
__b);
2995 return __builtin_ia32_movmskpd256((__v4df)
__a);
3013 return __builtin_ia32_movmskps256((__v8sf)
__a);
3022static __inline
void __attribute__((__always_inline__, __nodebug__, __target__(
"avx")))
3025 __builtin_ia32_vzeroall();
3033static __inline
void __attribute__((__always_inline__, __nodebug__, __target__(
"avx")))
3034_mm256_zeroupper(
void)
3036 __builtin_ia32_vzeroupper();
3055 struct __mm_broadcast_ss_struct {
3058 float __f = ((
const struct __mm_broadcast_ss_struct*)
__a)->__f;
3059 return __extension__ (__m128){ __f, __f, __f, __f };
3077 struct __mm256_broadcast_sd_struct {
3080 double __d = ((
const struct __mm256_broadcast_sd_struct*)
__a)->__d;
3081 return __extension__ (__m256d)(__v4df){ __d, __d, __d, __d };
3099 struct __mm256_broadcast_ss_struct {
3102 float __f = ((
const struct __mm256_broadcast_ss_struct*)
__a)->__f;
3103 return __extension__ (__m256)(__v8sf){ __f, __f, __f, __f, __f, __f, __f, __f };
3122 return (__m256d)__builtin_shufflevector((__v2df)
__b, (__v2df)
__b,
3142 return (__m256)__builtin_shufflevector((__v4sf)
__b, (__v4sf)
__b,
3143 0, 1, 2, 3, 0, 1, 2, 3);
3161 return *(
const __m256d *)
__p;
3177 return *(
const __m256 *)
__p;
3197 return ((
const struct __loadu_pd*)
__p)->__v;
3217 return ((
const struct __loadu_ps*)
__p)->__v;
3250 struct __loadu_si256 {
3253 return ((
const struct __loadu_si256*)
__p)->__v;
3271 return (__m256i)__builtin_ia32_lddqu256((
char const *)
__p);
3327 struct __storeu_pd {
3330 ((
struct __storeu_pd*)
__p)->__v =
__a;
3347 struct __storeu_ps {
3350 ((
struct __storeu_ps*)
__p)->__v =
__a;
3385 struct __storeu_si256 {
3388 ((
struct __storeu_si256*)
__p)->__v =
__a;
3413 return (__m128d)__builtin_ia32_maskloadpd((
const __v2df *)
__p, (__v2di)__m);
3437 return (__m256d)__builtin_ia32_maskloadpd256((
const __v4df *)
__p,
3462 return (__m128)__builtin_ia32_maskloadps((
const __v4sf *)
__p, (__v4si)__m);
3486 return (__m256)__builtin_ia32_maskloadps256((
const __v8sf *)
__p, (__v8si)__m);
3511 __builtin_ia32_maskstoreps256((__v8sf *)
__p, (__v8si)__m, (__v8sf)
__a);
3535 __builtin_ia32_maskstorepd((__v2df *)
__p, (__v2di)__m, (__v2df)
__a);
3559 __builtin_ia32_maskstorepd256((__v4df *)
__p, (__v4di)__m, (__v4df)
__a);
3583 __builtin_ia32_maskstoreps((__v4sf *)
__p, (__v4si)__m, (__v4sf)
__a);
3604 __builtin_nontemporal_store((__v4di_aligned)
__b, (__v4di_aligned*)
__a);
3624 __builtin_nontemporal_store((__v4df_aligned)
__b, (__v4df_aligned*)
__a);
3645 __builtin_nontemporal_store((__v8sf_aligned)
__a, (__v8sf_aligned*)
__p);
3659 return (__m256d)__builtin_ia32_undef256();
3672 return (__m256)__builtin_ia32_undef256();
3685 return (__m256i)__builtin_ia32_undef256();
3712 return __extension__ (__m256d){ __d,
__c,
__b,
__a };
3750 float __e,
float __f,
float __g,
float __h)
3752 return __extension__ (__m256){ __h, __g, __f, __e, __d,
__c,
__b,
__a };
3782 int __i4,
int __i5,
int __i6,
int __i7)
3784 return __extension__ (__m256i)(__v8si){ __i7, __i6, __i5, __i4, __i3, __i2, __i1, __i0 };
3830 short __w11,
short __w10,
short __w09,
short __w08,
3831 short __w07,
short __w06,
short __w05,
short __w04,
3832 short __w03,
short __w02,
short __w01,
short __w00)
3834 return __extension__ (__m256i)(__v16hi){ __w00, __w01, __w02, __w03, __w04, __w05, __w06,
3835 __w07, __w08, __w09, __w10, __w11, __w12, __w13, __w14, __w15 };
3913 char __b27,
char __b26,
char __b25,
char __b24,
3914 char __b23,
char __b22,
char __b21,
char __b20,
3915 char __b19,
char __b18,
char __b17,
char __b16,
3916 char __b15,
char __b14,
char __b13,
char __b12,
3917 char __b11,
char __b10,
char __b09,
char __b08,
3918 char __b07,
char __b06,
char __b05,
char __b04,
3919 char __b03,
char __b02,
char __b01,
char __b00)
3921 return __extension__ (__m256i)(__v32qi){
3922 __b00, __b01, __b02, __b03, __b04, __b05, __b06, __b07,
3923 __b08, __b09, __b10, __b11, __b12, __b13, __b14, __b15,
