30 typedef long long __m128i
__attribute__((__vector_size__(16)));
34 typedef long long __v2di
__attribute__ ((__vector_size__ (16)));
39 typedef unsigned long long __v2du
__attribute__ ((__vector_size__ (16)));
40 typedef unsigned short __v8hu
__attribute__((__vector_size__(16)));
41 typedef unsigned char __v16qu
__attribute__((__vector_size__(16)));
45 typedef signed char __v16qs
__attribute__((__vector_size__(16)));
50 #define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("sse2"))) 89 return (__m128d)((__v2df)__a + (__v2df)
__b);
131 return (__m128d)((__v2df)__a - (__v2df)
__b);
172 return (__m128d)((__v2df)__a * (__v2df)
__b);
215 return (__m128d)((__v2df)__a / (__v2df)
__b);
240 __m128d
__c = __builtin_ia32_sqrtsd((__v2df)__b);
241 return (__m128d) { __c[0], __a[1] };
258 return __builtin_ia32_sqrtpd((__v2df)__a);
282 return __builtin_ia32_minsd((__v2df)__a, (__v2df)__b);
302 return __builtin_ia32_minpd((__v2df)__a, (__v2df)__b);
326 return __builtin_ia32_maxsd((__v2df)__a, (__v2df)__b);
346 return __builtin_ia32_maxpd((__v2df)__a, (__v2df)__b);
364 return (__m128d)((__v2du)__a & (__v2du)
__b);
385 return (__m128d)(~(__v2du)__a & (__v2du)
__b);
403 return (__m128d)((__v2du)__a | (__v2du)
__b);
421 return (__m128d)((__v2du)__a ^ (__v2du)
__b);
440 return (__m128d)__builtin_ia32_cmpeqpd((__v2df)__a, (__v2df)__b);
460 return (__m128d)__builtin_ia32_cmpltpd((__v2df)__a, (__v2df)__b);
481 return (__m128d)__builtin_ia32_cmplepd((__v2df)__a, (__v2df)__b);
502 return (__m128d)__builtin_ia32_cmpltpd((__v2df)__b, (__v2df)__a);
523 return (__m128d)__builtin_ia32_cmplepd((__v2df)__b, (__v2df)__a);
546 return (__m128d)__builtin_ia32_cmpordpd((__v2df)__a, (__v2df)__b);
570 return (__m128d)__builtin_ia32_cmpunordpd((__v2df)__a, (__v2df)__b);
591 return (__m128d)__builtin_ia32_cmpneqpd((__v2df)__a, (__v2df)__b);
612 return (__m128d)__builtin_ia32_cmpnltpd((__v2df)__a, (__v2df)__b);
633 return (__m128d)__builtin_ia32_cmpnlepd((__v2df)__a, (__v2df)__b);
654 return (__m128d)__builtin_ia32_cmpnltpd((__v2df)__b, (__v2df)__a);
675 return (__m128d)__builtin_ia32_cmpnlepd((__v2df)__b, (__v2df)__a);
698 return (__m128d)__builtin_ia32_cmpeqsd((__v2df)__a, (__v2df)__b);
723 return (__m128d)__builtin_ia32_cmpltsd((__v2df)__a, (__v2df)__b);
748 return (__m128d)__builtin_ia32_cmplesd((__v2df)__a, (__v2df)__b);
773 __m128d
__c = __builtin_ia32_cmpltsd((__v2df)__b, (__v2df)__a);
774 return (__m128d) { __c[0], __a[1] };
799 __m128d
__c = __builtin_ia32_cmplesd((__v2df)__b, (__v2df)__a);
800 return (__m128d) { __c[0], __a[1] };
827 return (__m128d)__builtin_ia32_cmpordsd((__v2df)__a, (__v2df)__b);
855 return (__m128d)__builtin_ia32_cmpunordsd((__v2df)__a, (__v2df)__b);
880 return (__m128d)__builtin_ia32_cmpneqsd((__v2df)__a, (__v2df)__b);
905 return (__m128d)__builtin_ia32_cmpnltsd((__v2df)__a, (__v2df)__b);
930 return (__m128d)__builtin_ia32_cmpnlesd((__v2df)__a, (__v2df)__b);
955 __m128d
__c = __builtin_ia32_cmpnltsd((__v2df)__b, (__v2df)__a);
956 return (__m128d) { __c[0], __a[1] };
981 __m128d
__c = __builtin_ia32_cmpnlesd((__v2df)__b, (__v2df)__a);
982 return (__m128d) { __c[0], __a[1] };
1006 return __builtin_ia32_comisdeq((__v2df)__a, (__v2df)__b);
1032 return __builtin_ia32_comisdlt((__v2df)__a, (__v2df)__b);
1058 return __builtin_ia32_comisdle((__v2df)__a, (__v2df)__b);
1084 return __builtin_ia32_comisdgt((__v2df)__a, (__v2df)__b);
1110 return __builtin_ia32_comisdge((__v2df)__a, (__v2df)__b);
1136 return __builtin_ia32_comisdneq((__v2df)__a, (__v2df)__b);
1160 return __builtin_ia32_ucomisdeq((__v2df)__a, (__v2df)__b);
1186 return __builtin_ia32_ucomisdlt((__v2df)__a, (__v2df)__b);
1212 return __builtin_ia32_ucomisdle((__v2df)__a, (__v2df)__b);
1238 return __builtin_ia32_ucomisdgt((__v2df)__a, (__v2df)__b);
1264 return __builtin_ia32_ucomisdge((__v2df)__a, (__v2df)__b);
1290 return __builtin_ia32_ucomisdneq((__v2df)__a, (__v2df)__b);
1309 return __builtin_ia32_cvtpd2ps((__v2df)__a);
1329 return (__m128d) __builtin_convertvector(
1330 __builtin_shufflevector((__v4sf)__a, (__v4sf)__a, 0, 1), __v2df);
1352 return (__m128d) __builtin_convertvector(
1353 __builtin_shufflevector((__v4si)__a, (__v4si)__a, 0, 1), __v2df);
1372 return __builtin_ia32_cvtpd2dq((__v2df)__a);
1389 return __builtin_ia32_cvtsd2si((__v2df)__a);
1414 return (__m128)__builtin_ia32_cvtsd2ss((__v4sf)__a, (__v2df)__b);
1487 return (__m128i)__builtin_ia32_cvttpd2dq((__v2df)__a);
1505 return __builtin_ia32_cvttsd2si((__v2df)__a);
1522 return (__m64)__builtin_ia32_cvtpd2pi((__v2df)__a);
1542 return (__m64)__builtin_ia32_cvttpd2pi((__v2df)__a);
1559 return __builtin_ia32_cvtpi2pd((__v2si)__a);
1593 return *(__m128d*)__dp;
1611 struct __mm_load1_pd_struct {
1614 double __u = ((
struct __mm_load1_pd_struct*)__dp)->__u;
1615 return (__m128d){ __u, __u };
1618 #define _mm_load_pd1(dp) _mm_load1_pd(dp) 1637 __m128d __u = *(__m128d*)__dp;
1638 return __builtin_shufflevector((__v2df)__u, (__v2df)__u, 1, 0);
1658 return ((
struct __loadu_pd*)__dp)->__v;
1675 struct __loadu_si64 {
1678 long long __u = ((
struct __loadu_si64*)__a)->__v;
1679 return (__m128i){__u, 0L};
1696 struct __mm_load_sd_struct {
1699 double __u = ((
struct __mm_load_sd_struct*)__dp)->__u;
1700 return (__m128d){ __u, 0 };
1723 struct __mm_loadh_pd_struct {
1726 double __u = ((
struct __mm_loadh_pd_struct*)__dp)->__u;
1727 return (__m128d){ __a[0], __u };
1750 struct __mm_loadl_pd_struct {
1753 double __u = ((
struct __mm_loadl_pd_struct*)__dp)->__u;
