-/* Copyright (C) 2013-2014 Free Software Foundation, Inc.
+/* Copyright (C) 2013-2020 Free Software Foundation, Inc.
This file is part of GCC.
typedef double __v8df __attribute__ ((__vector_size__ (64)));
typedef float __v16sf __attribute__ ((__vector_size__ (64)));
typedef long long __v8di __attribute__ ((__vector_size__ (64)));
+typedef unsigned long long __v8du __attribute__ ((__vector_size__ (64)));
typedef int __v16si __attribute__ ((__vector_size__ (64)));
+typedef unsigned int __v16su __attribute__ ((__vector_size__ (64)));
+typedef short __v32hi __attribute__ ((__vector_size__ (64)));
+typedef unsigned short __v32hu __attribute__ ((__vector_size__ (64)));
+typedef char __v64qi __attribute__ ((__vector_size__ (64)));
+typedef unsigned char __v64qu __attribute__ ((__vector_size__ (64)));
/* The Intel API is flexible enough that we must allow aliasing with other
vector types, and their scalar components. */
typedef long long __m512i __attribute__ ((__vector_size__ (64), __may_alias__));
typedef double __m512d __attribute__ ((__vector_size__ (64), __may_alias__));
+/* Unaligned version of the same type. */
+typedef float __m512_u __attribute__ ((__vector_size__ (64), __may_alias__, __aligned__ (1)));
+typedef long long __m512i_u __attribute__ ((__vector_size__ (64), __may_alias__, __aligned__ (1)));
+typedef double __m512d_u __attribute__ ((__vector_size__ (64), __may_alias__, __aligned__ (1)));
+
typedef unsigned char __mmask8;
typedef unsigned short __mmask16;
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_int2mask (int __M)
+{
+ return (__mmask16) __M;
+}
+
+extern __inline int
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask2int (__mmask16 __M)
+{
+ return (int) __M;
+}
+
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_set_epi64 (long long __A, long long __B, long long __C,
__H, __G, __F, __E, __D, __C, __B, __A };
}
+extern __inline __m512i
+__attribute__((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set_epi16 (short __q31, short __q30, short __q29, short __q28,
+ short __q27, short __q26, short __q25, short __q24,
+ short __q23, short __q22, short __q21, short __q20,
+ short __q19, short __q18, short __q17, short __q16,
+ short __q15, short __q14, short __q13, short __q12,
+ short __q11, short __q10, short __q09, short __q08,
+ short __q07, short __q06, short __q05, short __q04,
+ short __q03, short __q02, short __q01, short __q00)
+{
+ return __extension__ (__m512i)(__v32hi){
+ __q00, __q01, __q02, __q03, __q04, __q05, __q06, __q07,
+ __q08, __q09, __q10, __q11, __q12, __q13, __q14, __q15,
+ __q16, __q17, __q18, __q19, __q20, __q21, __q22, __q23,
+ __q24, __q25, __q26, __q27, __q28, __q29, __q30, __q31
+ };
+}
+
+extern __inline __m512i
+__attribute__((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set_epi8 (char __q63, char __q62, char __q61, char __q60,
+ char __q59, char __q58, char __q57, char __q56,
+ char __q55, char __q54, char __q53, char __q52,
+ char __q51, char __q50, char __q49, char __q48,
+ char __q47, char __q46, char __q45, char __q44,
+ char __q43, char __q42, char __q41, char __q40,
+ char __q39, char __q38, char __q37, char __q36,
+ char __q35, char __q34, char __q33, char __q32,
+ char __q31, char __q30, char __q29, char __q28,
+ char __q27, char __q26, char __q25, char __q24,
+ char __q23, char __q22, char __q21, char __q20,
+ char __q19, char __q18, char __q17, char __q16,
+ char __q15, char __q14, char __q13, char __q12,
+ char __q11, char __q10, char __q09, char __q08,
+ char __q07, char __q06, char __q05, char __q04,
+ char __q03, char __q02, char __q01, char __q00)
+{
+ return __extension__ (__m512i)(__v64qi){
+ __q00, __q01, __q02, __q03, __q04, __q05, __q06, __q07,
+ __q08, __q09, __q10, __q11, __q12, __q13, __q14, __q15,
+ __q16, __q17, __q18, __q19, __q20, __q21, __q22, __q23,
+ __q24, __q25, __q26, __q27, __q28, __q29, __q30, __q31,
+ __q32, __q33, __q34, __q35, __q36, __q37, __q38, __q39,
+ __q40, __q41, __q42, __q43, __q44, __q45, __q46, __q47,
+ __q48, __q49, __q50, __q51, __q52, __q53, __q54, __q55,
+ __q56, __q57, __q58, __q59, __q60, __q61, __q62, __q63
+ };
+}
+
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_set_pd (double __A, double __B, double __C, double __D,
return __Y;
}
+#define _mm512_undefined _mm512_undefined_ps
+
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_undefined_pd (void)
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_undefined_si512 (void)
+_mm512_undefined_epi32 (void)
{
__m512i __Y = __Y;
return __Y;
}
+#define _mm512_undefined_si512 _mm512_undefined_epi32
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set1_epi8 (char __A)
+{
+ return __extension__ (__m512i)(__v64qi)
+ { __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A };
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set1_epi16 (short __A)
+{
+ return __extension__ (__m512i)(__v32hi)
+ { __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A,
+ __A, __A, __A, __A, __A, __A, __A, __A };
+}
+
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_set1_pd (double __A)
(__mmask16) -1);
}
+/* Create the vector [A B C D A B C D A B C D A B C D]. */
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set4_epi32 (int __A, int __B, int __C, int __D)
+{
+ return __extension__ (__m512i)(__v16si)
+ { __D, __C, __B, __A, __D, __C, __B, __A,
+ __D, __C, __B, __A, __D, __C, __B, __A };
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set4_epi64 (long long __A, long long __B, long long __C,
+ long long __D)
+{
+ return __extension__ (__m512i) (__v8di)
+ { __D, __C, __B, __A, __D, __C, __B, __A };
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set4_pd (double __A, double __B, double __C, double __D)
+{
+ return __extension__ (__m512d)
+ { __D, __C, __B, __A, __D, __C, __B, __A };
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_set4_ps (float __A, float __B, float __C, float __D)
+{
+ return __extension__ (__m512)
+ { __D, __C, __B, __A, __D, __C, __B, __A,
+ __D, __C, __B, __A, __D, __C, __B, __A };
+}
+
+#define _mm512_setr4_epi64(e0,e1,e2,e3) \
+ _mm512_set4_epi64(e3,e2,e1,e0)
+
+#define _mm512_setr4_epi32(e0,e1,e2,e3) \
+ _mm512_set4_epi32(e3,e2,e1,e0)
+
+#define _mm512_setr4_pd(e0,e1,e2,e3) \
+ _mm512_set4_pd(e3,e2,e1,e0)
+
+#define _mm512_setr4_ps(e0,e1,e2,e3) \
+ _mm512_set4_ps(e3,e2,e1,e0)
+
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_setzero_ps (void)
0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 };
}
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_setzero (void)
+{
+ return _mm512_setzero_ps ();
+}
+
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_setzero_pd (void)
return __extension__ (__m512d) { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 };
}
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_setzero_epi32 (void)
+{
+ return __extension__ (__m512i)(__v8di){ 0, 0, 0, 0, 0, 0, 0, 0 };
+}
+
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_setzero_si512 (void)
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mullo_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pmulld512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A * (__v16su) __B);
}
extern __inline __m512i
(__v16si) __W, __M);
}
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mullox_epi64 (__m512i __A, __m512i __B)
+{
+ return (__m512i) ((__v8du) __A * (__v8du) __B);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_mullox_epi64 (__m512i __W, __mmask8 __M, __m512i __A, __m512i __B)
+{
+ return _mm512_mask_mov_epi64 (__W, __M, _mm512_mullox_epi64 (__A, __B));
+}
+
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_sllv_epi32 (__m512i __X, __m512i __Y)
return (__m512i) __builtin_ia32_psllv16si_mask ((__v16si) __X,
(__v16si) __Y,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_psrav16si_mask ((__v16si) __X,
(__v16si) __Y,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_psrlv16si_mask ((__v16si) __X,
(__v16si) __Y,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_add_epi64 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_paddq512_mask ((__v8di) __A,
- (__v8di) __B,
- (__v8di)
- _mm512_undefined_si512 (),
- (__mmask8) -1);
+ return (__m512i) ((__v8du) __A + (__v8du) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_sub_epi64 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_psubq512_mask ((__v8di) __A,
- (__v8di) __B,
- (__v8di)
- _mm512_undefined_pd (),
- (__mmask8) -1);
+ return (__m512i) ((__v8du) __A - (__v8du) __B);
}
extern __inline __m512i
return (__m512i) __builtin_ia32_psrav8di_mask ((__v8di) __X,
(__v8di) __Y,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_psrlv8di_mask ((__v8di) __X,
(__v8di) __Y,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_add_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_paddd512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A + (__v16su) __B);
}
extern __inline __m512i
return (__m512i) __builtin_ia32_pmuldq512_mask ((__v16si) __X,
(__v16si) __Y,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_sub_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_psubd512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A - (__v16su) __B);
}
extern __inline __m512i
return (__m512i) __builtin_ia32_pmuludq512_mask ((__v16si) __X,
(__v16si) __Y,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_psllqi512_mask ((__v8di) __A, __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#else
#define _mm512_slli_epi64(X, C) \
((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X), (int)(C),\
- (__v8di)(__m512i)_mm512_undefined_si512 (),\
+ (__v8di)(__m512i)_mm512_undefined_epi32 (),\
(__mmask8)-1))
#define _mm512_mask_slli_epi64(W, U, X, C) \
return (__m512i) __builtin_ia32_psllq512_mask ((__v8di) __A,
(__v2di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_psrlqi512_mask ((__v8di) __A, __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#else
#define _mm512_srli_epi64(X, C) \
((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X), (int)(C),\
- (__v8di)(__m512i)_mm512_undefined_si512 (),\
+ (__v8di)(__m512i)_mm512_undefined_epi32 (),\
(__mmask8)-1))
#define _mm512_mask_srli_epi64(W, U, X, C) \
return (__m512i) __builtin_ia32_psrlq512_mask ((__v8di) __A,
(__v2di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_psraqi512_mask ((__v8di) __A, __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#else
#define _mm512_srai_epi64(X, C) \
((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X), (int)(C),\
- (__v8di)(__m512i)_mm512_undefined_si512 (),\
+ (__v8di)(__m512i)_mm512_undefined_epi32 (),\
(__mmask8)-1))
#define _mm512_mask_srai_epi64(W, U, X, C) \
return (__m512i) __builtin_ia32_psraq512_mask ((__v8di) __A,
(__v2di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pslldi512_mask ((__v16si) __A, __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
#else
#define _mm512_slli_epi32(X, C) \
((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X), (int)(C),\
- (__v16si)(__m512i)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_slli_epi32(W, U, X, C) \
return (__m512i) __builtin_ia32_pslld512_mask ((__v16si) __A,
(__v4si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_psrldi512_mask ((__v16si) __A, __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
#else
#define _mm512_srli_epi32(X, C) \
((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X), (int)(C),\