3924 __b16, __b17, __b18, __b19, __b20, __b21, __b22, __b23,
3925 __b24, __b25, __b26, __b27, __b28, __b29, __b30, __b31
3949 return __extension__ (__m256i)(__v4di){ __d,
__c,
__b,
__a };
4017 float __e,
float __f,
float __g,
float __h)
4049 int __i4,
int __i5,
int __i6,
int __i7)
4097 short __w11,
short __w10,
short __w09,
short __w08,
4098 short __w07,
short __w06,
short __w05,
short __w04,
4099 short __w03,
short __w02,
short __w01,
short __w00)
4102 __w04, __w05, __w06, __w07,
4103 __w08, __w09, __w10, __w11,
4104 __w12, __w13, __w14, __w15);
4182 char __b27,
char __b26,
char __b25,
char __b24,
4183 char __b23,
char __b22,
char __b21,
char __b20,
4184 char __b19,
char __b18,
char __b17,
char __b16,
4185 char __b15,
char __b14,
char __b13,
char __b12,
4186 char __b11,
char __b10,
char __b09,
char __b08,
4187 char __b07,
char __b06,
char __b05,
char __b04,
4188 char __b03,
char __b02,
char __b01,
char __b00)
4190 return _mm256_set_epi8(__b00, __b01, __b02, __b03, __b04, __b05, __b06, __b07,
4191 __b08, __b09, __b10, __b11, __b12, __b13, __b14, __b15,
4192 __b16, __b17, __b18, __b19, __b20, __b21, __b22, __b23,
4193 __b24, __b25, __b26, __b27, __b28, __b29, __b30, __b31);
4254 return _mm256_set_ps(__w, __w, __w, __w, __w, __w, __w, __w);
4292 __w, __w, __w, __w, __w, __w, __w, __w);
4343 return __extension__(__m256d){0.0, 0.0, 0.0, 0.0};
4355 return __extension__ (__m256){ 0.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.0f };
4367 return __extension__ (__m256i)(__v4di){ 0, 0, 0, 0 };
4402 return (__m256i)
__a;
4419 return (__m256d)
__a;
4436 return (__m256i)
__a;
4470 return (__m256d)
__a;
4487 return __builtin_shufflevector((__v4df)
__a, (__v4df)
__a, 0, 1);
4504 return __builtin_shufflevector((__v8sf)
__a, (__v8sf)
__a, 0, 1, 2, 3);
4520 return __builtin_shufflevector((__v4di)
__a, (__v4di)
__a, 0, 1);
4541 return __builtin_shufflevector(
4542 (__v2df)
__a, (__v2df)__builtin_nondeterministic_value(
__a), 0, 1, 2, 3);
4563 return __builtin_shufflevector((__v4sf)
__a,
4564 (__v4sf)__builtin_nondeterministic_value(
__a),
4565 0, 1, 2, 3, 4, 5, 6, 7);
4584 return __builtin_shufflevector(
4585 (__v2di)
__a, (__v2di)__builtin_nondeterministic_value(
__a), 0, 1, 2, 3);
4604 return __builtin_shufflevector((__v2df)
__a, (__v2df)
_mm_setzero_pd(), 0, 1, 2, 3);
4622 return __builtin_shufflevector((__v4sf)
__a, (__v4sf)
_mm_setzero_ps(), 0, 1, 2, 3, 4, 5, 6, 7);
4682#define _mm256_insertf128_ps(V1, V2, M) \
4683 ((__m256)__builtin_ia32_vinsertf128_ps256((__v8sf)(__m256)(V1), \
4684 (__v4sf)(__m128)(V2), (int)(M)))
4720#define _mm256_insertf128_pd(V1, V2, M) \
4721 ((__m256d)__builtin_ia32_vinsertf128_pd256((__v4df)(__m256d)(V1), \
4722 (__v2df)(__m128d)(V2), (int)(M)))
4758#define _mm256_insertf128_si256(V1, V2, M) \
4759 ((__m256i)__builtin_ia32_vinsertf128_si256((__v8si)(__m256i)(V1), \
4760 (__v4si)(__m128i)(V2), (int)(M)))
4788#define _mm256_extractf128_ps(V, M) \
4789 ((__m128)__builtin_ia32_vextractf128_ps256((__v8sf)(__m256)(V), (int)(M)))
4812#define _mm256_extractf128_pd(V, M) \
4813 ((__m128d)__builtin_ia32_vextractf128_pd256((__v4df)(__m256d)(V), (int)(M)))
4836#define _mm256_extractf128_si256(V, M) \
4837 ((__m128i)__builtin_ia32_vextractf128_si256((__v8si)(__m256i)(V), (int)(M)))
4857 return (__m256) __builtin_shufflevector((__v4sf)__lo, (__v4sf)__hi, 0, 1, 2, 3, 4, 5, 6, 7);
4878 return (__m256d) __builtin_shufflevector((__v2df)__lo, (__v2df)__hi, 0, 1, 2, 3);
4898 return (__m256i) __builtin_shufflevector((__v2di)__lo, (__v2di)__hi, 0, 1, 2, 3);
5135#undef __DEFAULT_FN_ATTRS
5136#undef __DEFAULT_FN_ATTRS_CONSTEXPR
5137#undef __DEFAULT_FN_ATTRS128
5138#undef __DEFAULT_FN_ATTRS128_CONSTEXPR
static __inline__ vector float vector float vector float __c
static __inline__ vector float vector float __b
static __inline__ uint32_t volatile uint32_t * __p
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_broadcast_sd(double const *__a)