1754 return (__m128d){ __u, __a[1] };
1771 return (__m128d)__builtin_ia32_undef128();
1791 return (__m128d){ __w, 0 };
1809 return (__m128d){ __w, __w };
1847 return (__m128d){
__x, __w };
1868 return (__m128d){ __w, __x };
1883 return (__m128d){ 0, 0 };
1904 return (__m128d){ __b[0], __a[1] };
1921 struct __mm_store_sd_struct {
1924 ((
struct __mm_store_sd_struct*)__dp)->__u = __a[0];
1943 *(__m128d*)__dp = __a;
1963 __a = __builtin_shufflevector((__v2df)__a, (__v2df)__a, 0, 0);
2002 struct __storeu_pd {
2005 ((
struct __storeu_pd*)__dp)->__v = __a;
2025 __a = __builtin_shufflevector((__v2df)__a, (__v2df)__a, 1, 0);
2026 *(__m128d *)__dp = __a;
2043 struct __mm_storeh_pd_struct {
2046 ((
struct __mm_storeh_pd_struct*)__dp)->__u = __a[1];
2063 struct __mm_storeh_pd_struct {
2066 ((
struct __mm_storeh_pd_struct*)__dp)->__u = __a[0];
2088 return (__m128i)((__v16qu)__a + (__v16qu)
__b);
2110 return (__m128i)((__v8hu)__a + (__v8hu)
__b);
2132 return (__m128i)((__v4su)__a + (__v4su)
__b);
2150 return (__m64)__builtin_ia32_paddq((__v1di)__a, (__v1di)__b);
2172 return (__m128i)((__v2du)__a + (__v2du)
__b);
2193 return (__m128i)__builtin_ia32_paddsb128((__v16qi)__a, (__v16qi)__b);
2215 return (__m128i)__builtin_ia32_paddsw128((__v8hi)__a, (__v8hi)__b);
2236 return (__m128i)__builtin_ia32_paddusb128((__v16qi)__a, (__v16qi)__b);
2257 return (__m128i)__builtin_ia32_paddusw128((__v8hi)__a, (__v8hi)__b);
2277 typedef unsigned short __v16hu
__attribute__ ((__vector_size__ (32)));
2278 return (__m128i)__builtin_convertvector(
2279 ((__builtin_convertvector((__v16qu)__a, __v16hu) +
2280 __builtin_convertvector((__v16qu)__b, __v16hu)) + 1)
2301 typedef unsigned int __v8su
__attribute__ ((__vector_size__ (32)));
2302 return (__m128i)__builtin_convertvector(
2303 ((__builtin_convertvector((__v8hu)__a, __v8su) +
2304 __builtin_convertvector((__v8hu)__b, __v8su)) + 1)
2331 return (__m128i)__builtin_ia32_pmaddwd128((__v8hi)__a, (__v8hi)__b);
2351 return (__m128i)__builtin_ia32_pmaxsw128((__v8hi)__a, (__v8hi)__b);
2371 return (__m128i)__builtin_ia32_pmaxub128((__v16qi)__a, (__v16qi)__b);
2391 return (__m128i)__builtin_ia32_pminsw128((__v8hi)__a, (__v8hi)__b);
2411 return (__m128i)__builtin_ia32_pminub128((__v16qi)__a, (__v16qi)__b);
2431 return (__m128i)__builtin_ia32_pmulhw128((__v8hi)__a, (__v8hi)__b);
2451 return (__m128i)__builtin_ia32_pmulhuw128((__v8hi)__a, (__v8hi)__b);
2471 return (__m128i)((__v8hu)__a * (__v8hu)
__b);
2490 return __builtin_ia32_pmuludq((__v2si)__a, (__v2si)__b);
2509 return __builtin_ia32_pmuludq128((__v4si)__a, (__v4si)__b);
2531 return __builtin_ia32_psadbw128((__v16qi)__a, (__v16qi)__b);
2549 return (__m128i)((__v16qu)__a - (__v16qu)
__b);
2567 return (__m128i)((__v8hu)__a - (__v8hu)
__b);
2585 return (__m128i)((__v4su)__a - (__v4su)
__b);
2604 return (__m64)__builtin_ia32_psubq((__v1di)__a, (__v1di)__b);
2622 return (__m128i)((__v2du)__a - (__v2du)
__b);
2643 return (__m128i)__builtin_ia32_psubsb128((__v16qi)__a, (__v16qi)__b);
2664 return (__m128i)__builtin_ia32_psubsw128((__v8hi)__a, (__v8hi)__b);
2684 return (__m128i)__builtin_ia32_psubusb128((__v16qi)__a, (__v16qi)__b);
2704 return (__m128i)__builtin_ia32_psubusw128((__v8hi)__a, (__v8hi)__b);
2722 return (__m128i)((__v2du)__a & (__v2du)
__b);
2742 return (__m128i)(~(__v2du)__a & (__v2du)
__b);
2759 return (__m128i)((__v2du)__a | (__v2du)
__b);
2777 return (__m128i)((__v2du)__a ^ (__v2du)
__b);
2797 #define _mm_slli_si128(a, imm) __extension__ ({ \ 2798 (__m128i)__builtin_shufflevector( \ 2799 (__v16qi)_mm_setzero_si128(), \ 2800 (__v16qi)(__m128i)(a), \ 2801 ((char)(imm)&0xF0) ? 0 : 16 - (char)(imm), \ 2802 ((char)(imm)&0xF0) ? 1 : 17 - (char)(imm), \ 2803 ((char)(imm)&0xF0) ? 2 : 18 - (char)(imm), \ 2804 ((char)(imm)&0xF0) ? 3 : 19 - (char)(imm), \ 2805 ((char)(imm)&0xF0) ? 4 : 20 - (char)(imm), \ 2806 ((char)(imm)&0xF0) ? 5 : 21 - (char)(imm), \ 2807 ((char)(imm)&0xF0) ? 6 : 22 - (char)(imm), \ 2808 ((char)(imm)&0xF0) ? 7 : 23 - (char)(imm), \ 2809 ((char)(imm)&0xF0) ? 8 : 24 - (char)(imm), \ 2810 ((char)(imm)&0xF0) ? 9 : 25 - (char)(imm), \ 2811 ((char)(imm)&0xF0) ? 10 : 26 - (char)(imm), \ 2812 ((char)(imm)&0xF0) ? 11 : 27 - (char)(imm), \ 2813 ((char)(imm)&0xF0) ? 12 : 28 - (char)(imm), \ 2814 ((char)(imm)&0xF0) ? 13 : 29 - (char)(imm), \ 2815 ((char)(imm)&0xF0) ? 14 : 30 - (char)(imm), \ 2816 ((char)(imm)&0xF0) ? 15 : 31 - (char)(imm)); }) 2818 #define _mm_bslli_si128(a, imm) \ 2819 _mm_slli_si128((a), (imm)) 2837 return (__m128i)__builtin_ia32_psllwi128((__v8hi)__a, __count);
2856 return (__m128i)__builtin_ia32_psllw128((__v8hi)__a, (__v8hi)__count);
2875 return (__m128i)__builtin_ia32_pslldi128((__v4si)__a, __count);
2894 return (__m128i)__builtin_ia32_pslld128((__v4si)__a, (__v4si)__count);
2913 return __builtin_ia32_psllqi128((__v2di)__a, __count);
2932 return __builtin_ia32_psllq128((__v2di)__a, (__v2di)__count);
2952 return (__m128i)__builtin_ia32_psrawi128((__v8hi)__a, __count);
2972 return (__m128i)__builtin_ia32_psraw128((__v8hi)__a, (__v8hi)__count);
2992 return (__m128i)__builtin_ia32_psradi128((__v4si)__a, __count);
3012 return (__m128i)__builtin_ia32_psrad128((__v4si)__a, (__v4si)__count);