- (__v16si)(__m512i)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_srli_epi32(W, U, X, C) \
return (__m512i) __builtin_ia32_psrld512_mask ((__v16si) __A,
(__v4si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_psradi512_mask ((__v16si) __A, __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
#else
#define _mm512_srai_epi32(X, C) \
((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X), (int)(C),\
- (__v16si)(__m512i)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_srai_epi32(W, U, X, C) \
return (__m512i) __builtin_ia32_psrad512_mask ((__v16si) __A,
(__v4si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_add_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_addsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_add_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_addsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_add_round_ss (__m128 __A, __m128 __B, const int __R)
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_add_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_addss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_add_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_addss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_sub_round_sd (__m128d __A, __m128d __B, const int __R)
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_sub_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_subsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_sub_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_subsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_sub_round_ss (__m128 __A, __m128 __B, const int __R)
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_sub_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_subss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_sub_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_subss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
#else
#define _mm_add_round_sd(A, B, C) \
(__m128d)__builtin_ia32_addsd_round(A, B, C)
+#define _mm_mask_add_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_addsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_add_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_addsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
+
#define _mm_add_round_ss(A, B, C) \
(__m128)__builtin_ia32_addss_round(A, B, C)
+#define _mm_mask_add_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_addss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_add_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_addss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
+
#define _mm_sub_round_sd(A, B, C) \
(__m128d)__builtin_ia32_subsd_round(A, B, C)
+#define _mm_mask_sub_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_subsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_sub_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_subsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
+
#define _mm_sub_round_ss(A, B, C) \
(__m128)__builtin_ia32_subss_round(A, B, C)
+
+#define _mm_mask_sub_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_subss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_sub_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_subss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
+
#endif
#ifdef __OPTIMIZE__
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_ternarylogic_epi64 (__m512i __A, __m512i __B, __m512i __C, const int imm)
+_mm512_ternarylogic_epi64 (__m512i __A, __m512i __B, __m512i __C,
+ const int __imm)
{
return (__m512i) __builtin_ia32_pternlogq512_mask ((__v8di) __A,
(__v8di) __B,
- (__v8di) __C, imm,
+ (__v8di) __C, __imm,
(__mmask8) -1);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_ternarylogic_epi64 (__m512i __A, __mmask8 __U, __m512i __B,
- __m512i __C, const int imm)
+ __m512i __C, const int __imm)
{
return (__m512i) __builtin_ia32_pternlogq512_mask ((__v8di) __A,
(__v8di) __B,
- (__v8di) __C, imm,
+ (__v8di) __C, __imm,
(__mmask8) __U);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_maskz_ternarylogic_epi64 (__mmask8 __U, __m512i __A, __m512i __B,
- __m512i __C, const int imm)
+ __m512i __C, const int __imm)
{
return (__m512i) __builtin_ia32_pternlogq512_maskz ((__v8di) __A,
(__v8di) __B,
(__v8di) __C,
- imm, (__mmask8) __U);
+ __imm, (__mmask8) __U);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_ternarylogic_epi32 (__m512i __A, __m512i __B, __m512i __C, const int imm)
+_mm512_ternarylogic_epi32 (__m512i __A, __m512i __B, __m512i __C,
+ const int __imm)
{
return (__m512i) __builtin_ia32_pternlogd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si) __C,
- imm, (__mmask16) -1);
+ __imm, (__mmask16) -1);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_ternarylogic_epi32 (__m512i __A, __mmask16 __U, __m512i __B,
- __m512i __C, const int imm)
+ __m512i __C, const int __imm)
{
return (__m512i) __builtin_ia32_pternlogd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si) __C,
- imm, (__mmask16) __U);
+ __imm, (__mmask16) __U);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_maskz_ternarylogic_epi32 (__mmask16 __U, __m512i __A, __m512i __B,
- __m512i __C, const int imm)
+ __m512i __C, const int __imm)
{
return (__m512i) __builtin_ia32_pternlogd512_maskz ((__v16si) __A,
(__v16si) __B,
(__v16si) __C,
- imm, (__mmask16) __U);
+ __imm, (__mmask16) __U);
}
#else
#define _mm512_ternarylogic_epi64(A, B, C, I) \
(__v2df) __A);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_rcp14_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_rcp14sd_mask ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_rcp14_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_rcp14sd_mask ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df) _mm_setzero_ps (),
+ (__mmask8) __U);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_rcp14_ss (__m128 __A, __m128 __B)
(__v4sf) __A);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_rcp14_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_rcp14ss_mask ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_rcp14_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_rcp14ss_mask ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf) _mm_setzero_ps (),
+ (__mmask8) __U);
+}
+
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_rsqrt14_pd (__m512d __A)
(__v2df) __A);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_rsqrt14_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_rsqrt14sd_mask ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_rsqrt14_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_rsqrt14sd_mask ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df) _mm_setzero_pd (),
+ (__mmask8) __U);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_rsqrt14_ss (__m128 __A, __m128 __B)
(__v4sf) __A);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_rsqrt14_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_rsqrt14ss_mask ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_rsqrt14_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_rsqrt14ss_mask ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf) _mm_setzero_ps (),
+ (__mmask8) __U);
+}
+
#ifdef __OPTIMIZE__
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_sqrt_round_sd (__m128d __A, __m128d __B, const int __R)
{
- return (__m128d) __builtin_ia32_sqrtsd_round ((__v2df) __B,
- (__v2df) __A,
- __R);
+ return (__m128d) __builtin_ia32_sqrtsd_mask_round ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) -1, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_sqrt_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_sqrtsd_mask_round ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_sqrt_round_sd (__mmask8 __U, __m128d __A, __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_sqrtsd_mask_round ((__v2df) __B,
+ (__v2df) __A,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
}
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_sqrt_round_ss (__m128 __A, __m128 __B, const int __R)
{
- return (__m128) __builtin_ia32_sqrtss_round ((__v4sf) __B,
- (__v4sf) __A,
- __R);
+ return (__m128) __builtin_ia32_sqrtss_mask_round ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) -1, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_sqrt_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_sqrtss_mask_round ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_sqrt_round_ss (__mmask8 __U, __m128 __A, __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_sqrtss_mask_round ((__v4sf) __B,
+ (__v4sf) __A,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
}
#else
#define _mm512_sqrt_round_pd(A, C) \
#define _mm512_maskz_sqrt_round_ps(U, A, C) \
(__m512)__builtin_ia32_sqrtps512_mask(A, (__v16sf)_mm512_setzero_ps(), U, C)
-#define _mm_sqrt_round_sd(A, B, C) \
- (__m128d)__builtin_ia32_sqrtsd_round(A, B, C)
+#define _mm_sqrt_round_sd(A, B, C) \
+ (__m128d)__builtin_ia32_sqrtsd_mask_round (B, A, \
+ (__v2df) _mm_setzero_pd (), -1, C)
+
+#define _mm_mask_sqrt_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_sqrtsd_mask_round (B, A, W, U, C)
+
+#define _mm_maskz_sqrt_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_sqrtsd_mask_round (B, A, \
+ (__v2df) _mm_setzero_pd (), U, C)
-#define _mm_sqrt_round_ss(A, B, C) \
- (__m128)__builtin_ia32_sqrtss_round(A, B, C)
+#define _mm_sqrt_round_ss(A, B, C) \
+ (__m128)__builtin_ia32_sqrtss_mask_round (B, A, \
+ (__v4sf) _mm_setzero_ps (), -1, C)
+
+#define _mm_mask_sqrt_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_sqrtss_mask_round (B, A, W, U, C)
+
+#define _mm_maskz_sqrt_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_sqrtss_mask_round (B, A, \
+ (__v4sf) _mm_setzero_ps (), U, C)
#endif
extern __inline __m512i
{
return (__m512i) __builtin_ia32_pmovsxbd512_mask ((__v16qi) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pmovsxbq512_mask ((__v16qi) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pmovsxwd512_mask ((__v16hi) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pmovsxwq512_mask ((__v8hi) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pmovsxdq512_mask ((__v8si) __X,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pmovzxbd512_mask ((__v16qi) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pmovzxbq512_mask ((__v16qi) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pmovzxwd512_mask ((__v16hi) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pmovzxwq512_mask ((__v8hi) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pmovzxdq512_mask ((__v8si) __X,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
__R);
}
-extern __inline __m128
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_mul_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_mulsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_mul_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_mulsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_mul_round_ss (__m128 __A, __m128 __B, const int __R)
{
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_mul_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_mulss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_mul_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_mulss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_div_round_sd (__m128d __A, __m128d __B, const int __R)