Loads a scalar double-precision floating point value from the specified address pointed to by __a and...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_set_m128(__m128 __hi, __m128 __lo)
Constructs a 256-bit floating-point vector of [8 x float] by concatenating two 128-bit floating-point...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_broadcast_pd(__m128d const *__a)
Loads the data from a 128-bit vector of [2 x double] from the specified address pointed to by __a and...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_hsub_pd(__m256d __a, __m256d __b)
Horizontally subtracts the adjacent pairs of values contained in two 256-bit vectors of [4 x double].
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu_pd(double *__p, __m256d __a)
Stores double-precision floating point values from a 256-bit vector of [4 x double] to an unaligned m...
static __inline int __DEFAULT_FN_ATTRS _mm256_testc_si256(__m256i __a, __m256i __b)
Given two 256-bit integer vectors, perform a bit-by-bit comparison of the two source vectors.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_add_ps(__m256 __a, __m256 __b)
Adds two 256-bit vectors of [8 x float].
static __inline void __DEFAULT_FN_ATTRS _mm256_stream_pd(void *__a, __m256d __b)
Moves double-precision values from a 256-bit vector of [4 x double] to a 32-byte aligned memory locat...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_castsi256_ps(__m256i __a)
Casts a 256-bit integer vector into a 256-bit floating-point vector of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_broadcast_ps(__m128 const *__a)
Loads the data from a 128-bit vector of [4 x float] from the specified address pointed to by __a and ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setr_epi8(char __b31, char __b30, char __b29, char __b28, char __b27, char __b26, char __b25, char __b24, char __b23, char __b22, char __b21, char __b20, char __b19, char __b18, char __b17, char __b16, char __b15, char __b14, char __b13, char __b12, char __b11, char __b10, char __b09, char __b08, char __b07, char __b06, char __b05, char __b04, char __b03, char __b02, char __b01, char __b00)
Constructs a 256-bit integer vector, initialized in reverse order with the specified 8-bit integral v...
static __inline __m128i __DEFAULT_FN_ATTRS _mm256_cvtpd_epi32(__m256d __a)
Converts a 256-bit vector of [4 x double] into a 128-bit vector of [4 x i32].
static __inline void __DEFAULT_FN_ATTRS _mm256_store_pd(double *__p, __m256d __a)
Stores double-precision floating point values from a 256-bit vector of [4 x double] to a 32-byte alig...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_zextsi128_si256(__m128i __a)
Constructs a 256-bit integer vector from a 128-bit integer vector.
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu_ps(float *__p, __m256 __a)
Stores single-precision floating point values from a 256-bit vector of [8 x float] to an unaligned me...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_unpacklo_pd(__m256d __a, __m256d __b)
Unpacks the even-indexed vector elements from two 256-bit vectors of [4 x double] and interleaves the...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_loadu2_m128(float const *__addr_hi, float const *__addr_lo)
Loads two 128-bit floating-point vectors of [4 x float] from unaligned memory locations and construct...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_sqrt_ps(__m256 __a)
Calculates the square roots of the values in a 256-bit vector of [8 x float].
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_maskload_pd(double const *__p, __m256i __m)
Conditionally loads double-precision floating point elements from a memory location pointed to by __p...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_castps_si256(__m256 __a)
Casts a 256-bit floating-point vector of [8 x float] into a 256-bit integer vector.