3032 #define _mm_srli_si128(a, imm) __extension__ ({ \ 3033 (__m128i)__builtin_shufflevector( \ 3034 (__v16qi)(__m128i)(a), \ 3035 (__v16qi)_mm_setzero_si128(), \ 3036 ((char)(imm)&0xF0) ? 16 : (char)(imm) + 0, \ 3037 ((char)(imm)&0xF0) ? 17 : (char)(imm) + 1, \ 3038 ((char)(imm)&0xF0) ? 18 : (char)(imm) + 2, \ 3039 ((char)(imm)&0xF0) ? 19 : (char)(imm) + 3, \ 3040 ((char)(imm)&0xF0) ? 20 : (char)(imm) + 4, \ 3041 ((char)(imm)&0xF0) ? 21 : (char)(imm) + 5, \ 3042 ((char)(imm)&0xF0) ? 22 : (char)(imm) + 6, \ 3043 ((char)(imm)&0xF0) ? 23 : (char)(imm) + 7, \ 3044 ((char)(imm)&0xF0) ? 24 : (char)(imm) + 8, \ 3045 ((char)(imm)&0xF0) ? 25 : (char)(imm) + 9, \ 3046 ((char)(imm)&0xF0) ? 26 : (char)(imm) + 10, \ 3047 ((char)(imm)&0xF0) ? 27 : (char)(imm) + 11, \ 3048 ((char)(imm)&0xF0) ? 28 : (char)(imm) + 12, \ 3049 ((char)(imm)&0xF0) ? 29 : (char)(imm) + 13, \ 3050 ((char)(imm)&0xF0) ? 30 : (char)(imm) + 14, \ 3051 ((char)(imm)&0xF0) ? 31 : (char)(imm) + 15); }) 3053 #define _mm_bsrli_si128(a, imm) \ 3054 _mm_srli_si128((a), (imm)) 3072 return (__m128i)__builtin_ia32_psrlwi128((__v8hi)__a, __count);
3091 return (__m128i)__builtin_ia32_psrlw128((__v8hi)__a, (__v8hi)__count);
3110 return (__m128i)__builtin_ia32_psrldi128((__v4si)__a, __count);
3129 return (__m128i)__builtin_ia32_psrld128((__v4si)__a, (__v4si)__count);
3148 return __builtin_ia32_psrlqi128((__v2di)__a, __count);
3167 return __builtin_ia32_psrlq128((__v2di)__a, (__v2di)__count);
3186 return (__m128i)((__v16qi)__a == (__v16qi)
__b);
3205 return (__m128i)((__v8hi)__a == (__v8hi)
__b);
3224 return (__m128i)((__v4si)__a == (__v4si)
__b);
3246 return (__m128i)((__v16qs)__a > (__v16qs)
__b);
3267 return (__m128i)((__v8hi)__a > (__v8hi)
__b);
3288 return (__m128i)((__v4si)__a > (__v4si)
__b);
3373 _mm_cvtsi64_sd(__m128d __a,
long long __b)
3391 _mm_cvtsd_si64(__m128d __a)
3393 return __builtin_ia32_cvtsd2si64((__v2df)__a);
3409 _mm_cvttsd_si64(__m128d __a)
3411 return __builtin_ia32_cvttsd2si64((__v2df)__a);
3427 return __builtin_ia32_cvtdq2ps((__v4si)__a);
3443 return (__m128i)__builtin_ia32_cvtps2dq((__v4sf)__a);
3460 return (__m128i)__builtin_ia32_cvttps2dq((__v4sf)__a);
3476 return (__m128i)(__v4si){ __a, 0, 0, 0 };
3491 _mm_cvtsi64_si128(
long long __a)
3493 return (__m128i){ __a, 0 };
3511 __v4si __b = (__v4si)__a;
3528 _mm_cvtsi128_si64(__m128i __a)
3563 struct __loadu_si128 {
3566 return ((
struct __loadu_si128*)__p)->__v;
3584 struct __mm_loadl_epi64_struct {
3587 return (__m128i) { ((
struct __mm_loadl_epi64_struct*)__p)->__u, 0};
3602 return (__m128i)__builtin_ia32_undef128();
3624 return (__m128i){ __q0, __q1 };
3646 return (__m128i){ (
long long)__q0, (
long long)__q1 };
3674 return (__m128i)(__v4si){ __i0, __i1, __i2, __i3};
3712 _mm_set_epi16(
short __w7,
short __w6,
short __w5,
short __w4,
short __w3,
short __w2,
short __w1,
short __w0)
3714 return (__m128i)(__v8hi){ __w0, __w1, __w2, __w3, __w4, __w5, __w6, __w7 };
3760 _mm_set_epi8(
char __b15,
char __b14,
char __b13,
char __b12,
char __b11,
char __b10,
char __b9,
char __b8,
char __b7,
char __b6,
char __b5,
char __b4,
char __b3,
char __b2,
char __b1,
char __b0)
3762 return (__m128i)(__v16qi){ __b0, __b1, __b2, __b3, __b4, __b5, __b6, __b7, __b8, __b9, __b10, __b11, __b12, __b13, __b14, __b15 };
3781 return (__m128i){ __q, __q };
3800 return (__m128i){ (
long long)__q, (
long long)__q };
3819 return (__m128i)(__v4si){ __i, __i, __i, __i };
3838 return (__m128i)(__v8hi){ __w, __w, __w, __w, __w, __w, __w, __w };
3857 return (__m128i)(__v16qi){
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b,
__b, __b };
3877 return (__m128i){ (
long long)__q0, (
long long)__q1 };
3900 return (__m128i)(__v4si){ __i0, __i1, __i2, __i3};
3929 _mm_setr_epi16(
short __w0,
short __w1,
short __w2,
short __w3,
short __w4,
short __w5,
short __w6,
short __w7)
3931 return (__m128i)(__v8hi){ __w0, __w1, __w2, __w3, __w4, __w5, __w6, __w7 };
3976 _mm_setr_epi8(
char __b0,
char __b1,
char __b2,
char __b3,
char __b4,
char __b5,
char __b6,
char __b7,
char __b8,
char __b9,
char __b10,
char __b11,
char __b12,
char __b13,
char __b14,
char __b15)
3978 return (__m128i)(__v16qi){ __b0, __b1, __b2, __b3, __b4, __b5, __b6, __b7, __b8, __b9, __b10, __b11, __b12, __b13, __b14, __b15 };
3992 return (__m128i){ 0LL, 0LL };
4026 struct __storeu_si128 {
4029 ((
struct __storeu_si128*)__p)->__v =
__b;
4056 __builtin_ia32_maskmovdqu((__v16qi)__d, (__v16qi)__n, __p);
4075 struct __mm_storel_epi64_struct {
4078 ((
struct __mm_storel_epi64_struct*)__p)->__u = __a[0];
4098 __builtin_nontemporal_store((__v2df)__a, (__v2df*)__p);
4117 __builtin_nontemporal_store((__v2di)__a, (__v2di*)__p);
4136 __builtin_ia32_movnti(__p, __a);
4154 _mm_stream_si64(
long long *
__p,
long long __a)
4156 __builtin_ia32_movnti64(__p, __a);
4160 #if defined(__cplusplus) 4198 #if defined(__cplusplus) 4227 return (__m128i)__builtin_ia32_packsswb128((__v8hi)__a, (__v8hi)__b);
4255 return (__m128i)__builtin_ia32_packssdw128((__v4si)__a, (__v4si)__b);
4283 return (__m128i)__builtin_ia32_packuswb128((__v8hi)__a, (__v8hi)__b);
4311 __v8hi __b = (__v8hi)__a;
4312 return (
unsigned short)__b[__imm & 7];
4338 __v8hi
__c = (__v8hi)__a;
4339 __c[__imm & 7] =
__b;
4340 return (__m128i)
__c;
4358 return __builtin_ia32_pmovmskb128((__v16qi)__a);
4389 #define _mm_shuffle_epi32(a, imm) __extension__ ({ \ 4390 (__m128i)__builtin_shufflevector((__v4si)(__m128i)(a), \ 4391 (__v4si)_mm_undefined_si128(), \ 4392 ((imm) >> 0) & 0x3, ((imm) >> 2) & 0x3, \ 4393 ((imm) >> 4) & 0x3, ((imm) >> 6) & 0x3); }) 4422 #define _mm_shufflelo_epi16(a, imm) __extension__ ({ \ 4423 (__m128i)__builtin_shufflevector((__v8hi)(__m128i)(a), \ 4424 (__v8hi)_mm_undefined_si128(), \ 4425 ((imm) >> 0) & 0x3, ((imm) >> 2) & 0x3, \ 4426 ((imm) >> 4) & 0x3, ((imm) >> 6) & 0x3, \ 4456 #define _mm_shufflehi_epi16(a, imm) __extension__ ({ \ 4457 (__m128i)__builtin_shufflevector((__v8hi)(__m128i)(a), \ 4458 (__v8hi)_mm_undefined_si128(), \ 4460 4 + (((imm) >> 0) & 0x3), \ 4461 4 + (((imm) >> 2) & 0x3), \ 4462 4 + (((imm) >> 4) & 0x3), \ 4463 4 + (((imm) >> 6) & 0x3)); }) 4497 return (__m128i)__builtin_shufflevector((__v16qi)__a, (__v16qi)__b, 8, 16+8, 9, 16+9, 10, 16+10, 11, 16+11, 12, 16+12, 13, 16+13, 14, 16+14, 15, 16+15);