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_div_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_divsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_div_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_divsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_div_round_ss (__m128 __A, __m128 __B, const int __R)
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_div_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_divss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_div_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_divss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
#else
#define _mm512_mul_round_pd(A, B, C) \
(__m512d)__builtin_ia32_mulpd512_mask(A, B, (__v8df)_mm512_undefined_pd(), -1, C)
#define _mm_mul_round_sd(A, B, C) \
(__m128d)__builtin_ia32_mulsd_round(A, B, C)
+#define _mm_mask_mul_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_mulsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_mul_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_mulsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
+
#define _mm_mul_round_ss(A, B, C) \
(__m128)__builtin_ia32_mulss_round(A, B, C)
+#define _mm_mask_mul_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_mulss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_mul_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_mulss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
+
#define _mm_div_round_sd(A, B, C) \
(__m128d)__builtin_ia32_divsd_round(A, B, C)
+#define _mm_mask_div_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_divsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_div_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_divsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
+
#define _mm_div_round_ss(A, B, C) \
(__m128)__builtin_ia32_divss_round(A, B, C)
+
+#define _mm_mask_div_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_divss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_div_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_divss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
+
#endif
#ifdef __OPTIMIZE__
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_scalef_round_sd (__m128d __A, __m128d __B, const int __R)
{
- return (__m128d) __builtin_ia32_scalefsd_round ((__v2df) __A,
- (__v2df) __B,
- __R);
+ return (__m128d) __builtin_ia32_scalefsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) -1, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_scalef_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_scalefsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_scalef_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_scalefsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
}
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_scalef_round_ss (__m128 __A, __m128 __B, const int __R)
{
- return (__m128) __builtin_ia32_scalefss_round ((__v4sf) __A,
- (__v4sf) __B,
- __R);
+ return (__m128) __builtin_ia32_scalefss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) -1, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_scalef_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_scalefss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_scalef_round_ss (__mmask8 __U, __m128 __A, __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_scalefss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
}
#else
#define _mm512_scalef_round_pd(A, B, C) \
(__m512)__builtin_ia32_scalefps512_mask(A, B, (__v16sf)_mm512_setzero_ps(), U, C)
#define _mm_scalef_round_sd(A, B, C) \
- (__m128d)__builtin_ia32_scalefsd_round(A, B, C)
+ (__m128d)__builtin_ia32_scalefsd_mask_round (A, B, \
+ (__v2df)_mm_setzero_pd (), -1, C)
#define _mm_scalef_round_ss(A, B, C) \
- (__m128)__builtin_ia32_scalefss_round(A, B, C)
+ (__m128)__builtin_ia32_scalefss_mask_round (A, B, \
+ (__v4sf)_mm_setzero_ps (), -1, C)
#endif
#ifdef __OPTIMIZE__
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fmsub_round_pd (__m512d __A, __m512d __B, __m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
+ return (__m512d) __builtin_ia32_vfmsubpd512_mask ((__v8df) __A,
(__v8df) __B,
- -(__v8df) __C,
+ (__v8df) __C,
(__mmask8) -1, __R);
}
_mm512_mask_fmsub_round_pd (__m512d __A, __mmask8 __U, __m512d __B,
__m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
+ return (__m512d) __builtin_ia32_vfmsubpd512_mask ((__v8df) __A,
(__v8df) __B,
- -(__v8df) __C,
+ (__v8df) __C,
(__mmask8) __U, __R);
}
_mm512_maskz_fmsub_round_pd (__mmask8 __U, __m512d __A, __m512d __B,
__m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
+ return (__m512d) __builtin_ia32_vfmsubpd512_maskz ((__v8df) __A,
(__v8df) __B,
- -(__v8df) __C,
+ (__v8df) __C,
(__mmask8) __U, __R);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fmsub_round_ps (__m512 __A, __m512 __B, __m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
+ return (__m512) __builtin_ia32_vfmsubps512_mask ((__v16sf) __A,
(__v16sf) __B,
- -(__v16sf) __C,
+ (__v16sf) __C,
(__mmask16) -1, __R);
}
_mm512_mask_fmsub_round_ps (__m512 __A, __mmask16 __U, __m512 __B,
__m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
+ return (__m512) __builtin_ia32_vfmsubps512_mask ((__v16sf) __A,
(__v16sf) __B,
- -(__v16sf) __C,
+ (__v16sf) __C,
(__mmask16) __U, __R);
}
_mm512_maskz_fmsub_round_ps (__mmask16 __U, __m512 __A, __m512 __B,
__m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
+ return (__m512) __builtin_ia32_vfmsubps512_maskz ((__v16sf) __A,
(__v16sf) __B,
- -(__v16sf) __C,
+ (__v16sf) __C,
(__mmask16) __U, __R);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fnmadd_round_pd (__m512d __A, __m512d __B, __m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) -1, __R);
+ return (__m512d) __builtin_ia32_vfnmaddpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1, __R);
}
extern __inline __m512d
_mm512_mask3_fnmadd_round_pd (__m512d __A, __m512d __B, __m512d __C,
__mmask8 __U, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask3 (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U, __R);
+ return (__m512d) __builtin_ia32_vfnmaddpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U, __R);
}
extern __inline __m512d
_mm512_maskz_fnmadd_round_pd (__mmask8 __U, __m512d __A, __m512d __B,
__m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U, __R);
+ return (__m512d) __builtin_ia32_vfnmaddpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U, __R);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fnmadd_round_ps (__m512 __A, __m512 __B, __m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) -1, __R);
+ return (__m512) __builtin_ia32_vfnmaddps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1, __R);
}
extern __inline __m512
_mm512_mask3_fnmadd_round_ps (__m512 __A, __m512 __B, __m512 __C,
__mmask16 __U, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask3 (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U, __R);
+ return (__m512) __builtin_ia32_vfnmaddps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U, __R);
}
extern __inline __m512
_mm512_maskz_fnmadd_round_ps (__mmask16 __U, __m512 __A, __m512 __B,
__m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U, __R);
+ return (__m512) __builtin_ia32_vfnmaddps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U, __R);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fnmsub_round_pd (__m512d __A, __m512d __B, __m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) -1, __R);
+ return (__m512d) __builtin_ia32_vfnmsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1, __R);
}
extern __inline __m512d
_mm512_maskz_fnmsub_round_pd (__mmask8 __U, __m512d __A, __m512d __B,
__m512d __C, const int __R)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U, __R);
+ return (__m512d) __builtin_ia32_vfnmsubpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U, __R);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_fnmsub_round_ps (__m512 __A, __m512 __B, __m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) -1, __R);
+ return (__m512) __builtin_ia32_vfnmsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1, __R);
}
extern __inline __m512
_mm512_maskz_fnmsub_round_ps (__mmask16 __U, __m512 __A, __m512 __B,
__m512 __C, const int __R)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U, __R);
+ return (__m512) __builtin_ia32_vfnmsubps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U, __R);
}
#else
#define _mm512_fmadd_round_pd(A, B, C, R) \
(__m512)__builtin_ia32_vfmaddps512_maskz(A, B, C, U, R)
#define _mm512_fmsub_round_pd(A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask(A, B, -(C), -1, R)
+ (__m512d)__builtin_ia32_vfmsubpd512_mask(A, B, C, -1, R)
#define _mm512_mask_fmsub_round_pd(A, U, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask(A, B, -(C), U, R)
+ (__m512d)__builtin_ia32_vfmsubpd512_mask(A, B, C, U, R)
#define _mm512_mask3_fmsub_round_pd(A, B, C, U, R) \
(__m512d)__builtin_ia32_vfmsubpd512_mask3(A, B, C, U, R)
#define _mm512_maskz_fmsub_round_pd(U, A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_maskz(A, B, -(C), U, R)
+ (__m512d)__builtin_ia32_vfmsubpd512_maskz(A, B, C, U, R)
#define _mm512_fmsub_round_ps(A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_mask(A, B, -(C), -1, R)
+ (__m512)__builtin_ia32_vfmsubps512_mask(A, B, C, -1, R)
#define _mm512_mask_fmsub_round_ps(A, U, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_mask(A, B, -(C), U, R)
+ (__m512)__builtin_ia32_vfmsubps512_mask(A, B, C, U, R)
#define _mm512_mask3_fmsub_round_ps(A, B, C, U, R) \
(__m512)__builtin_ia32_vfmsubps512_mask3(A, B, C, U, R)
#define _mm512_maskz_fmsub_round_ps(U, A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_maskz(A, B, -(C), U, R)
+ (__m512)__builtin_ia32_vfmsubps512_maskz(A, B, C, U, R)
#define _mm512_fmaddsub_round_pd(A, B, C, R) \
(__m512d)__builtin_ia32_vfmaddsubpd512_mask(A, B, C, -1, R)
#define _mm512_mask_fmaddsub_round_pd(A, U, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask(A, B, C, U, R)
+ (__m512d)__builtin_ia32_vfmaddsubpd512_mask(A, B, C, U, R)
#define _mm512_mask3_fmaddsub_round_pd(A, B, C, U, R) \
(__m512d)__builtin_ia32_vfmaddsubpd512_mask3(A, B, C, U, R)
(__m512)__builtin_ia32_vfmaddsubps512_maskz(A, B, -(C), U, R)
#define _mm512_fnmadd_round_pd(A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask(-(A), B, C, -1, R)
+ (__m512d)__builtin_ia32_vfnmaddpd512_mask(A, B, C, -1, R)
#define _mm512_mask_fnmadd_round_pd(A, U, B, C, R) \
- (__m512d)__builtin_ia32_vfnmaddpd512_mask(-(A), B, C, U, R)
+ (__m512d)__builtin_ia32_vfnmaddpd512_mask(A, B, C, U, R)
#define _mm512_mask3_fnmadd_round_pd(A, B, C, U, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask3(-(A), B, C, U, R)
+ (__m512d)__builtin_ia32_vfnmaddpd512_mask3(A, B, C, U, R)
#define _mm512_maskz_fnmadd_round_pd(U, A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(A), B, C, U, R)
+ (__m512d)__builtin_ia32_vfnmaddpd512_maskz(A, B, C, U, R)
#define _mm512_fnmadd_round_ps(A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_mask(-(A), B, C, -1, R)
+ (__m512)__builtin_ia32_vfnmaddps512_mask(A, B, C, -1, R)
#define _mm512_mask_fnmadd_round_ps(A, U, B, C, R) \
- (__m512)__builtin_ia32_vfnmaddps512_mask(-(A), B, C, U, R)
+ (__m512)__builtin_ia32_vfnmaddps512_mask(A, B, C, U, R)
#define _mm512_mask3_fnmadd_round_ps(A, B, C, U, R) \
- (__m512)__builtin_ia32_vfmaddps512_mask3(-(A), B, C, U, R)
+ (__m512)__builtin_ia32_vfnmaddps512_mask3(A, B, C, U, R)
#define _mm512_maskz_fnmadd_round_ps(U, A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_maskz(-(A), B, C, U, R)