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_xor_pd(__m256d __a, __m256d __b)
Performs a bitwise XOR of two 256-bit vectors of [4 x double].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_cvtepi32_ps(__m256i __a)
Converts a vector of [8 x i32] into a vector of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_rcp_ps(__m256 __a)
Calculates the reciprocals of the values in a 256-bit vector of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_moveldup_ps(__m256 __a)
Moves and duplicates even-indexed values from a 256-bit vector of [8 x float] to float values in a 25...
static __inline __m128d __DEFAULT_FN_ATTRS _mm256_castpd256_pd128(__m256d __a)
Returns the lower 128 bits of a 256-bit floating-point vector of [4 x double] as a 128-bit floating-p...
static __inline int __DEFAULT_FN_ATTRS _mm256_movemask_pd(__m256d __a)
Extracts the sign bits of double-precision floating point elements in a 256-bit vector of [4 x double...
static __inline __m128 __DEFAULT_FN_ATTRS _mm256_cvtpd_ps(__m256d __a)
Converts a 256-bit vector of [4 x double] into a 128-bit vector of [4 x float].
static __inline__ __m256 __DEFAULT_FN_ATTRS _mm256_undefined_ps(void)
Create a 256-bit vector of [8 x float] with undefined values.
static __inline int __DEFAULT_FN_ATTRS _mm256_testnzc_ps(__m256 __a, __m256 __b)
Given two 256-bit floating-point vectors of [8 x float], perform an element-by-element comparison of ...
static __inline __m128 __DEFAULT_FN_ATTRS _mm256_castps256_ps128(__m256 __a)
Returns the lower 128 bits of a 256-bit floating-point vector of [8 x float] as a 128-bit floating-po...
static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_setr_ps(float __a, float __b, float __c, float __d, float __e, float __f, float __g, float __h)
Constructs a 256-bit floating-point vector of [8 x float], initialized in reverse order with the spec...
static __inline __m128 __DEFAULT_FN_ATTRS128 _mm_maskload_ps(float const *__p, __m128i __m)
Conditionally loads single-precision floating point elements from a memory location pointed to by __p...
static __inline __m128d __DEFAULT_FN_ATTRS128 _mm_maskload_pd(double const *__p, __m128i __m)
Conditionally loads double-precision floating point elements from a memory location pointed to by __p...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set1_epi8(char __b)
Constructs a 256-bit integer vector of [32 x i8], with each of the 8-bit integral vector elements set...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_permutevar_ps(__m256 __a, __m256i __c)
Copies the values stored in a 256-bit vector of [8 x float] as specified by the 256-bit integer vecto...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_cvtps_pd(__m128 __a)
Converts a 128-bit vector of [4 x float] into a 256-bit vector of [4 x double].
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu_si256(__m256i_u *__p, __m256i __a)
Stores integer values from a 256-bit integer vector to an unaligned memory location pointed to by __p...
#define _mm256_extractf128_ps(V, M)
Extracts either the upper or the lower 128 bits from a 256-bit vector of [8 x float],...
#define _mm256_extractf128_si256(V, M)
Extracts either the upper or the lower 128 bits from a 256-bit integer vector, as determined by the i...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_load_si256(__m256i const *__p)
Loads 256 bits of integer data from a 32-byte aligned memory location pointed to by __p into elements...
static __inline int __DEFAULT_FN_ATTRS _mm256_testnzc_si256(__m256i __a, __m256i __b)
Given two 256-bit integer vectors, perform a bit-by-bit comparison of the two source vectors.
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_blendv_pd(__m256d __a, __m256d __b, __m256d __c)
Merges 64-bit double-precision data values stored in either of the two 256-bit vectors of [4 x double...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_set_m128d(__m128d __hi, __m128d __lo)
Constructs a 256-bit floating-point vector of [4 x double] by concatenating two 128-bit floating-poin...
static __inline __m256d __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_set_pd(double __a, double __b, double __c, double __d)
Constructs a 256-bit floating-point vector of [4 x double] initialized with the specified double-prec...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_castpd_ps(__m256d __a)
Casts a 256-bit floating-point vector of [4 x double] into a 256-bit floating-point vector of [8 x fl...
static __inline void __DEFAULT_FN_ATTRS _mm256_stream_si256(void *__a, __m256i __b)
Moves integer data from a 256-bit integer vector to a 32-byte aligned memory location.
static __inline __m128 __DEFAULT_FN_ATTRS128 _mm_permutevar_ps(__m128 __a, __m128i __c)
Copies the values stored in a 128-bit vector of [4 x float] as specified by the 128-bit integer vecto...
static __inline__ __m256d __DEFAULT_FN_ATTRS _mm256_undefined_pd(void)
Create a 256-bit vector of [4 x double] with undefined values.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_rsqrt_ps(__m256 __a)
Calculates the reciprocal square roots of the values in a 256-bit vector of [8 x float].
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setr_epi16(short __w15, short __w14, short __w13, short __w12, short __w11, short __w10, short __w09, short __w08, short __w07, short __w06, short __w05, short __w04, short __w03, short __w02, short __w01, short __w00)
Constructs a 256-bit integer vector, initialized in reverse order with the specified 16-bit integral ...