4524 return (__m128i)__builtin_shufflevector((__v8hi)__a, (__v8hi)__b, 4, 8+4, 5, 8+5, 6, 8+6, 7, 8+7);
4547 return (__m128i)__builtin_shufflevector((__v4si)__a, (__v4si)__b, 2, 4+2, 3, 4+3);
4568 return (__m128i)__builtin_shufflevector((__v2di)__a, (__v2di)__b, 1, 2+1);
4603 return (__m128i)__builtin_shufflevector((__v16qi)__a, (__v16qi)__b, 0, 16+0, 1, 16+1, 2, 16+2, 3, 16+3, 4, 16+4, 5, 16+5, 6, 16+6, 7, 16+7);
4631 return (__m128i)__builtin_shufflevector((__v8hi)__a, (__v8hi)__b, 0, 8+0, 1, 8+1, 2, 8+2, 3, 8+3);
4654 return (__m128i)__builtin_shufflevector((__v4si)__a, (__v4si)__b, 0, 4+0, 1, 4+1);
4675 return (__m128i)__builtin_shufflevector((__v2di)__a, (__v2di)__b, 0, 2+0);
4692 return (__m64)__a[0];
4709 return (__m128i){ (
long long)__a, 0 };
4727 return __builtin_shufflevector((__v2di)__a, (__m128i){ 0 }, 0, 2);
4748 return __builtin_shufflevector((__v2df)__a, (__v2df)__b, 1, 2+1);
4769 return __builtin_shufflevector((__v2df)__a, (__v2df)__b, 0, 2+0);
4788 return __builtin_ia32_movmskpd((__v2df)__a);
4816 #define _mm_shuffle_pd(a, b, i) __extension__ ({ \ 4817 (__m128d)__builtin_shufflevector((__v2df)(__m128d)(a), (__v2df)(__m128d)(b), \ 4818 0 + (((i) >> 0) & 0x1), \ 4819 2 + (((i) >> 1) & 0x1)); }) 4852 return (__m128i)__a;
4869 return (__m128d)__a;
4886 return (__m128i)__a;
4920 return (__m128d)__a;
4923 #if defined(__cplusplus) 4936 #if defined(__cplusplus) 4939 #undef __DEFAULT_FN_ATTRS 4941 #define _MM_SHUFFLE2(x, y) (((x) << 1) | (y)) 4943 #define _MM_DENORMALS_ZERO_ON (0x0040) 4944 #define _MM_DENORMALS_ZERO_OFF (0x0000) 4946 #define _MM_DENORMALS_ZERO_MASK (0x0040) 4948 #define _MM_GET_DENORMALS_ZERO_MODE() (_mm_getcsr() & _MM_DENORMALS_ZERO_MASK) 4949 #define _MM_SET_DENORMALS_ZERO_MODE(x) (_mm_setcsr((_mm_getcsr() & ~_MM_DENORMALS_ZERO_MASK) | (x)))
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpgt_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_slli_epi32(__m128i __a, int __count)
Left-shifts each 32-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srli_epi16(__m128i __a, int __count)
Right-shifts each of 16-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ int __DEFAULT_FN_ATTRS _mm_cvtsd_si32(__m128d __a)
Converts the low-order element of a 128-bit vector of [2 x double] into a 32-bit signed integer value...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomige_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set_epi16(short __w7, short __w6, short __w5, short __w4, short __w3, short __w2, short __w1, short __w0)
Initializes the 16-bit values in a 128-bit vector of [8 x i16] with the specified 16-bit integer valu...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_sqrt_pd(__m128d __a)
Calculates the square root of the each of two values stored in a 128-bit vector of [2 x double]...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setr_epi8(char __b0, char __b1, char __b2, char __b3, char __b4, char __b5, char __b6, char __b7, char __b8, char __b9, char __b10, char __b11, char __b12, char __b13, char __b14, char __b15)
Constructs a 128-bit integer vector, initialized in reverse order with the specified 8-bit integral v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sll_epi32(__m128i __a, __m128i __count)
Left-shifts each 32-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srai_epi16(__m128i __a, int __count)
Right-shifts each 16-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomile_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cvttpd_epi32(__m128d __a)
Converts the two double-precision floating-point elements of a 128-bit vector of [2 x double] into tw...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtepi32_pd(__m128i __a)
Converts the lower two integer elements of a 128-bit vector of [4 x i32] into two double-precision fl...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set1_epi64(__m64 __q)
Initializes both values in a 128-bit vector of [2 x i64] with the specified 64-bit value...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srli_epi32(__m128i __a, int __count)
Right-shifts each of 32-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setr_epi32(int __i0, int __i1, int __i2, int __i3)
Constructs a 128-bit integer vector, initialized in reverse order with the specified 32-bit integral ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_packs_epi32(__m128i __a, __m128i __b)
Converts 32-bit signed integers from both 128-bit integer vector operands into 16-bit signed integers...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set1_epi64x(long long __q)
Initializes both values in a 128-bit integer vector with the specified 64-bit integer value...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_set_sd(double __w)
Constructs a 128-bit floating-point vector of [2 x double].
static __inline__ int __DEFAULT_FN_ATTRS _mm_comile_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ void __DEFAULT_FN_ATTRS _mm_storel_pd(double *__dp, __m128d __a)