+ (__m512)__builtin_ia32_vfnmaddps512_maskz(A, B, C, U, R)
#define _mm512_fnmsub_round_pd(A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_mask(-(A), B, -(C), -1, R)
+ (__m512d)__builtin_ia32_vfnmsubpd512_mask(A, B, C, -1, R)
#define _mm512_mask_fnmsub_round_pd(A, U, B, C, R) \
(__m512d)__builtin_ia32_vfnmsubpd512_mask(A, B, C, U, R)
(__m512d)__builtin_ia32_vfnmsubpd512_mask3(A, B, C, U, R)
#define _mm512_maskz_fnmsub_round_pd(U, A, B, C, R) \
- (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(A), B, -(C), U, R)
+ (__m512d)__builtin_ia32_vfnmsubpd512_maskz(A, B, C, U, R)
#define _mm512_fnmsub_round_ps(A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_mask(-(A), B, -(C), -1, R)
+ (__m512)__builtin_ia32_vfnmsubps512_mask(A, B, C, -1, R)
#define _mm512_mask_fnmsub_round_ps(A, U, B, C, R) \
(__m512)__builtin_ia32_vfnmsubps512_mask(A, B, C, U, R)
(__m512)__builtin_ia32_vfnmsubps512_mask3(A, B, C, U, R)
#define _mm512_maskz_fnmsub_round_ps(U, A, B, C, R) \
- (__m512)__builtin_ia32_vfmaddps512_maskz(-(A), B, -(C), U, R)
+ (__m512)__builtin_ia32_vfnmsubps512_maskz(A, B, C, U, R)
#endif
extern __inline __m512i
{
return (__m512i) __builtin_ia32_pabsq512_mask ((__v8di) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_pabsd512_mask ((__v16si) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pbroadcastd512 ((__v4si) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_pbroadcastd512_gpr_mask (__A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16)(-1));
}
{
return (__m512i) __builtin_ia32_pbroadcastq512 ((__v2di) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_set1_epi64 (long long __A)
{
-#ifdef TARGET_64BIT
return (__m512i) __builtin_ia32_pbroadcastq512_gpr_mask (__A,
(__v8di)
- _mm512_undefined_si512 (),
- (__mmask8)(-1));
-#else
- return (__m512i) __builtin_ia32_pbroadcastq512_mem_mask (__A,
- (__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8)(-1));
-#endif
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_set1_epi64 (__m512i __O, __mmask8 __M, long long __A)
{
-#ifdef TARGET_64BIT
return (__m512i) __builtin_ia32_pbroadcastq512_gpr_mask (__A, (__v8di) __O,
__M);
-#else
- return (__m512i) __builtin_ia32_pbroadcastq512_mem_mask (__A, (__v8di) __O,
- __M);
-#endif
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_maskz_set1_epi64 (__mmask8 __M, long long __A)
{
-#ifdef TARGET_64BIT
return (__m512i)
__builtin_ia32_pbroadcastq512_gpr_mask (__A,
(__v8di) _mm512_setzero_si512 (),
__M);
-#else
- return (__m512i)
- __builtin_ia32_pbroadcastq512_mem_mask (__A,
- (__v8di) _mm512_setzero_si512 (),
- __M);
-#endif
}
extern __inline __m512
{
return (__m512i) __builtin_ia32_broadcasti32x4_512 ((__v4si) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_broadcasti64x4_512 ((__v4di) __A,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_pshufd512_mask ((__v16si) __A,
__mask,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_shuf_i64x2_mask ((__v8di) __A,
(__v8di) __B, __imm,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
(__v16si) __B,
__imm,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
#else
#define _mm512_shuffle_epi32(X, C) \
((__m512i) __builtin_ia32_pshufd512_mask ((__v16si)(__m512i)(X), (int)(C),\
- (__v16si)(__m512i)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_shuffle_epi32(W, U, X, C) \
#define _mm512_shuffle_i64x2(X, Y, C) \
((__m512i) __builtin_ia32_shuf_i64x2_mask ((__v8di)(__m512i)(X), \
(__v8di)(__m512i)(Y), (int)(C),\
- (__v8di)(__m512i)_mm512_undefined_si512 (),\
+ (__v8di)(__m512i)_mm512_undefined_epi32 (),\
(__mmask8)-1))
#define _mm512_mask_shuffle_i64x2(W, U, X, Y, C) \
#define _mm512_shuffle_i32x4(X, Y, C) \
((__m512i) __builtin_ia32_shuf_i32x4_mask ((__v16si)(__m512i)(X), \
(__v16si)(__m512i)(Y), (int)(C),\
- (__v16si)(__m512i)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_shuffle_i32x4(W, U, X, Y, C) \
return (__m512i) __builtin_ia32_prolvd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_prorvd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_prolvq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_prorvq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1, __R);
}
{
return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1, __R);
}
}
#else
#define _mm512_cvtt_roundps_epi32(A, B) \
- ((__m512i)__builtin_ia32_cvttps2dq512_mask(A, (__v16si)_mm512_undefined_si512 (), -1, B))
+ ((__m512i)__builtin_ia32_cvttps2dq512_mask(A, (__v16si)_mm512_undefined_epi32 (), -1, B))
#define _mm512_mask_cvtt_roundps_epi32(W, U, A, B) \
((__m512i)__builtin_ia32_cvttps2dq512_mask(A, (__v16si)(W), U, B))
((__m512i)__builtin_ia32_cvttps2dq512_mask(A, (__v16si)_mm512_setzero_si512 (), U, B))
#define _mm512_cvtt_roundps_epu32(A, B) \
- ((__m512i)__builtin_ia32_cvttps2udq512_mask(A, (__v16si)_mm512_undefined_si512 (), -1, B))
+ ((__m512i)__builtin_ia32_cvttps2udq512_mask(A, (__v16si)_mm512_undefined_epi32 (), -1, B))
#define _mm512_mask_cvtt_roundps_epu32(W, U, A, B) \
((__m512i)__builtin_ia32_cvttps2udq512_mask(A, (__v16si)(W), U, B))
{
return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1, __R);
}
{
return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1, __R);
}
}
#else
#define _mm512_cvt_roundps_epi32(A, B) \
- ((__m512i)__builtin_ia32_cvtps2dq512_mask(A, (__v16si)_mm512_undefined_si512 (), -1, B))
+ ((__m512i)__builtin_ia32_cvtps2dq512_mask(A, (__v16si)_mm512_undefined_epi32 (), -1, B))
#define _mm512_mask_cvt_roundps_epi32(W, U, A, B) \
((__m512i)__builtin_ia32_cvtps2dq512_mask(A, (__v16si)(W), U, B))
((__m512i)__builtin_ia32_cvtps2dq512_mask(A, (__v16si)_mm512_setzero_si512 (), U, B))
#define _mm512_cvt_roundps_epu32(A, B) \
- ((__m512i)__builtin_ia32_cvtps2udq512_mask(A, (__v16si)_mm512_undefined_si512 (), -1, B))
+ ((__m512i)__builtin_ia32_cvtps2udq512_mask(A, (__v16si)_mm512_undefined_epi32 (), -1, B))
#define _mm512_mask_cvt_roundps_epu32(W, U, A, B) \
((__m512i)__builtin_ia32_cvtps2udq512_mask(A, (__v16si)(W), U, B))
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cvtsepi64_epi32 (__m512i __A)
{
- __v8si __O;
return (__m256i) __builtin_ia32_pmovsqd512_mask ((__v8di) __A,
(__v8si)
_mm256_undefined_si256 (),
(__v4di) __B,
__imm,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#define _mm512_inserti64x4(X, Y, C) \
((__m512i) __builtin_ia32_inserti64x4_mask ((__v8di)(__m512i) (X), \
(__v4di)(__m256i) (Y), (int) (C), \
- (__v8di)(__m512i)_mm512_undefined_si512 (), \
+ (__v8di)(__m512i)_mm512_undefined_epi32 (), \
(__mmask8)-1))
#define _mm512_mask_inserti64x4(W, U, X, Y, C) \
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_loadu_pd (void const *__P)
{
- return (__m512d) __builtin_ia32_loadupd512_mask ((const __v8df *) __P,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1);
+ return *(__m512d_u *)__P;
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_loadu_pd (__m512d __W, __mmask8 __U, void const *__P)
{
- return (__m512d) __builtin_ia32_loadupd512_mask ((const __v8df *) __P,
+ return (__m512d) __builtin_ia32_loadupd512_mask ((const double *) __P,
(__v8df) __W,
(__mmask8) __U);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_maskz_loadu_pd (__mmask8 __U, void const *__P)
{
- return (__m512d) __builtin_ia32_loadupd512_mask ((const __v8df *) __P,
+ return (__m512d) __builtin_ia32_loadupd512_mask ((const double *) __P,
(__v8df)
_mm512_setzero_pd (),
(__mmask8) __U);
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_storeu_pd (void *__P, __m512d __A)
{
- __builtin_ia32_storeupd512_mask ((__v8df *) __P, (__v8df) __A,
- (__mmask8) -1);
+ *(__m512d_u *)__P = __A;
}
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_storeu_pd (void *__P, __mmask8 __U, __m512d __A)
{
- __builtin_ia32_storeupd512_mask ((__v8df *) __P, (__v8df) __A,
+ __builtin_ia32_storeupd512_mask ((double *) __P, (__v8df) __A,
(__mmask8) __U);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_loadu_ps (void const *__P)
{
- return (__m512) __builtin_ia32_loadups512_mask ((const __v16sf *) __P,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1);
+ return *(__m512_u *)__P;
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_loadu_ps (__m512 __W, __mmask16 __U, void const *__P)
{
- return (__m512) __builtin_ia32_loadups512_mask ((const __v16sf *) __P,
+ return (__m512) __builtin_ia32_loadups512_mask ((const float *) __P,
(__v16sf) __W,
(__mmask16) __U);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_maskz_loadu_ps (__mmask16 __U, void const *__P)
{
- return (__m512) __builtin_ia32_loadups512_mask ((const __v16sf *) __P,
+ return (__m512) __builtin_ia32_loadups512_mask ((const float *) __P,
(__v16sf)
_mm512_setzero_ps (),
(__mmask16) __U);
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_storeu_ps (void *__P, __m512 __A)
{
- __builtin_ia32_storeups512_mask ((__v16sf *) __P, (__v16sf) __A,
- (__mmask16) -1);
+ *(__m512_u *)__P = __A;
}
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_storeu_ps (void *__P, __mmask16 __U, __m512 __A)
{
- __builtin_ia32_storeups512_mask ((__v16sf *) __P, (__v16sf) __A,
+ __builtin_ia32_storeups512_mask ((float *) __P, (__v16sf) __A,
(__mmask16) __U);
}
-extern __inline __m512i
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_loadu_epi64 (__m512i __W, __mmask8 __U, void const *__P)
+_mm_mask_load_ss (__m128 __W, __mmask8 __U, const float *__P)
{
- return (__m512i) __builtin_ia32_loaddqudi512_mask ((const __v8di *) __P,
- (__v8di) __W,
- (__mmask8) __U);
+ return (__m128) __builtin_ia32_loadss_mask (__P, (__v4sf) __W, __U);
}
-extern __inline __m512i
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
+_mm_maskz_load_ss (__mmask8 __U, const float *__P)
{
- return (__m512i) __builtin_ia32_loaddqudi512_mask ((const __v8di *) __P,
- (__v8di)
- _mm512_setzero_si512 (),
- (__mmask8) __U);
+ return (__m128) __builtin_ia32_loadss_mask (__P, (__v4sf) _mm_setzero_ps (),
+ __U);
}
-extern __inline void
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_storeu_epi64 (void *__P, __mmask8 __U, __m512i __A)
+_mm_mask_load_sd (__m128d __W, __mmask8 __U, const double *__P)
{
- __builtin_ia32_storedqudi512_mask ((__v8di *) __P, (__v8di) __A,
- (__mmask8) __U);
+ return (__m128d) __builtin_ia32_loadsd_mask (__P, (__v2df) __W, __U);
}
-extern __inline __m512i
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_loadu_si512 (void const *__P)
+_mm_maskz_load_sd (__mmask8 __U, const double *__P)
{
- return (__m512i) __builtin_ia32_loaddqusi512_mask ((const __v16si *) __P,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) -1);
+ return (__m128d) __builtin_ia32_loadsd_mask (__P, (__v2df) _mm_setzero_pd (),
+ __U);
}
-extern __inline __m512i
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_loadu_epi32 (__m512i __W, __mmask16 __U, void const *__P)
+_mm_mask_move_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512i) __builtin_ia32_loaddqusi512_mask ((const __v16si *) __P,
- (__v16si) __W,
- (__mmask16) __U);
+ return (__m128) __builtin_ia32_movess_mask ((__v4sf) __A, (__v4sf) __B,
+ (__v4sf) __W, __U);
}
-extern __inline __m512i
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_loadu_epi32 (__mmask16 __U, void const *__P)
+_mm_maskz_move_ss (__mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512i) __builtin_ia32_loaddqusi512_mask ((const __v16si *) __P,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) __U);