#define __DEFAULT_FN_ATTRS
static __inline void __DEFAULT_FN_ATTRS128 _mm_maskstore_pd(double *__p, __m128i __m, __m128d __a)
Moves double-precision values from a 128-bit vector of [2 x double] to a memory location pointed to b...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_castpd128_pd256(__m128d __a)
Constructs a 256-bit floating-point vector of [4 x double] from a 128-bit floating-point vector of [2...
static __inline __m256d __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_set1_pd(double __w)
Constructs a 256-bit floating-point vector of [4 x double], with each of the four double-precision fl...
static __inline __m128i __DEFAULT_FN_ATTRS _mm256_cvttpd_epi32(__m256d __a)
Converts a 256-bit vector of [4 x double] into four signed truncated (rounded toward zero) 32-bit int...
static __inline__ __m256i __DEFAULT_FN_ATTRS _mm256_undefined_si256(void)
Create a 256-bit integer vector with undefined values.
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_cvtps_epi32(__m256 __a)
Converts a vector of [8 x float] into a vector of [8 x i32].
static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_set1_ps(float __w)
Constructs a 256-bit floating-point vector of [8 x float], with each of the eight single-precision fl...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set1_epi64x(long long __q)
Constructs a 256-bit integer vector of [4 x i64], with each of the 64-bit integral vector elements se...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_castsi256_pd(__m256i __a)
Casts a 256-bit integer vector into a 256-bit floating-point vector of [4 x double].
static __inline int __DEFAULT_FN_ATTRS128 _mm_testnzc_pd(__m128d __a, __m128d __b)
Given two 128-bit floating-point vectors of [2 x double], perform an element-by-element comparison of...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_min_ps(__m256 __a, __m256 __b)
Compares two 256-bit vectors of [8 x float] and returns the lesser of each pair of values.
static __inline int __DEFAULT_FN_ATTRS _mm256_testc_ps(__m256 __a, __m256 __b)
Given two 256-bit floating-point vectors of [8 x float], perform an element-by-element comparison of ...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_load_ps(float const *__p)
Loads 8 single-precision floating point values from a 32-byte aligned memory location pointed to by _...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_xor_ps(__m256 __a, __m256 __b)
Performs a bitwise XOR of two 256-bit vectors of [8 x float].
#define _mm256_extractf128_pd(V, M)
Extracts either the upper or the lower 128 bits from a 256-bit vector of [4 x double],...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_cvttps_epi32(__m256 __a)
Converts a vector of [8 x float] into eight signed truncated (rounded toward zero) 32-bit integers re...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_castps128_ps256(__m128 __a)
Constructs a 256-bit floating-point vector of [8 x float] from a 128-bit floating-point vector of [4 ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setr_m128i(__m128i __lo, __m128i __hi)
Constructs a 256-bit integer vector by concatenating two 128-bit integer vectors.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_sub_ps(__m256 __a, __m256 __b)
Subtracts two 256-bit vectors of [8 x float].
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set_epi64x(long long __a, long long __b, long long __c, long long __d)
Constructs a 256-bit integer vector initialized with the specified 64-bit integral values.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_broadcast_ss(float const *__a)
Loads a scalar single-precision floating point value from the specified address pointed to by __a and...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_max_ps(__m256 __a, __m256 __b)
Compares two 256-bit vectors of [8 x float] and returns the greater of each pair of values.
static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_setzero_ps(void)
Constructs a 256-bit floating-point vector of [8 x float] with all vector elements initialized to zer...
static __inline int __DEFAULT_FN_ATTRS _mm256_testnzc_pd(__m256d __a, __m256d __b)
Given two 256-bit floating-point vectors of [4 x double], perform an element-by-element comparison of...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_loadu2_m128d(double const *__addr_hi, double const *__addr_lo)
Loads two 128-bit floating-point vectors of [2 x double] from unaligned memory locations and construc...
static __inline float __DEFAULT_FN_ATTRS _mm256_cvtss_f32(__m256 __a)
Returns the first element of the input vector of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_movehdup_ps(__m256 __a)
Moves and duplicates odd-indexed values from a 256-bit vector of [8 x float] to float values in a 256...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_div_pd(__m256d __a, __m256d __b)
Divides two 256-bit vectors of [4 x double].
static __inline int __DEFAULT_FN_ATTRS128 _mm_testnzc_ps(__m128 __a, __m128 __b)
Given two 128-bit floating-point vectors of [4 x float], perform an element-by-element comparison of ...
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu2_m128d(double *__addr_hi, double *__addr_lo, __m256d __a)
Stores the upper and lower 128 bits of a 256-bit floating-point vector of [4 x double] into two diffe...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_sqrt_pd(__m256d __a)
Calculates the square roots of the values in a 256-bit vector of [4 x double].