Stores the lower 64 bits of a 128-bit vector of [2 x double] to a memory location.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_castps_si128(__m128 __a)
Casts a 128-bit floating-point vector of [4 x float] into a 128-bit integer vector.
void _mm_pause(void)
Indicates that a spin loop is being executed for the purposes of optimizing power consumption during ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_set1_pd(double __w)
Constructs a 128-bit floating-point vector of [2 x double], with each of the two double-precision flo...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_add_epi32(__m128i __a, __m128i __b)
Adds the corresponding elements of two 128-bit vectors of [4 x i32], saving the lower 32 bits of each...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_slli_epi16(__m128i __a, int __count)
Left-shifts each 16-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpgt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ int __DEFAULT_FN_ATTRS _mm_comige_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_loadu_si64(void const *__a)
Loads a 64-bit integer value to the low element of a 128-bit integer vector and clears the upper elem...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sub_epi64(__m128i __a, __m128i __b)
Subtracts the corresponding elements of two [2 x i64] vectors.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_max_sd(__m128d __a, __m128d __b)
Compares lower 64-bit double-precision values of both operands, and returns the greater of the pair o...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srli_epi64(__m128i __a, int __count)
Right-shifts each of 64-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomigt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_unpacklo_pd(__m128d __a, __m128d __b)
Unpacks the low-order 64-bit elements from two 128-bit vectors of [2 x double] and interleaves them i...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_cvtpd_ps(__m128d __a)
Converts the two double-precision floating-point elements of a 128-bit vector of [2 x double] into tw...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sub_epi8(__m128i __a, __m128i __b)
Subtracts the corresponding 8-bit integer values in the operands.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cvtpd_epi32(__m128d __a)
Converts the two double-precision floating-point elements of a 128-bit vector of [2 x double] into tw...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_loadl_pd(__m128d __a, double const *__dp)
Loads a double-precision value into the low-order bits of a 128-bit vector of [2 x double]...
static __inline__ int __DEFAULT_FN_ATTRS _mm_extract_epi16(__m128i __a, int __imm)
Extracts 16 bits from a 128-bit integer vector of [8 x i16], using the immediate-value parameter as a...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_subs_epu16(__m128i __a, __m128i __b)
Subtracts corresponding 16-bit unsigned integer values in the input and returns the differences in th...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpackhi_epi32(__m128i __a, __m128i __b)
Unpacks the high-order (index 2,3) values from two 128-bit vectors of [4 x i32] and interleaves them ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpeq_epi16(__m128i __a, __m128i __b)
Compares each of the corresponding 16-bit values of the 128-bit integer vectors for equality...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setr_epi16(short __w0, short __w1, short __w2, short __w3, short __w4, short __w5, short __w6, short __w7)
Constructs a 128-bit integer vector, initialized in reverse order with the specified 16-bit integral ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpngt_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ void __DEFAULT_FN_ATTRS _mm_maskmoveu_si128(__m128i __d, __m128i __n, char *__p)
Moves bytes selected by the mask from the first operand to the specified unaligned memory location...
double __m128d __attribute__((__vector_size__(16)))
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmplt_epi16(__m128i __a, __m128i __b)
Compares each of the corresponding signed 16-bit values of the 128-bit integer vectors to determine i...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set_epi64(__m64 __q1, __m64 __q0)
Initializes both 64-bit values in a 128-bit vector of [2 x i64] with the specified 64-bit integer val...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_castsi128_ps(__m128i __a)
Casts a 128-bit integer vector into a 128-bit floating-point vector of [4 x float].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_loadl_epi64(__m128i const *__p)
Returns a vector of [2 x i64] where the lower element is taken from the lower element of the operand...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomineq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_min_sd(__m128d __a, __m128d __b)
Compares lower 64-bit double-precision values of both operands, and returns the lesser of the pair of...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_add_sd(__m128d __a, __m128d __b)
Adds lower double-precision values in both operands and returns the sum in the lower 64 bits of the r...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set1_epi8(char __b)
Initializes all values in a 128-bit vector of [16 x i8] with the specified 8-bit value.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpge_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_movpi64_epi64(__m64 __a)
Moves the 64-bit operand to a 128-bit integer vector, zeroing the upper bits.