+ return (__m128) __builtin_ia32_movess_mask ((__v4sf) __A, (__v4sf) __B,
+ (__v4sf) _mm_setzero_ps (), __U);
}
-extern __inline void
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_storeu_si512 (void *__P, __m512i __A)
+_mm_mask_move_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- __builtin_ia32_storedqusi512_mask ((__v16si *) __P, (__v16si) __A,
- (__mmask16) -1);
+ return (__m128d) __builtin_ia32_movesd_mask ((__v2df) __A, (__v2df) __B,
+ (__v2df) __W, __U);
}
-extern __inline void
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_storeu_epi32 (void *__P, __mmask16 __U, __m512i __A)
+_mm_maskz_move_sd (__mmask8 __U, __m128d __A, __m128d __B)
{
- __builtin_ia32_storedqusi512_mask ((__v16si *) __P, (__v16si) __A,
- (__mmask16) __U);
+ return (__m128d) __builtin_ia32_movesd_mask ((__v2df) __A, (__v2df) __B,
+ (__v2df) _mm_setzero_pd (),
+ __U);
}
-extern __inline __m512d
+extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_permutevar_pd (__m512d __A, __m512i __C)
+_mm_mask_store_ss (float *__P, __mmask8 __U, __m128 __A)
{
- return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
- (__v8di) __C,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1);
+ __builtin_ia32_storess_mask (__P, (__v4sf) __A, (__mmask8) __U);
}
-extern __inline __m512d
+extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_permutevar_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512i __C)
+_mm_mask_store_sd (double *__P, __mmask8 __U, __m128d __A)
{
- return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
- (__v8di) __C,
- (__v8df) __W,
- (__mmask8) __U);
+ __builtin_ia32_storesd_mask (__P, (__v2df) __A, (__mmask8) __U);
}
-extern __inline __m512d
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_permutevar_pd (__mmask8 __U, __m512d __A, __m512i __C)
+_mm512_loadu_epi64 (void const *__P)
{
- return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
- (__v8di) __C,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U);
+ return *(__m512i_u *) __P;
}
-extern __inline __m512
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_loadu_epi64 (__m512i __W, __mmask8 __U, void const *__P)
+{
+ return (__m512i) __builtin_ia32_loaddqudi512_mask ((const long long *) __P,
+ (__v8di) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
+{
+ return (__m512i) __builtin_ia32_loaddqudi512_mask ((const long long *) __P,
+ (__v8di)
+ _mm512_setzero_si512 (),
+ (__mmask8) __U);
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_storeu_epi64 (void *__P, __m512i __A)
+{
+ *(__m512i_u *) __P = (__m512i_u) __A;
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_storeu_epi64 (void *__P, __mmask8 __U, __m512i __A)
+{
+ __builtin_ia32_storedqudi512_mask ((long long *) __P, (__v8di) __A,
+ (__mmask8) __U);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_loadu_si512 (void const *__P)
+{
+ return *(__m512i_u *)__P;
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_loadu_epi32 (void const *__P)
+{
+ return *(__m512i_u *) __P;
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_loadu_epi32 (__m512i __W, __mmask16 __U, void const *__P)
+{
+ return (__m512i) __builtin_ia32_loaddqusi512_mask ((const int *) __P,
+ (__v16si) __W,
+ (__mmask16) __U);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_loadu_epi32 (__mmask16 __U, void const *__P)
+{
+ return (__m512i) __builtin_ia32_loaddqusi512_mask ((const int *) __P,
+ (__v16si)
+ _mm512_setzero_si512 (),
+ (__mmask16) __U);
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_storeu_si512 (void *__P, __m512i __A)
+{
+ *(__m512i_u *)__P = __A;
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_storeu_epi32 (void *__P, __m512i __A)
+{
+ *(__m512i_u *) __P = (__m512i_u) __A;
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_storeu_epi32 (void *__P, __mmask16 __U, __m512i __A)
+{
+ __builtin_ia32_storedqusi512_mask ((int *) __P, (__v16si) __A,
+ (__mmask16) __U);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_permutevar_pd (__m512d __A, __m512i __C)
+{
+ return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
+ (__v8di) __C,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_permutevar_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512i __C)
+{
+ return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
+ (__v8di) __C,
+ (__v8df) __W,
+ (__mmask8) __U);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_permutevar_pd (__mmask8 __U, __m512d __A, __m512i __C)
+{
+ return (__m512d) __builtin_ia32_vpermilvarpd512_mask ((__v8df) __A,
+ (__v8di) __C,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U);
+}
+
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_permutevar_ps (__m512 __A, __m512i __C)
{
{
return (__m512i) __builtin_ia32_permdi512_mask ((__v8di) __X, __I,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) (-1));
}
((__m512i) __builtin_ia32_permdi512_mask ((__v8di)(__m512i)(X), \
(int)(I), \
(__v8di)(__m512i) \
- (_mm512_undefined_si512 ()),\
+ (_mm512_undefined_epi32 ()),\
(__mmask8)(-1)))
#define _mm512_maskz_permutex_epi64(M, X, I) \
return (__m512i) __builtin_ia32_permvardi512_mask ((__v8di) __Y,
(__v8di) __X,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_permvarsi512_mask ((__v16si) __Y,
(__v16si) __X,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_or_si512 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pord512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A | (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_or_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pord512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A | (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_or_epi64 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_porq512_mask ((__v8di) __A,
- (__v8di) __B,
- (__v8di)
- _mm512_undefined_si512 (),
- (__mmask8) -1);
+ return (__m512i) ((__v8du) __A | (__v8du) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_xor_si512 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pxord512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A ^ (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_xor_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pxord512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A ^ (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_xor_epi64 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pxorq512_mask ((__v8di) __A,
- (__v8di) __B,
- (__v8di)
- _mm512_undefined_si512 (),
- (__mmask8) -1);
+ return (__m512i) ((__v8du) __A ^ (__v8du) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_xor_epi64 (__m512i __W, __mmask16 __U, __m512i __A, __m512i __B)
+_mm512_mask_xor_epi64 (__m512i __W, __mmask8 __U, __m512i __A, __m512i __B)
{
return (__m512i) __builtin_ia32_pxorq512_mask ((__v8di) __A,
(__v8di) __B,
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_xor_epi64 (__mmask16 __U, __m512i __A, __m512i __B)
+_mm512_maskz_xor_epi64 (__mmask8 __U, __m512i __A, __m512i __B)
{
return (__m512i) __builtin_ia32_pxorq512_mask ((__v8di) __A,
(__v8di) __B,
{
return (__m512i) __builtin_ia32_prold512_mask ((__v16si) __A, __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_prord512_mask ((__v16si) __A, __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
{
return (__m512i) __builtin_ia32_prolq512_mask ((__v8di) __A, __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
{
return (__m512i) __builtin_ia32_prorq512_mask ((__v8di) __A, __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#define _mm512_rol_epi32(A, B) \
((__m512i)__builtin_ia32_prold512_mask ((__v16si)(__m512i)(A), \
(int)(B), \
- (__v16si)_mm512_undefined_si512 (), \
+ (__v16si)_mm512_undefined_epi32 (), \
(__mmask16)(-1)))
#define _mm512_mask_rol_epi32(W, U, A, B) \
((__m512i)__builtin_ia32_prold512_mask ((__v16si)(__m512i)(A), \
#define _mm512_ror_epi32(A, B) \
((__m512i)__builtin_ia32_prord512_mask ((__v16si)(__m512i)(A), \
(int)(B), \
- (__v16si)_mm512_undefined_si512 (), \
+ (__v16si)_mm512_undefined_epi32 (), \
(__mmask16)(-1)))
#define _mm512_mask_ror_epi32(W, U, A, B) \
((__m512i)__builtin_ia32_prord512_mask ((__v16si)(__m512i)(A), \
#define _mm512_rol_epi64(A, B) \
((__m512i)__builtin_ia32_prolq512_mask ((__v8di)(__m512i)(A), \
(int)(B), \
- (__v8di)_mm512_undefined_si512 (), \
+ (__v8di)_mm512_undefined_epi32 (), \
(__mmask8)(-1)))
#define _mm512_mask_rol_epi64(W, U, A, B) \
((__m512i)__builtin_ia32_prolq512_mask ((__v8di)(__m512i)(A), \
#define _mm512_ror_epi64(A, B) \
((__m512i)__builtin_ia32_prorq512_mask ((__v8di)(__m512i)(A), \
(int)(B), \
- (__v8di)_mm512_undefined_si512 (), \
+ (__v8di)_mm512_undefined_epi32 (), \
(__mmask8)(-1)))
#define _mm512_mask_ror_epi64(W, U, A, B) \
((__m512i)__builtin_ia32_prorq512_mask ((__v8di)(__m512i)(A), \
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_and_si512 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pandd512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A & (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_and_epi32 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pandd512_mask ((__v16si) __A,
- (__v16si) __B,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1);
+ return (__m512i) ((__v16su) __A & (__v16su) __B);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_and_epi64 (__m512i __A, __m512i __B)
{
- return (__m512i) __builtin_ia32_pandq512_mask ((__v8di) __A,
- (__v8di) __B,
- (__v8di)
- _mm512_undefined_si512 (),
- (__mmask8) -1);
+ return (__m512i) ((__v8du) __A & (__v8du) __B);
}
extern __inline __m512i
return (__m512i) __builtin_ia32_pandnd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_pandnd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_pandnq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
(__v8di) __B, __U);
}
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_abs_ps (__m512 __A)
+{
+ return (__m512) _mm512_and_epi32 ((__m512i) __A,
+ _mm512_set1_epi32 (0x7fffffff));
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_abs_ps (__m512 __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512) _mm512_mask_and_epi32 ((__m512i) __W, __U, (__m512i) __A,
+ _mm512_set1_epi32 (0x7fffffff));
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_abs_pd (__m512d __A)
+{
+ return (__m512d) _mm512_and_epi64 ((__m512i) __A,
+ _mm512_set1_epi64 (0x7fffffffffffffffLL));
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_abs_pd (__m512d __W, __mmask8 __U, __m512d __A)
+{
+ return (__m512d)
+ _mm512_mask_and_epi64 ((__m512i) __W, __U, (__m512i) __A,
+ _mm512_set1_epi64 (0x7fffffffffffffffLL));
+}
+
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_unpackhi_epi32 (__m512i __A, __m512i __B)
return (__m512i) __builtin_ia32_punpckhdq512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_punpckhqdq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_punpckldq512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_punpcklqdq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return __builtin_ia32_movntdqa512 ((__v8di *)__P);
}
+/* Constants for mantissa extraction */
+typedef enum
+{
+ _MM_MANT_NORM_1_2, /* interval [1, 2) */
+ _MM_MANT_NORM_p5_2, /* interval [0.5, 2) */
+ _MM_MANT_NORM_p5_1, /* interval [0.5, 1) */
+ _MM_MANT_NORM_p75_1p5 /* interval [0.75, 1.5) */
+} _MM_MANTISSA_NORM_ENUM;
+
+typedef enum
+{
+ _MM_MANT_SIGN_src, /* sign = sign(SRC) */
+ _MM_MANT_SIGN_zero, /* sign = 0 */
+ _MM_MANT_SIGN_nan /* DEST = NaN if sign(SRC) = 1 */
+} _MM_MANTISSA_SIGN_ENUM;