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_addsub_pd(__m256d __a, __m256d __b)
Adds the even-indexed values and subtracts the odd-indexed values of two 256-bit vectors of [4 x doub...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_hadd_ps(__m256 __a, __m256 __b)
Horizontally adds the adjacent pairs of values contained in two 256-bit vectors of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_addsub_ps(__m256 __a, __m256 __b)
Adds the even-indexed values and subtracts the odd-indexed values of two 256-bit vectors of [8 x floa...
static __inline int __DEFAULT_FN_ATTRS _mm256_testz_pd(__m256d __a, __m256d __b)
Given two 256-bit floating-point vectors of [4 x double], perform an element-by-element comparison of...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_mul_pd(__m256d __a, __m256d __b)
Multiplies two 256-bit vectors of [4 x double].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_and_ps(__m256 __a, __m256 __b)
Performs a bitwise AND of two 256-bit vectors of [8 x float].
static __inline void __DEFAULT_FN_ATTRS _mm256_maskstore_pd(double *__p, __m256i __m, __m256d __a)
Moves double-precision values from a 256-bit vector of [4 x double] to a memory location pointed to b...
static __inline void __DEFAULT_FN_ATTRS _mm256_maskstore_ps(float *__p, __m256i __m, __m256 __a)
Moves single-precision floating point values from a 256-bit vector of [8 x float] to a memory locatio...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_or_ps(__m256 __a, __m256 __b)
Performs a bitwise OR of two 256-bit vectors of [8 x float].
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_loadu_pd(double const *__p)
Loads 4 double-precision floating point values from an unaligned memory location pointed to by __p in...
static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_set_ps(float __a, float __b, float __c, float __d, float __e, float __f, float __g, float __h)
Constructs a 256-bit floating-point vector of [8 x float] initialized with the specified single-preci...
static __inline int __DEFAULT_FN_ATTRS _mm256_testz_si256(__m256i __a, __m256i __b)
Given two 256-bit integer vectors, perform a bit-by-bit comparison of the two source vectors.
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_zextpd128_pd256(__m128d __a)
Constructs a 256-bit floating-point vector of [4 x double] from a 128-bit floating-point vector of [2...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setr_epi64x(long long __a, long long __b, long long __c, long long __d)
Constructs a 256-bit integer vector, initialized in reverse order with the specified 64-bit integral ...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_castps_pd(__m256 __a)
Casts a 256-bit floating-point vector of [8 x float] into a 256-bit floating-point vector of [4 x dou...
static __inline double __DEFAULT_FN_ATTRS _mm256_cvtsd_f64(__m256d __a)
Returns the first element of the input vector of [4 x double].
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_hadd_pd(__m256d __a, __m256d __b)
Horizontally adds the adjacent pairs of values contained in two 256-bit vectors of [4 x double].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_hsub_ps(__m256 __a, __m256 __b)
Horizontally subtracts the adjacent pairs of values contained in two 256-bit vectors of [8 x float].
static __inline int __DEFAULT_FN_ATTRS _mm256_testc_pd(__m256d __a, __m256d __b)
Given two 256-bit floating-point vectors of [4 x double], perform an element-by-element comparison of...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_cvtepi32_pd(__m128i __a)
Converts a vector of [4 x i32] into a vector of [4 x double].
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_setr_m128d(__m128d __lo, __m128d __hi)
Constructs a 256-bit floating-point vector of [4 x double] by concatenating two 128-bit floating-poin...
#define __DEFAULT_FN_ATTRS128
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_andnot_pd(__m256d __a, __m256d __b)
Performs a bitwise AND of two 256-bit vectors of [4 x double], using the one's complement of the valu...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_setr_epi32(int __i0, int __i1, int __i2, int __i3, int __i4, int __i5, int __i6, int __i7)
Constructs a 256-bit integer vector, initialized in reverse order with the specified 32-bit integral ...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
Merges 32-bit single-precision data values stored in either of the two 256-bit vectors of [8 x float]...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_unpackhi_pd(__m256d __a, __m256d __b)
Unpacks the odd-indexed vector elements from two 256-bit vectors of [4 x double] and interleaves them...
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu2_m128i(__m128i_u *__addr_hi, __m128i_u *__addr_lo, __m256i __a)
Stores the upper and lower 128 bits of a 256-bit integer vector into two different unaligned memory l...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_sub_pd(__m256d __a, __m256d __b)
Subtracts two 256-bit vectors of [4 x double].
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set_m128i(__m128i __hi, __m128i __lo)
Constructs a 256-bit integer vector by concatenating two 128-bit integer vectors.
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_or_pd(__m256d __a, __m256d __b)
Performs a bitwise OR of two 256-bit vectors of [4 x double].