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_movepi64_pi64(__m128i __a)
Returns the lower 64 bits of a 128-bit integer vector as a 64-bit integer.
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 _mm_sll_epi64(__m128i __a, __m128i __count)
Left-shifts each 64-bit value in the 128-bit integer vector operand by the specified number of bits...
void _mm_mfence(void)
Forces strong memory ordering (serialization) between load and store instructions preceding this inst...
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_cvtpd_pi32(__m128d __a)
Converts the two double-precision floating-point elements of a 128-bit vector of [2 x double] into tw...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_load1_pd(double const *__dp)
Loads a double-precision floating-point value from a specified memory location and duplicates it to b...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_castpd_ps(__m128d __a)
Casts a 128-bit floating-point vector of [2 x double] into a 128-bit floating-point vector of [4 x fl...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmplt_epi32(__m128i __a, __m128i __b)
Compares each of the corresponding signed 32-bit values of the 128-bit integer vectors to determine i...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpeq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_loadu_si128(__m128i const *__p)
Moves packed integer values from an unaligned 128-bit memory location to elements in a 128-bit intege...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set_epi64x(long long __q1, long long __q0)
Initializes both 64-bit values in a 128-bit vector of [2 x i64] with the specified 64-bit integer val...
static __inline unsigned char unsigned int unsigned int unsigned int * __p
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_adds_epi16(__m128i __a, __m128i __b)
Adds, with saturation, the corresponding elements of two 128-bit signed [8 x i16] vectors...
static __inline__ void __DEFAULT_FN_ATTRS _mm_store_pd(double *__dp, __m128d __a)
Moves packed double-precision values from a 128-bit vector of [2 x double] to a memory location...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomilt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_packs_epi16(__m128i __a, __m128i __b)
Converts 16-bit signed integers from both 128-bit integer vector operands into 8-bit signed integers...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srai_epi32(__m128i __a, int __count)
Right-shifts each 32-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sub_epi32(__m128i __a, __m128i __b)
Subtracts the corresponding 32-bit integer values in the operands.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_load_sd(double const *__dp)
Loads a 64-bit double-precision value to the low element of a 128-bit integer vector and clears the u...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpgt_epi32(__m128i __a, __m128i __b)
Compares each of the corresponding signed 32-bit values of the 128-bit integer vectors to determine i...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_setr_pd(double __w, double __x)
Constructs a 128-bit floating-point vector of [2 x double], initialized in reverse order with the spe...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cvtsi32_si128(int __a)
Returns a vector of [4 x i32] where the lowest element is the input operand and the remaining element...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnge_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_load_si128(__m128i const *__p)
Moves packed integer values from an aligned 128-bit memory location to elements in a 128-bit integer ...
static __inline__ void __DEFAULT_FN_ATTRS _mm_storer_pd(double *__dp, __m128d __a)
Stores two double-precision values, in reverse order, from a 128-bit vector of [2 x double] to a 16-b...
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__ __m128d __DEFAULT_FN_ATTRS _mm_sub_pd(__m128d __a, __m128d __b)
Subtracts two 128-bit vectors of [2 x double].
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castps_pd(__m128 __a)
Casts a 128-bit floating-point vector of [4 x float] into a 128-bit floating-point vector of [2 x dou...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_div_sd(__m128d __a, __m128d __b)
Divides the lower double-precision value of the first operand by the lower double-precision value of ...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_cvtsd_ss(__m128 __a, __m128d __b)
Converts the lower double-precision floating-point element of a 128-bit vector of [2 x double]...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtps_pd(__m128 __a)
Converts the lower two single-precision floating-point elements of a 128-bit vector of [4 x float] in...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_add_epi64(__m128i __a, __m128i __b)
Adds the corresponding elements of two 128-bit vectors of [2 x i64], saving the lower 64 bits of each...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_set_pd1(double __w)
Constructs a 128-bit floating-point vector of [2 x double], with each of the two double-precision flo...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmplt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmple_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cvttps_epi32(__m128 __a)
Converts a vector of [4 x float] into a vector of [4 x i32], truncating the result when it is inexact...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_set_pd(double __w, double __x)
Constructs a 128-bit floating-point vector of [2 x double] initialized with the specified double-prec...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_slli_epi64(__m128i __a, int __count)
Left-shifts each 64-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ int __DEFAULT_FN_ATTRS _mm_ucomieq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_min_epi16(__m128i __a, __m128i __b)
Compares corresponding elements of two 128-bit signed [8 x i16] vectors, saving the smaller value fro...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_load_pd(double const *__dp)
Loads a 128-bit floating-point vector of [2 x double] from an aligned memory location.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sra_epi16(__m128i __a, __m128i __count)
Right-shifts each 16-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpngt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_add_pd(__m128d __a, __m128d __b)
Adds two 128-bit vectors of [2 x double].
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpord_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnge_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_add_epi8(__m128i __a, __m128i __b)
Adds the corresponding elements of two 128-bit vectors of [16 x i8], saving the lower 8 bits of each ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_adds_epi8(__m128i __a, __m128i __b)
Adds, with saturation, the corresponding elements of two 128-bit signed [16 x i8] vectors...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpeq_epi32(__m128i __a, __m128i __b)
Compares each of the corresponding 32-bit values of the 128-bit integer vectors for equality...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_move_epi64(__m128i __a)
Moves the lower 64 bits of a 128-bit integer vector to a 128-bit integer vector, zeroing the upper bi...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_or_pd(__m128d __a, __m128d __b)
Performs a bitwise OR of two 128-bit vectors of [2 x double].
static __inline__ void __DEFAULT_FN_ATTRS _mm_store_sd(double *__dp, __m128d __a)
Stores the lower 64 bits of a 128-bit vector of [2 x double] to a memory location.
static __inline__ vector float vector float __b
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnle_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpgt_epi8(__m128i __a, __m128i __b)
Compares each of the corresponding signed 8-bit values of the 128-bit integer vectors to determine if...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnlt_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_div_pd(__m128d __a, __m128d __b)
Performs an element-by-element division of two 128-bit vectors of [2 x double].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpacklo_epi64(__m128i __a, __m128i __b)
Unpacks the low-order 64-bit elements from two 128-bit vectors of [2 x i64] and interleaves them into...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_sqrt_sd(__m128d __a, __m128d __b)
Calculates the square root of the lower double-precision value of the second operand and returns it i...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set_epi32(int __i3, int __i2, int __i1, int __i0)
Initializes the 32-bit values in a 128-bit vector of [4 x i32] with the specified 32-bit integer valu...