+
#ifdef __OPTIMIZE__
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getexp_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_getexpss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getexp_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_getexpss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_getexp_round_sd (__m128d __A, __m128d __B, const int __R)
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getexp_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_getexpsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getexp_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_getexpsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_getexp_round_ps (__m512 __A, const int __R)
(__mmask8) __U, __R);
}
-/* Constants for mantissa extraction */
-typedef enum
-{
- _MM_MANT_NORM_1_2, /* interval [1, 2) */
- _MM_MANT_NORM_p5_2, /* interval [0.5, 2) */
- _MM_MANT_NORM_p5_1, /* interval [0.5, 1) */
- _MM_MANT_NORM_p75_1p5 /* interval [0.75, 1.5) */
-} _MM_MANTISSA_NORM_ENUM;
-
-typedef enum
-{
- _MM_MANT_SIGN_src, /* sign = sign(SRC) */
- _MM_MANT_SIGN_zero, /* sign = 0 */
- _MM_MANT_SIGN_nan /* DEST = NaN if sign(SRC) = 1 */
-} _MM_MANTISSA_SIGN_ENUM;
-
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_getmant_round_pd (__m512d __A, _MM_MANTISSA_NORM_ENUM __B,
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getmant_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D, const int __R)
+{
+ return (__m128d) __builtin_ia32_getmantsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__D << 2) | __C,
+ (__v2df) __W,
+ __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getmant_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D, const int __R)
+{
+ return (__m128d) __builtin_ia32_getmantsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__D << 2) | __C,
+ (__v2df)
+ _mm_setzero_pd(),
+ __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_getmant_round_ss (__m128 __A, __m128 __B,
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getmant_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D, const int __R)
+{
+ return (__m128) __builtin_ia32_getmantss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__D << 2) | __C,
+ (__v4sf) __W,
+ __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getmant_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D, const int __R)
+{
+ return (__m128) __builtin_ia32_getmantss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__D << 2) | __C,
+ (__v4sf)
+ _mm_setzero_ps(),
+ __U, __R);
+}
+
#else
#define _mm512_getmant_round_pd(X, B, C, R) \
((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
(int)(((D)<<2) | (C)), \
(R)))
+#define _mm_mask_getmant_round_sd(W, U, X, Y, C, D, R) \
+ ((__m128d)__builtin_ia32_getmantsd_mask_round ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v2df)(__m128d)(W), \
+ (__mmask8)(U),\
+ (R)))
+
+#define _mm_maskz_getmant_round_sd(U, X, Y, C, D, R) \
+ ((__m128d)__builtin_ia32_getmantsd_mask_round ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v2df)(__m128d)_mm_setzero_pd(), \
+ (__mmask8)(U),\
+ (R)))
+
#define _mm_getmant_round_ss(X, Y, C, D, R) \
((__m128)__builtin_ia32_getmantss_round ((__v4sf)(__m128)(X), \
(__v4sf)(__m128)(Y), \
(int)(((D)<<2) | (C)), \
(R)))
-#define _mm_getexp_round_ss(A, B, R) \
- ((__m128)__builtin_ia32_getexpss128_round((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), R))
+#define _mm_mask_getmant_round_ss(W, U, X, Y, C, D, R) \
+ ((__m128)__builtin_ia32_getmantss_mask_round ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v4sf)(__m128)(W), \
+ (__mmask8)(U),\
+ (R)))
+
+#define _mm_maskz_getmant_round_ss(U, X, Y, C, D, R) \
+ ((__m128)__builtin_ia32_getmantss_mask_round ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v4sf)(__m128)_mm_setzero_ps(), \
+ (__mmask8)(U),\
+ (R)))
+
+#define _mm_getexp_round_ss(A, B, R) \
+ ((__m128)__builtin_ia32_getexpss128_round((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), R))
+
+#define _mm_mask_getexp_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_getexpss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_getexp_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_getexpss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
#define _mm_getexp_round_sd(A, B, R) \
((__m128d)__builtin_ia32_getexpsd128_round((__v2df)(__m128d)(A), (__v2df)(__m128d)(B), R))
+#define _mm_mask_getexp_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_getexpsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_getexp_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_getexpsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
+
+
#define _mm512_getexp_round_ps(A, R) \
((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
(__v16sf)_mm512_undefined_ps(), (__mmask16)-1, R))
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_roundscale_round_ss (__m128 __A, __m128 __B, const int __imm, const int __R)
+_mm_roundscale_round_ss (__m128 __A, __m128 __B, const int __imm,
+ const int __R)
+{
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __A,
+ (__v4sf) __B, __imm,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) -1,
+ __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_roundscale_round_ss (__m128 __A, __mmask8 __B, __m128 __C,
+ __m128 __D, const int __imm, const int __R)
+{
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __C,
+ (__v4sf) __D, __imm,
+ (__v4sf) __A,
+ (__mmask8) __B,
+ __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_roundscale_round_ss (__mmask8 __A, __m128 __B, __m128 __C,
+ const int __imm, const int __R)
{
- return (__m128) __builtin_ia32_rndscaless_round ((__v4sf) __A,
- (__v4sf) __B, __imm, __R);
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __B,
+ (__v4sf) __C, __imm,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __A,
+ __R);
}
extern __inline __m128d
_mm_roundscale_round_sd (__m128d __A, __m128d __B, const int __imm,
const int __R)
{
- return (__m128d) __builtin_ia32_rndscalesd_round ((__v2df) __A,
- (__v2df) __B, __imm, __R);
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __A,
+ (__v2df) __B, __imm,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) -1,
+ __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_roundscale_round_sd (__m128d __A, __mmask8 __B, __m128d __C,
+ __m128d __D, const int __imm, const int __R)
+{
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __C,
+ (__v2df) __D, __imm,
+ (__v2df) __A,
+ (__mmask8) __B,
+ __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_roundscale_round_sd (__mmask8 __A, __m128d __B, __m128d __C,
+ const int __imm, const int __R)
+{
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __B,
+ (__v2df) __C, __imm,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __A,
+ __R);
}
#else
(int)(C), \
(__v8df)_mm512_setzero_pd(),\
(__mmask8)(A), R))
-#define _mm_roundscale_round_ss(A, B, C, R) \
- ((__m128) __builtin_ia32_rndscaless_round ((__v4sf)(__m128)(A), \
- (__v4sf)(__m128)(B), (int)(C), R))
-#define _mm_roundscale_round_sd(A, B, C, R) \
- ((__m128d) __builtin_ia32_rndscalesd_round ((__v2df)(__m128d)(A), \
- (__v2df)(__m128d)(B), (int)(C), R))
+#define _mm_roundscale_round_ss(A, B, I, R) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (A), \
+ (__v4sf) (__m128) (B), \
+ (int) (I), \
+ (__v4sf) _mm_setzero_ps (), \
+ (__mmask8) (-1), \
+ (int) (R)))
+#define _mm_mask_roundscale_round_ss(A, U, B, C, I, R) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (B), \
+ (__v4sf) (__m128) (C), \
+ (int) (I), \
+ (__v4sf) (__m128) (A), \
+ (__mmask8) (U), \
+ (int) (R)))
+#define _mm_maskz_roundscale_round_ss(U, A, B, I, R) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (A), \
+ (__v4sf) (__m128) (B), \
+ (int) (I), \
+ (__v4sf) _mm_setzero_ps (), \
+ (__mmask8) (U), \
+ (int) (R)))
+#define _mm_roundscale_round_sd(A, B, I, R) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (A), \
+ (__v2df) (__m128d) (B), \
+ (int) (I), \
+ (__v2df) _mm_setzero_pd (), \
+ (__mmask8) (-1), \
+ (int) (R)))
+#define _mm_mask_roundscale_round_sd(A, U, B, C, I, R) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (B), \
+ (__v2df) (__m128d) (C), \
+ (int) (I), \
+ (__v2df) (__m128d) (A), \
+ (__mmask8) (U), \
+ (int) (R)))
+#define _mm_maskz_roundscale_round_sd(U, A, B, I, R) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (A), \
+ (__v2df) (__m128d) (B), \
+ (int) (I), \
+ (__v2df) _mm_setzero_pd (), \
+ (__mmask8) (U), \
+ (int) (R)))
#endif
extern __inline __m512
return (__m512i) __builtin_ia32_alignd512_mask ((__v16si) __A,
(__v16si) __B, __imm,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_alignq512_mask ((__v8di) __A,
(__v8di) __B, __imm,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
#else
#define _mm512_alignr_epi32(X, Y, C) \
((__m512i)__builtin_ia32_alignd512_mask ((__v16si)(__m512i)(X), \
- (__v16si)(__m512i)(Y), (int)(C), (__v16si)_mm512_undefined_si512 (),\
+ (__v16si)(__m512i)(Y), (int)(C), (__v16si)_mm512_undefined_epi32 (),\
(__mmask16)-1))
#define _mm512_mask_alignr_epi32(W, U, X, Y, C) \
#define _mm512_alignr_epi64(X, Y, C) \
((__m512i)__builtin_ia32_alignq512_mask ((__v8di)(__m512i)(X), \
- (__v8di)(__m512i)(Y), (int)(C), (__v8di)_mm512_undefined_si512 (), \
+ (__v8di)(__m512i)(Y), (int)(C), (__v8di)_mm512_undefined_epi32 (), \
(__mmask8)-1))
#define _mm512_mask_alignr_epi64(W, U, X, Y, C) \
(__mmask16) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpge_epi32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_cmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 5,
+ (__mmask16) __M);
+}
+
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpge_epu32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 5,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpge_epu32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpge_epi64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_cmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 5,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpge_epi64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpge_epu64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 5,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpge_epu64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmple_epi32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_cmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 2,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmple_epi32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmple_epu32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 2,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmple_epu32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmple_epi64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_cmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 2,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmple_epi64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmple_epu64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 2,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmple_epu64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmplt_epi32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_cmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 1,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmplt_epi32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmplt_epu32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 1,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmplt_epu32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmplt_epi64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_cmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 1,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmplt_epi64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmplt_epu64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 1,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmplt_epu64_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpneq_epi32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_cmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 4,