#define __DEFAULT_FN_ATTRS_CONSTEXPR
static __inline int __DEFAULT_FN_ATTRS128 _mm_testz_pd(__m128d __a, __m128d __b)
Given two 128-bit floating-point vectors of [2 x double], perform an element-by-element comparison of...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_movedup_pd(__m256d __a)
Moves and duplicates double-precision floating point values from a 256-bit vector of [4 x double] to ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set_epi16(short __w15, short __w14, short __w13, short __w12, short __w11, short __w10, short __w09, short __w08, short __w07, short __w06, short __w05, short __w04, short __w03, short __w02, short __w01, short __w00)
Constructs a 256-bit integer vector initialized with the specified 16-bit integral values.
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_loadu2_m128i(__m128i_u const *__addr_hi, __m128i_u const *__addr_lo)
Loads two 128-bit integer vectors from unaligned memory locations and constructs a 256-bit integer ve...
static __inline __m128d __DEFAULT_FN_ATTRS128 _mm_permutevar_pd(__m128d __a, __m128i __c)
Copies the values in a 128-bit vector of [2 x double] as specified by the 128-bit integer vector oper...
static __inline __m256d __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_setzero_pd(void)
Constructs a 256-bit floating-point vector of [4 x double] with all vector elements initialized to ze...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_castpd_si256(__m256d __a)
Casts a 256-bit floating-point vector of [4 x double] into a 256-bit integer vector.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_maskload_ps(float const *__p, __m256i __m)
Conditionally loads single-precision floating point elements from a memory location pointed to by __p...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_div_ps(__m256 __a, __m256 __b)
Divides two 256-bit vectors of [8 x float].
static __inline int __DEFAULT_FN_ATTRS _mm256_cvtsi256_si32(__m256i __a)
Returns the first element of the input vector of [8 x i32].
static __inline void __DEFAULT_FN_ATTRS _mm256_stream_ps(void *__p, __m256 __a)
Moves single-precision floating point values from a 256-bit vector of [8 x float] to a 32-byte aligne...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_setr_m128(__m128 __lo, __m128 __hi)
Constructs a 256-bit floating-point vector of [8 x float] by concatenating two 128-bit floating-point...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_loadu_si256(__m256i_u const *__p)
Loads 256 bits of integer data from an unaligned memory location pointed to by __p into a 256-bit int...
static __inline void __DEFAULT_FN_ATTRS _mm256_store_si256(__m256i *__p, __m256i __a)
Stores integer values from a 256-bit integer vector to a 32-byte aligned memory location pointed to b...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_min_pd(__m256d __a, __m256d __b)
Compares two 256-bit vectors of [4 x double] and returns the lesser of each pair of values.
static __inline __m128i __DEFAULT_FN_ATTRS _mm256_castsi256_si128(__m256i __a)
Truncates a 256-bit integer vector into a 128-bit integer vector.
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set_epi32(int __i0, int __i1, int __i2, int __i3, int __i4, int __i5, int __i6, int __i7)
Constructs a 256-bit integer vector initialized with the specified 32-bit integral values.
static __inline __m256i __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_setzero_si256(void)
Constructs a 256-bit integer vector initialized to zero.
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_castsi128_si256(__m128i __a)
Constructs a 256-bit integer vector from a 128-bit integer vector.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_unpackhi_ps(__m256 __a, __m256 __b)
Unpacks the 32-bit vector elements 2, 3, 6 and 7 from each of the two 256-bit vectors of [8 x float] ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set_epi8(char __b31, char __b30, char __b29, char __b28, char __b27, char __b26, char __b25, char __b24, char __b23, char __b22, char __b21, char __b20, char __b19, char __b18, char __b17, char __b16, char __b15, char __b14, char __b13, char __b12, char __b11, char __b10, char __b09, char __b08, char __b07, char __b06, char __b05, char __b04, char __b03, char __b02, char __b01, char __b00)
Constructs a 256-bit integer vector initialized with the specified 8-bit integral values.
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_loadu_ps(float const *__p)
Loads 8 single-precision floating point values from an unaligned memory location pointed to by __p in...
static __inline __m128 __DEFAULT_FN_ATTRS128 _mm_broadcast_ss(float const *__a)
Loads a scalar single-precision floating point value from the specified address pointed to by __a and...
static __inline int __DEFAULT_FN_ATTRS128 _mm_testc_ps(__m128 __a, __m128 __b)
Given two 128-bit floating-point vectors of [4 x float], perform an element-by-element comparison of ...
static __inline int __DEFAULT_FN_ATTRS _mm256_movemask_ps(__m256 __a)
Extracts the sign bits of single-precision floating point elements in a 256-bit vector of [8 x float]...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_mul_ps(__m256 __a, __m256 __b)
Multiplies two 256-bit vectors of [8 x float].