void _mm_clflush(void const *__p)
The cache line containing __p is flushed and invalidated from all caches in the coherency domain...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpgt_epi16(__m128i __a, __m128i __b)
Compares each of the corresponding signed 16-bit values of the 128-bit integer vectors to determine i...
static __inline unsigned char unsigned int __x
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_add_epi16(__m128i __a, __m128i __b)
Adds the corresponding elements of two 128-bit vectors of [8 x i16], saving the lower 16 bits of each...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_unpackhi_pd(__m128d __a, __m128d __b)
Unpacks the high-order 64-bit elements from two 128-bit vectors of [2 x double] and interleaves them ...
static __inline__ void __DEFAULT_FN_ATTRS _mm_stream_si128(__m128i *__p, __m128i __a)
Stores a 128-bit integer vector to a 128-bit aligned memory location.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_min_epu8(__m128i __a, __m128i __b)
Compares corresponding elements of two 128-bit unsigned [16 x i8] vectors, saving the smaller value f...
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__ double __DEFAULT_FN_ATTRS _mm_cvtsd_f64(__m128d __a)
Returns the low-order element of a 128-bit vector of [2 x double] as a double-precision floating-poin...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mul_sd(__m128d __a, __m128d __b)
Multiplies lower double-precision values in both operands and returns the product in the lower 64 bit...
static __inline__ void __DEFAULT_FN_ATTRS _mm_storel_epi64(__m128i *__p, __m128i __a)
Stores the lower 64 bits of a 128-bit integer vector of [2 x i64] to a memory location.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sra_epi32(__m128i __a, __m128i __count)
Right-shifts each 32-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_sub_sd(__m128d __a, __m128d __b)
Subtracts the lower double-precision value of the second operand from the lower double-precision valu...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_and_pd(__m128d __a, __m128d __b)
Performs a bitwise AND of two 128-bit vectors of [2 x double].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_castpd_si128(__m128d __a)
Casts a 128-bit floating-point vector of [2 x double] into a 128-bit integer vector.
static __inline__ int __DEFAULT_FN_ATTRS _mm_comineq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mullo_epi16(__m128i __a, __m128i __b)
Multiplies the corresponding elements of two signed [8 x i16] vectors, saving the lower 16 bits of ea...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_insert_epi16(__m128i __a, int __b, int __imm)
Constructs a 128-bit integer vector by first making a copy of the 128-bit integer vector parameter...
static __inline__ int __DEFAULT_FN_ATTRS _mm_cvttsd_si32(__m128d __a)
Converts the low-order element of a [2 x double] vector into a 32-bit signed integer value...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_and_si128(__m128i __a, __m128i __b)
Performs a bitwise AND of two 128-bit integer vectors.
static __inline__ int __DEFAULT_FN_ATTRS _mm_cvtsi128_si32(__m128i __a)
Moves the least significant 32 bits of a vector of [4 x i32] to a 32-bit signed integer value...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set_epi8(char __b15, char __b14, char __b13, char __b12, char __b11, char __b10, char __b9, char __b8, char __b7, char __b6, char __b5, char __b4, char __b3, char __b2, char __b1, char __b0)
Initializes the 8-bit values in a 128-bit vector of [16 x i8] with the specified 8-bit integer values...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtpi32_pd(__m64 __a)
Converts the two signed 32-bit integer elements of a 64-bit vector of [2 x i32] into two double-preci...
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_mul_su32(__m64 __a, __m64 __b)
Multiplies 32-bit unsigned integer values contained in the lower bits of the two 64-bit integer vecto...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srl_epi32(__m128i __a, __m128i __count)
Right-shifts each of 32-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpacklo_epi32(__m128i __a, __m128i __b)
Unpacks the low-order (index 0,1) values from two 128-bit vectors of [4 x i32] and interleaves them i...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnlt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_sub_si64(__m64 __a, __m64 __b)
Subtracts signed or unsigned 64-bit integer values and writes the difference to the corresponding bit...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a)
Casts a 128-bit integer vector into a 128-bit floating-point vector of [2 x double].
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_min_pd(__m128d __a, __m128d __b)
Performs element-by-element comparison of the two 128-bit vectors of [2 x double] and returns the vec...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_max_epi16(__m128i __a, __m128i __b)
Compares corresponding elements of two 128-bit signed [8 x i16] vectors, saving the greater value fro...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_max_epu8(__m128i __a, __m128i __b)
Compares corresponding elements of two 128-bit unsigned [16 x i8] vectors, saving the greater value f...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpord_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_xor_pd(__m128d __a, __m128d __b)
Performs a bitwise XOR of two 128-bit vectors of [2 x double].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setr_epi64(__m64 __q0, __m64 __q1)
Constructs a 128-bit integer vector, initialized in reverse order with the specified 64-bit integral ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_adds_epu16(__m128i __a, __m128i __b)
Adds, with saturation, the corresponding elements of two 128-bit unsigned [8 x i16] vectors...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpacklo_epi8(__m128i __a, __m128i __b)
Unpacks the low-order (index 0-7) values from two 128-bit vectors of [16 x i8] and interleaves them i...
static __inline__ void __DEFAULT_FN_ATTRS _mm_storeh_pd(double *__dp, __m128d __a)
Stores the upper 64 bits of a 128-bit vector of [2 x double] to a memory location.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtsi32_sd(__m128d __a, int __b)
Converts a 32-bit signed integer value, in the second parameter, into a double-precision floating-poi...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_setzero_si128(void)
Creates a 128-bit integer vector initialized to zero.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mul_epu32(__m128i __a, __m128i __b)
Multiplies 32-bit unsigned integer values contained in the lower bits of the corresponding elements o...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_andnot_pd(__m128d __a, __m128d __b)
Performs a bitwise AND of two 128-bit vectors of [2 x double], using the one's complement of the valu...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpackhi_epi64(__m128i __a, __m128i __b)
Unpacks the high-order 64-bit elements from two 128-bit vectors of [2 x i64] and interleaves them int...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmple_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_madd_epi16(__m128i __a, __m128i __b)
Multiplies the corresponding elements of two 128-bit signed [8 x i16] vectors, producing eight interm...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set1_epi32(int __i)
Initializes all values in a 128-bit vector of [4 x i32] with the specified 32-bit value...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpacklo_epi16(__m128i __a, __m128i __b)
Unpacks the low-order (index 0-3) values from each of the two 128-bit vectors of [8 x i16] and interl...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_subs_epi16(__m128i __a, __m128i __b)
Subtracts corresponding 16-bit signed integer values in the input and returns the differences in the ...
static __inline__ int __DEFAULT_FN_ATTRS _mm_comieq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mulhi_epi16(__m128i __a, __m128i __b)
Multiplies the corresponding elements of two signed [8 x i16] vectors, saving the upper 16 bits of ea...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_packus_epi16(__m128i __a, __m128i __b)
Converts 16-bit signed integers from both 128-bit integer vector operands into 8-bit unsigned integer...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_subs_epu8(__m128i __a, __m128i __b)
Subtracts corresponding 8-bit unsigned integer values in the input and returns the differences in the...
static __inline__ void __DEFAULT_FN_ATTRS _mm_store_si128(__m128i *__p, __m128i __b)
Stores a 128-bit integer vector to a memory location aligned on a 128-bit boundary.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmplt_epi8(__m128i __a, __m128i __b)
Compares each of the corresponding signed 8-bit values of the 128-bit integer vectors to determine if...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_undefined_si128(void)
Generates a 128-bit vector of [4 x i32] with unspecified content.