+ (__mmask16) __M);
+}
+
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpneq_epu32_mask (__mmask16 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __X,
+ (__v16si) __Y, 4,
+ (__mmask16) __M);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpneq_epu32_mask (__m512i __X, __m512i __Y)
(__mmask16) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpneq_epi64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_cmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 4,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpneq_epi64_mask (__m512i __X, __m512i __Y)
(__mmask8) -1);
}
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cmpneq_epu64_mask (__mmask8 __M, __m512i __X, __m512i __Y)
+{
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __X,
+ (__v8di) __Y, 4,
+ (__mmask8) __M);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmpneq_epu64_mask (__m512i __X, __m512i __Y)
#define _MM_CMPINT_GT 0x6
#ifdef __OPTIMIZE__
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kshiftli_mask16 (__mmask16 __A, unsigned int __B)
+{
+ return (__mmask16) __builtin_ia32_kshiftlihi ((__mmask16) __A,
+ (__mmask8) __B);
+}
+
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kshiftri_mask16 (__mmask16 __A, unsigned int __B)
+{
+ return (__mmask16) __builtin_ia32_kshiftrihi ((__mmask16) __A,
+ (__mmask8) __B);
+}
+
extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_cmp_epi64_mask (__m512i __X, __m512i __Y, const int __P)
}
#else
+#define _kshiftli_mask16(X, Y) \
+ ((__mmask16) __builtin_ia32_kshiftlihi ((__mmask16)(X), (__mmask8)(Y)))
+
+#define _kshiftri_mask16(X, Y) \
+ ((__mmask16) __builtin_ia32_kshiftrihi ((__mmask16)(X), (__mmask8)(Y)))
+
#define _mm512_cmp_epi64_mask(X, Y, P) \
((__mmask8) __builtin_ia32_cmpq512_mask ((__v8di)(__m512i)(X), \
(__v8di)(__m512i)(Y), (int)(P),\
(__mmask8)-1))
#define _mm512_cmp_epi32_mask(X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpd512_mask ((__v16si)(__m512i)(X), \
- (__v16si)(__m512i)(Y), (int)(P),\
- (__mmask16)-1))
+ ((__mmask16) __builtin_ia32_cmpd512_mask ((__v16si)(__m512i)(X), \
+ (__v16si)(__m512i)(Y), (int)(P), \
+ (__mmask16)-1))
#define _mm512_cmp_epu64_mask(X, Y, P) \
((__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di)(__m512i)(X), \
(__mmask8)-1))
#define _mm512_cmp_epu32_mask(X, Y, P) \
- ((__mmask8) __builtin_ia32_ucmpd512_mask ((__v16si)(__m512i)(X), \
- (__v16si)(__m512i)(Y), (int)(P),\
- (__mmask16)-1))
+ ((__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si)(__m512i)(X), \
+ (__v16si)(__m512i)(Y), (int)(P), \
+ (__mmask16)-1))
-#define _mm512_cmp_round_pd_mask(X, Y, P, R) \
+#define _mm512_cmp_round_pd_mask(X, Y, P, R) \
((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
(__v8df)(__m512d)(Y), (int)(P),\
(__mmask8)-1, R))
-#define _mm512_cmp_round_ps_mask(X, Y, P, R) \
+#define _mm512_cmp_round_ps_mask(X, Y, P, R) \
((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
(__v16sf)(__m512)(Y), (int)(P),\
(__mmask16)-1, R))
-#define _mm512_mask_cmp_epi64_mask(M, X, Y, P) \
+#define _mm512_mask_cmp_epi64_mask(M, X, Y, P) \
((__mmask8) __builtin_ia32_cmpq512_mask ((__v8di)(__m512i)(X), \
(__v8di)(__m512i)(Y), (int)(P),\
(__mmask8)M))
-#define _mm512_mask_cmp_epi32_mask(M, X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpd512_mask ((__v16si)(__m512i)(X), \
- (__v16si)(__m512i)(Y), (int)(P),\
- (__mmask16)M))
+#define _mm512_mask_cmp_epi32_mask(M, X, Y, P) \
+ ((__mmask16) __builtin_ia32_cmpd512_mask ((__v16si)(__m512i)(X), \
+ (__v16si)(__m512i)(Y), (int)(P), \
+ (__mmask16)M))
-#define _mm512_mask_cmp_epu64_mask(M, X, Y, P) \
+#define _mm512_mask_cmp_epu64_mask(M, X, Y, P) \
((__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di)(__m512i)(X), \
(__v8di)(__m512i)(Y), (int)(P),\
(__mmask8)M))
-#define _mm512_mask_cmp_epu32_mask(M, X, Y, P) \
- ((__mmask8) __builtin_ia32_ucmpd512_mask ((__v16si)(__m512i)(X), \
- (__v16si)(__m512i)(Y), (int)(P),\
- (__mmask16)M))
+#define _mm512_mask_cmp_epu32_mask(M, X, Y, P) \
+ ((__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si)(__m512i)(X), \
+ (__v16si)(__m512i)(Y), (int)(P), \
+ (__mmask16)M))
-#define _mm512_mask_cmp_round_pd_mask(M, X, Y, P, R) \
+#define _mm512_mask_cmp_round_pd_mask(M, X, Y, P, R) \
((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
(__v8df)(__m512d)(Y), (int)(P),\
(__mmask8)M, R))
-#define _mm512_mask_cmp_round_ps_mask(M, X, Y, P, R) \
+#define _mm512_mask_cmp_round_ps_mask(M, X, Y, P, R) \
((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
(__v16sf)(__m512)(Y), (int)(P),\
(__mmask16)M, R))
-#define _mm_cmp_round_sd_mask(X, Y, P, R) \
+#define _mm_cmp_round_sd_mask(X, Y, P, R) \
((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
(__v2df)(__m128d)(Y), (int)(P),\
(__mmask8)-1, R))
-#define _mm_mask_cmp_round_sd_mask(M, X, Y, P, R) \
+#define _mm_mask_cmp_round_sd_mask(M, X, Y, P, R) \
((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
(__v2df)(__m128d)(Y), (int)(P),\
(M), R))
-#define _mm_cmp_round_ss_mask(X, Y, P, R) \
+#define _mm_cmp_round_ss_mask(X, Y, P, R) \
((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
(__v4sf)(__m128)(Y), (int)(P), \
(__mmask8)-1, R))
-#define _mm_mask_cmp_round_ss_mask(M, X, Y, P, R) \
+#define _mm_mask_cmp_round_ss_mask(M, X, Y, P, R) \
((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
(__v4sf)(__m128)(Y), (int)(P), \
(M), R))
#ifdef __OPTIMIZE__
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32gather_ps (__m512i __index, float const *__addr, int __scale)
+_mm512_i32gather_ps (__m512i __index, void const *__addr, int __scale)
{
- __m512 v1_old = _mm512_undefined_ps ();
- __mmask16 mask = 0xFFFF;
+ __m512 __v1_old = _mm512_undefined_ps ();
+ __mmask16 __mask = 0xFFFF;
- return (__m512) __builtin_ia32_gathersiv16sf ((__v16sf) v1_old,
+ return (__m512) __builtin_ia32_gathersiv16sf ((__v16sf) __v1_old,
__addr,
(__v16si) __index,
- mask, __scale);
+ __mask, __scale);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i32gather_ps (__m512 v1_old, __mmask16 __mask,
- __m512i __index, float const *__addr, int __scale)
+_mm512_mask_i32gather_ps (__m512 __v1_old, __mmask16 __mask,
+ __m512i __index, void const *__addr, int __scale)
{
- return (__m512) __builtin_ia32_gathersiv16sf ((__v16sf) v1_old,
+ return (__m512) __builtin_ia32_gathersiv16sf ((__v16sf) __v1_old,
__addr,
(__v16si) __index,
__mask, __scale);
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32gather_pd (__m256i __index, double const *__addr, int __scale)
+_mm512_i32gather_pd (__m256i __index, void const *__addr, int __scale)
{
- __m512d v1_old = _mm512_undefined_pd ();
- __mmask8 mask = 0xFF;
+ __m512d __v1_old = _mm512_undefined_pd ();
+ __mmask8 __mask = 0xFF;
- return (__m512d) __builtin_ia32_gathersiv8df ((__v8df) v1_old,
+ return (__m512d) __builtin_ia32_gathersiv8df ((__v8df) __v1_old,
__addr,
- (__v8si) __index, mask,
+ (__v8si) __index, __mask,
__scale);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i32gather_pd (__m512d __v1_old, __mmask8 __mask,
- __m256i __index, double const *__addr, int __scale)
+ __m256i __index, void const *__addr, int __scale)
{
return (__m512d) __builtin_ia32_gathersiv8df ((__v8df) __v1_old,
__addr,
extern __inline __m256
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64gather_ps (__m512i __index, float const *__addr, int __scale)
+_mm512_i64gather_ps (__m512i __index, void const *__addr, int __scale)
{
- __m256 v1_old = _mm256_undefined_ps ();
- __mmask8 mask = 0xFF;
+ __m256 __v1_old = _mm256_undefined_ps ();
+ __mmask8 __mask = 0xFF;
- return (__m256) __builtin_ia32_gatherdiv16sf ((__v8sf) v1_old,
+ return (__m256) __builtin_ia32_gatherdiv16sf ((__v8sf) __v1_old,
__addr,
- (__v8di) __index, mask,
+ (__v8di) __index, __mask,
__scale);
}
extern __inline __m256
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i64gather_ps (__m256 __v1_old, __mmask8 __mask,
- __m512i __index, float const *__addr, int __scale)
+ __m512i __index, void const *__addr, int __scale)
{
return (__m256) __builtin_ia32_gatherdiv16sf ((__v8sf) __v1_old,
__addr,
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64gather_pd (__m512i __index, double const *__addr, int __scale)
+_mm512_i64gather_pd (__m512i __index, void const *__addr, int __scale)
{
- __m512d v1_old = _mm512_undefined_pd ();
- __mmask8 mask = 0xFF;
+ __m512d __v1_old = _mm512_undefined_pd ();
+ __mmask8 __mask = 0xFF;
- return (__m512d) __builtin_ia32_gatherdiv8df ((__v8df) v1_old,
+ return (__m512d) __builtin_ia32_gatherdiv8df ((__v8df) __v1_old,
__addr,
- (__v8di) __index, mask,
+ (__v8di) __index, __mask,
__scale);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i64gather_pd (__m512d __v1_old, __mmask8 __mask,
- __m512i __index, double const *__addr, int __scale)
+ __m512i __index, void const *__addr, int __scale)
{
return (__m512d) __builtin_ia32_gatherdiv8df ((__v8df) __v1_old,
__addr,
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32gather_epi32 (__m512i __index, int const *__addr, int __scale)
+_mm512_i32gather_epi32 (__m512i __index, void const *__addr, int __scale)
{
- __m512i v1_old = _mm512_undefined_si512 ();
- __mmask16 mask = 0xFFFF;
+ __m512i __v1_old = _mm512_undefined_epi32 ();
+ __mmask16 __mask = 0xFFFF;
- return (__m512i) __builtin_ia32_gathersiv16si ((__v16si) v1_old,
+ return (__m512i) __builtin_ia32_gathersiv16si ((__v16si) __v1_old,
__addr,
(__v16si) __index,
- mask, __scale);
+ __mask, __scale);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i32gather_epi32 (__m512i __v1_old, __mmask16 __mask,
- __m512i __index, int const *__addr, int __scale)
+ __m512i __index, void const *__addr, int __scale)
{
return (__m512i) __builtin_ia32_gathersiv16si ((__v16si) __v1_old,
__addr,
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32gather_epi64 (__m256i __index, long long const *__addr, int __scale)
+_mm512_i32gather_epi64 (__m256i __index, void const *__addr, int __scale)
{
- __m512i v1_old = _mm512_undefined_si512 ();
- __mmask8 mask = 0xFF;
+ __m512i __v1_old = _mm512_undefined_epi32 ();
+ __mmask8 __mask = 0xFF;
- return (__m512i) __builtin_ia32_gathersiv8di ((__v8di) v1_old,
+ return (__m512i) __builtin_ia32_gathersiv8di ((__v8di) __v1_old,
__addr,
- (__v8si) __index, mask,
+ (__v8si) __index, __mask,
__scale);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i32gather_epi64 (__m512i __v1_old, __mmask8 __mask,
- __m256i __index, long long const *__addr,
+ __m256i __index, void const *__addr,
int __scale)
{
return (__m512i) __builtin_ia32_gathersiv8di ((__v8di) __v1_old,
extern __inline __m256i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64gather_epi32 (__m512i __index, int const *__addr, int __scale)