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_unpacklo_ps(__m256 __a, __m256 __b)
Unpacks the 32-bit vector elements 0, 1, 4 and 5 from each of the two 256-bit vectors of [8 x float] ...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_permutevar_pd(__m256d __a, __m256i __c)
Copies the values in a 256-bit vector of [4 x double] as specified by the 256-bit integer vector oper...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_max_pd(__m256d __a, __m256d __b)
Compares two 256-bit vectors of [4 x double] and returns the greater of each pair of values.
static __inline int __DEFAULT_FN_ATTRS128 _mm_testz_ps(__m128 __a, __m128 __b)
Given two 128-bit floating-point vectors of [4 x float], perform an element-by-element comparison of ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set1_epi16(short __w)
Constructs a 256-bit integer vector of [16 x i16], with each of the 16-bit integral vector elements s...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_lddqu_si256(__m256i_u const *__p)
Loads 256 bits of integer data from an unaligned memory location pointed to by __p into a 256-bit int...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_and_pd(__m256d __a, __m256d __b)
Performs a bitwise AND of two 256-bit vectors of [4 x double].
static __inline void __DEFAULT_FN_ATTRS _mm256_store_ps(float *__p, __m256 __a)
Stores single-precision floating point values from a 256-bit vector of [8 x float] to a 32-byte align...
static __inline __m256d __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_setr_pd(double __a, double __b, double __c, double __d)
Constructs a 256-bit floating-point vector of [4 x double], initialized in reverse order with the spe...
static __inline void __DEFAULT_FN_ATTRS _mm256_storeu2_m128(float *__addr_hi, float *__addr_lo, __m256 __a)
Stores the upper and lower 128 bits of a 256-bit floating-point vector of [8 x float] into two differ...
static __inline __m256i __DEFAULT_FN_ATTRS _mm256_set1_epi32(int __i)
Constructs a 256-bit integer vector of [8 x i32], with each of the 32-bit integral vector elements se...
double __v4df __attribute__((__vector_size__(32)))
static __inline int __DEFAULT_FN_ATTRS128 _mm_testc_pd(__m128d __a, __m128d __b)
Given two 128-bit floating-point vectors of [2 x double], perform an element-by-element comparison of...
static __inline int __DEFAULT_FN_ATTRS _mm256_testz_ps(__m256 __a, __m256 __b)
Given two 256-bit floating-point vectors of [8 x float], perform an element-by-element comparison of ...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_zextps128_ps256(__m128 __a)
Constructs a 256-bit floating-point vector of [8 x float] from a 128-bit floating-point vector of [4 ...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_add_pd(__m256d __a, __m256d __b)
Adds two 256-bit vectors of [4 x double].
static __inline void __DEFAULT_FN_ATTRS128 _mm_maskstore_ps(float *__p, __m128i __m, __m128 __a)
Moves single-precision floating point values from a 128-bit vector of [4 x float] to a memory locatio...
static __inline __m256d __DEFAULT_FN_ATTRS _mm256_load_pd(double const *__p)
Loads 4 double-precision floating point values from a 32-byte aligned memory location pointed to by _...
static __inline __m256 __DEFAULT_FN_ATTRS _mm256_andnot_ps(__m256 __a, __m256 __b)
Performs a bitwise AND of two 256-bit vectors of [8 x float], using the one's complement of the value...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_loadu_pd(double const *__dp)
Loads a 128-bit floating-point vector of [2 x double] from an unaligned memory location.
static __inline__ __m128i __DEFAULT_FN_ATTRS_CONSTEXPR _mm_setzero_si128(void)
Creates a 128-bit integer vector initialized to zero.
static __inline__ void int __a
static __inline__ __m128d __DEFAULT_FN_ATTRS_CONSTEXPR _mm_setzero_pd(void)
Constructs a 128-bit floating-point vector of [2 x double] initialized to zero.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_loadu_si128(__m128i_u const *__p)
Moves packed integer values from an unaligned 128-bit memory location to elements in a 128-bit intege...
static __inline__ void __DEFAULT_FN_ATTRS _mm_storeu_pd(double *__dp, __m128d __a)
Stores a 128-bit vector of [2 x double] into an unaligned memory location.
static __inline__ void __DEFAULT_FN_ATTRS _mm_storeu_si128(__m128i_u *__p, __m128i __b)
Stores a 128-bit integer vector to an unaligned memory location.
struct __storeu_i16 *__P __v
static __inline__ void __DEFAULT_FN_ATTRS _mm_storeu_ps(float *__p, __m128 __a)
Stores a 128-bit vector of [4 x float] to an unaligned memory location.
static __inline__ __m128 __DEFAULT_FN_ATTRS_CONSTEXPR _mm_setzero_ps(void)
Constructs a 128-bit floating-point vector of [4 x float] initialized to zero.
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_loadu_ps(const float *__p)
Loads a 128-bit floating-point vector of [4 x float] from an unaligned memory location.