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sad_epu8(__m128i __a, __m128i __b)
Computes the absolute differences of corresponding 8-bit integer values in two 128-bit vectors...
static __inline__ int __DEFAULT_FN_ATTRS _mm_comigt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mulhi_epu16(__m128i __a, __m128i __b)
Multiplies the corresponding elements of two unsigned [8 x i16] vectors, saving the upper 16 bits of ...
static __inline__ int __DEFAULT_FN_ATTRS _mm_comilt_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpackhi_epi8(__m128i __a, __m128i __b)
Unpacks the high-order (index 8-15) values from two 128-bit vectors of [16 x i8] and interleaves them...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srl_epi64(__m128i __a, __m128i __count)
Right-shifts each of 64-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ void __DEFAULT_FN_ATTRS _mm_stream_pd(double *__p, __m128d __a)
Stores a 128-bit floating point vector of [2 x double] to a 128-bit aligned memory location...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_set1_epi16(short __w)
Initializes all values in a 128-bit vector of [8 x i16] with the specified 16-bit value...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_move_sd(__m128d __a, __m128d __b)
Constructs a 128-bit floating-point vector of [2 x double].
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpunord_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpge_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpneq_sd(__m128d __a, __m128d __b)
Compares the lower double-precision floating-point values in each of the two 128-bit floating-point v...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_avg_epu16(__m128i __a, __m128i __b)
Computes the rounded avarages of corresponding elements of two 128-bit unsigned [8 x i16] vectors...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmplt_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_srl_epi16(__m128i __a, __m128i __count)
Right-shifts each of 16-bit values in the 128-bit integer vector operand by the specified number of b...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sll_epi16(__m128i __a, __m128i __count)
Left-shifts each 16-bit value in the 128-bit integer vector operand by the specified number of bits...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_loadh_pd(__m128d __a, double const *__dp)
Loads a double-precision value into the high-order bits of a 128-bit vector of [2 x double]...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpneq_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
#define __DEFAULT_FN_ATTRS
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cvtps_epi32(__m128 __a)
Converts a vector of [4 x float] into a vector of [4 x i32].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sub_epi16(__m128i __a, __m128i __b)
Subtracts the corresponding 16-bit integer values in the operands.
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpeq_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] for...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_max_pd(__m128d __a, __m128d __b)
Performs element-by-element comparison of the two 128-bit vectors of [2 x double] and returns the vec...
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_add_si64(__m64 __a, __m64 __b)
Adds two signed or unsigned 64-bit integer values, returning the lower 64 bits of the sum...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cvtss_sd(__m128d __a, __m128 __b)
Converts the lower single-precision floating-point element of a 128-bit vector of [4 x float]...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_xor_si128(__m128i __a, __m128i __b)
Performs a bitwise exclusive OR of two 128-bit integer vectors.
static __inline__ int __DEFAULT_FN_ATTRS _mm_movemask_pd(__m128d __a)
Extracts the sign bits of the double-precision values in the 128-bit vector of [2 x double]...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_subs_epi8(__m128i __a, __m128i __b)
Subtracts corresponding 8-bit signed integer values in the input and returns the differences in the c...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_or_si128(__m128i __a, __m128i __b)
Performs a bitwise OR of two 128-bit integer vectors.
static __inline__ void __DEFAULT_FN_ATTRS _mm_storeu_si128(__m128i *__p, __m128i __b)
Stores a 128-bit integer vector to an unaligned memory location.
static __inline__ void __DEFAULT_FN_ATTRS _mm_stream_si32(int *__p, int __a)
Stores a 32-bit integer value in the specified memory location.
static __inline__ int __DEFAULT_FN_ATTRS _mm_movemask_epi8(__m128i __a)
Copies the values of the most significant bits from each 8-bit element in a 128-bit integer vector of...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_mul_pd(__m128d __a, __m128d __b)
Multiplies two 128-bit vectors of [2 x double].
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_loadr_pd(double const *__dp)
Loads two double-precision values, in reverse order, from an aligned memory location into a 128-bit v...
static __inline__ void __DEFAULT_FN_ATTRS _mm_store1_pd(double *__dp, __m128d __a)
Moves the lower 64 bits of a 128-bit vector of [2 x double] twice to the upper and lower 64 bits of a...
static __inline__ __m64 __DEFAULT_FN_ATTRS _mm_cvttpd_pi32(__m128d __a)
Converts the two double-precision floating-point elements of a 128-bit vector of [2 x double] into tw...
static __inline__ __m128 __DEFAULT_FN_ATTRS _mm_cvtepi32_ps(__m128i __a)
Converts a vector of [4 x i32] into a vector of [4 x float].
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_avg_epu8(__m128i __a, __m128i __b)
Computes the rounded avarages of corresponding elements of two 128-bit unsigned [16 x i8] vectors...
static __inline__ vector float vector float vector float __c
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_adds_epu8(__m128i __a, __m128i __b)
Adds, with saturation, the corresponding elements of two 128-bit unsigned [16 x i8] vectors...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_undefined_pd(void)
Constructs a 128-bit floating-point vector of [2 x double] with unspecified content.
void _mm_lfence(void)
Forces strong memory ordering (serialization) between load instructions preceding this instruction an...
static __inline__ void __DEFAULT_FN_ATTRS _mm_store_pd1(double *__dp, __m128d __a)
Moves the lower 64 bits of a 128-bit vector of [2 x double] twice to the upper and lower 64 bits of a...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpnle_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_andnot_si128(__m128i __a, __m128i __b)
Performs a bitwise AND of two 128-bit integer vectors, using the one's complement of the values conta...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_cmpeq_epi8(__m128i __a, __m128i __b)
Compares each of the corresponding 8-bit values of the 128-bit integer vectors for equality...
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_unpackhi_epi16(__m128i __a, __m128i __b)
Unpacks the high-order (index 4-7) values from two 128-bit vectors of [8 x i16] and interleaves them ...
static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_cmpunord_pd(__m128d __a, __m128d __b)
Compares each of the corresponding double-precision values of the 128-bit vectors of [2 x double] to ...