+_mm512_i64gather_epi32 (__m512i __index, void const *__addr, int __scale)
{
- __m256i v1_old = _mm256_undefined_si256 ();
- __mmask8 mask = 0xFF;
+ __m256i __v1_old = _mm256_undefined_si256 ();
+ __mmask8 __mask = 0xFF;
- return (__m256i) __builtin_ia32_gatherdiv16si ((__v8si) v1_old,
+ return (__m256i) __builtin_ia32_gatherdiv16si ((__v8si) __v1_old,
__addr,
(__v8di) __index,
- mask, __scale);
+ __mask, __scale);
}
extern __inline __m256i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i64gather_epi32 (__m256i __v1_old, __mmask8 __mask,
- __m512i __index, int const *__addr, int __scale)
+ __m512i __index, void const *__addr, int __scale)
{
return (__m256i) __builtin_ia32_gatherdiv16si ((__v8si) __v1_old,
__addr,
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64gather_epi64 (__m512i __index, long long const *__addr, int __scale)
+_mm512_i64gather_epi64 (__m512i __index, void const *__addr, int __scale)
{
- __m512i v1_old = _mm512_undefined_si512 ();
- __mmask8 mask = 0xFF;
+ __m512i __v1_old = _mm512_undefined_epi32 ();
+ __mmask8 __mask = 0xFF;
- return (__m512i) __builtin_ia32_gatherdiv8di ((__v8di) v1_old,
+ return (__m512i) __builtin_ia32_gatherdiv8di ((__v8di) __v1_old,
__addr,
- (__v8di) __index, mask,
+ (__v8di) __index, __mask,
__scale);
}
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask_i64gather_epi64 (__m512i __v1_old, __mmask8 __mask,
- __m512i __index, long long const *__addr,
+ __m512i __index, void const *__addr,
int __scale)
{
return (__m512i) __builtin_ia32_gatherdiv8di ((__v8di) __v1_old,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32scatter_ps (float *__addr, __m512i __index, __m512 __v1, int __scale)
+_mm512_i32scatter_ps (void *__addr, __m512i __index, __m512 __v1, int __scale)
{
__builtin_ia32_scattersiv16sf (__addr, (__mmask16) 0xFFFF,
(__v16si) __index, (__v16sf) __v1, __scale);
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i32scatter_ps (float *__addr, __mmask16 __mask,
+_mm512_mask_i32scatter_ps (void *__addr, __mmask16 __mask,
__m512i __index, __m512 __v1, int __scale)
{
__builtin_ia32_scattersiv16sf (__addr, __mask, (__v16si) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32scatter_pd (double *__addr, __m256i __index, __m512d __v1,
+_mm512_i32scatter_pd (void *__addr, __m256i __index, __m512d __v1,
int __scale)
{
__builtin_ia32_scattersiv8df (__addr, (__mmask8) 0xFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i32scatter_pd (double *__addr, __mmask8 __mask,
+_mm512_mask_i32scatter_pd (void *__addr, __mmask8 __mask,
__m256i __index, __m512d __v1, int __scale)
{
__builtin_ia32_scattersiv8df (__addr, __mask, (__v8si) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64scatter_ps (float *__addr, __m512i __index, __m256 __v1, int __scale)
+_mm512_i64scatter_ps (void *__addr, __m512i __index, __m256 __v1, int __scale)
{
__builtin_ia32_scatterdiv16sf (__addr, (__mmask8) 0xFF,
(__v8di) __index, (__v8sf) __v1, __scale);
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i64scatter_ps (float *__addr, __mmask8 __mask,
+_mm512_mask_i64scatter_ps (void *__addr, __mmask8 __mask,
__m512i __index, __m256 __v1, int __scale)
{
__builtin_ia32_scatterdiv16sf (__addr, __mask, (__v8di) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64scatter_pd (double *__addr, __m512i __index, __m512d __v1,
+_mm512_i64scatter_pd (void *__addr, __m512i __index, __m512d __v1,
int __scale)
{
__builtin_ia32_scatterdiv8df (__addr, (__mmask8) 0xFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i64scatter_pd (double *__addr, __mmask8 __mask,
+_mm512_mask_i64scatter_pd (void *__addr, __mmask8 __mask,
__m512i __index, __m512d __v1, int __scale)
{
__builtin_ia32_scatterdiv8df (__addr, __mask, (__v8di) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32scatter_epi32 (int *__addr, __m512i __index,
+_mm512_i32scatter_epi32 (void *__addr, __m512i __index,
__m512i __v1, int __scale)
{
__builtin_ia32_scattersiv16si (__addr, (__mmask16) 0xFFFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i32scatter_epi32 (int *__addr, __mmask16 __mask,
+_mm512_mask_i32scatter_epi32 (void *__addr, __mmask16 __mask,
__m512i __index, __m512i __v1, int __scale)
{
__builtin_ia32_scattersiv16si (__addr, __mask, (__v16si) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i32scatter_epi64 (long long *__addr, __m256i __index,
+_mm512_i32scatter_epi64 (void *__addr, __m256i __index,
__m512i __v1, int __scale)
{
__builtin_ia32_scattersiv8di (__addr, (__mmask8) 0xFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i32scatter_epi64 (long long *__addr, __mmask8 __mask,
+_mm512_mask_i32scatter_epi64 (void *__addr, __mmask8 __mask,
__m256i __index, __m512i __v1, int __scale)
{
__builtin_ia32_scattersiv8di (__addr, __mask, (__v8si) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64scatter_epi32 (int *__addr, __m512i __index,
+_mm512_i64scatter_epi32 (void *__addr, __m512i __index,
__m256i __v1, int __scale)
{
__builtin_ia32_scatterdiv16si (__addr, (__mmask8) 0xFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i64scatter_epi32 (int *__addr, __mmask8 __mask,
+_mm512_mask_i64scatter_epi32 (void *__addr, __mmask8 __mask,
__m512i __index, __m256i __v1, int __scale)
{
__builtin_ia32_scatterdiv16si (__addr, __mask, (__v8di) __index,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_i64scatter_epi64 (long long *__addr, __m512i __index,
+_mm512_i64scatter_epi64 (void *__addr, __m512i __index,
__m512i __v1, int __scale)
{
__builtin_ia32_scatterdiv8di (__addr, (__mmask8) 0xFF,
extern __inline void
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_i64scatter_epi64 (long long *__addr, __mmask8 __mask,
+_mm512_mask_i64scatter_epi64 (void *__addr, __mmask8 __mask,
__m512i __index, __m512i __v1, int __scale)
{
__builtin_ia32_scatterdiv8di (__addr, __mask, (__v8di) __index,
#else
#define _mm512_i32gather_ps(INDEX, ADDR, SCALE) \
(__m512) __builtin_ia32_gathersiv16sf ((__v16sf)_mm512_undefined_ps(),\
- (float const *)ADDR, \
+ (void const *)ADDR, \
(__v16si)(__m512i)INDEX, \
(__mmask16)0xFFFF, (int)SCALE)
#define _mm512_mask_i32gather_ps(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512) __builtin_ia32_gathersiv16sf ((__v16sf)(__m512)V1OLD, \
- (float const *)ADDR, \
+ (void const *)ADDR, \
(__v16si)(__m512i)INDEX, \
(__mmask16)MASK, (int)SCALE)
#define _mm512_i32gather_pd(INDEX, ADDR, SCALE) \
(__m512d) __builtin_ia32_gathersiv8df ((__v8df)_mm512_undefined_pd(), \
- (double const *)ADDR, \
+ (void const *)ADDR, \
(__v8si)(__m256i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i32gather_pd(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512d) __builtin_ia32_gathersiv8df ((__v8df)(__m512d)V1OLD, \
- (double const *)ADDR, \
+ (void const *)ADDR, \
(__v8si)(__m256i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i64gather_ps(INDEX, ADDR, SCALE) \
(__m256) __builtin_ia32_gatherdiv16sf ((__v8sf)_mm256_undefined_ps(), \
- (float const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i64gather_ps(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m256) __builtin_ia32_gatherdiv16sf ((__v8sf)(__m256)V1OLD, \
- (float const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i64gather_pd(INDEX, ADDR, SCALE) \
(__m512d) __builtin_ia32_gatherdiv8df ((__v8df)_mm512_undefined_pd(), \
- (double const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i64gather_pd(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512d) __builtin_ia32_gatherdiv8df ((__v8df)(__m512d)V1OLD, \
- (double const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i32gather_epi32(INDEX, ADDR, SCALE) \
- (__m512i) __builtin_ia32_gathersiv16si ((__v16si)_mm512_undefined_si512 (), \
- (int const *)ADDR, \
+ (__m512i) __builtin_ia32_gathersiv16si ((__v16si)_mm512_undefined_epi32 (), \
+ (void const *)ADDR, \
(__v16si)(__m512i)INDEX, \
(__mmask16)0xFFFF, (int)SCALE)
#define _mm512_mask_i32gather_epi32(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512i) __builtin_ia32_gathersiv16si ((__v16si)(__m512i)V1OLD, \
- (int const *)ADDR, \
+ (void const *)ADDR, \
(__v16si)(__m512i)INDEX, \
(__mmask16)MASK, (int)SCALE)
#define _mm512_i32gather_epi64(INDEX, ADDR, SCALE) \
- (__m512i) __builtin_ia32_gathersiv8di ((__v8di)_mm512_undefined_si512 (), \
- (long long const *)ADDR, \
+ (__m512i) __builtin_ia32_gathersiv8di ((__v8di)_mm512_undefined_epi32 (), \
+ (void const *)ADDR, \
(__v8si)(__m256i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i32gather_epi64(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512i) __builtin_ia32_gathersiv8di ((__v8di)(__m512i)V1OLD, \
- (long long const *)ADDR, \
+ (void const *)ADDR, \
(__v8si)(__m256i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i64gather_epi32(INDEX, ADDR, SCALE) \
(__m256i) __builtin_ia32_gatherdiv16si ((__v8si)_mm256_undefined_si256(), \
- (int const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i64gather_epi32(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m256i) __builtin_ia32_gatherdiv16si ((__v8si)(__m256i)V1OLD, \
- (int const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i64gather_epi64(INDEX, ADDR, SCALE) \
- (__m512i) __builtin_ia32_gatherdiv8di ((__v8di)_mm512_undefined_si512 (), \
- (long long const *)ADDR, \
+ (__m512i) __builtin_ia32_gatherdiv8di ((__v8di)_mm512_undefined_epi32 (), \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)0xFF, (int)SCALE)
#define _mm512_mask_i64gather_epi64(V1OLD, MASK, INDEX, ADDR, SCALE) \
(__m512i) __builtin_ia32_gatherdiv8di ((__v8di)(__m512i)V1OLD, \
- (long long const *)ADDR, \
+ (void const *)ADDR, \
(__v8di)(__m512i)INDEX, \
(__mmask8)MASK, (int)SCALE)
#define _mm512_i32scatter_ps(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv16sf ((float *)ADDR, (__mmask16)0xFFFF, \
+ __builtin_ia32_scattersiv16sf ((void *)ADDR, (__mmask16)0xFFFF, \
(__v16si)(__m512i)INDEX, \
(__v16sf)(__m512)V1, (int)SCALE)
#define _mm512_mask_i32scatter_ps(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv16sf ((float *)ADDR, (__mmask16)MASK, \
+ __builtin_ia32_scattersiv16sf ((void *)ADDR, (__mmask16)MASK, \
(__v16si)(__m512i)INDEX, \
(__v16sf)(__m512)V1, (int)SCALE)
#define _mm512_i32scatter_pd(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv8df ((double *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scattersiv8df ((void *)ADDR, (__mmask8)0xFF, \
(__v8si)(__m256i)INDEX, \
(__v8df)(__m512d)V1, (int)SCALE)
#define _mm512_mask_i32scatter_pd(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv8df ((double *)ADDR, (__mmask8)MASK, \
+ __builtin_ia32_scattersiv8df ((void *)ADDR, (__mmask8)MASK, \
(__v8si)(__m256i)INDEX, \
(__v8df)(__m512d)V1, (int)SCALE)
#define _mm512_i64scatter_ps(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv16sf ((float *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scatterdiv16sf ((void *)ADDR, (__mmask8)0xFF, \
(__v8di)(__m512i)INDEX, \
(__v8sf)(__m256)V1, (int)SCALE)
#define _mm512_mask_i64scatter_ps(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv16sf ((float *)ADDR, (__mmask16)MASK, \
+ __builtin_ia32_scatterdiv16sf ((void *)ADDR, (__mmask16)MASK, \
(__v8di)(__m512i)INDEX, \
(__v8sf)(__m256)V1, (int)SCALE)
#define _mm512_i64scatter_pd(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv8df ((double *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scatterdiv8df ((void *)ADDR, (__mmask8)0xFF, \
(__v8di)(__m512i)INDEX, \
(__v8df)(__m512d)V1, (int)SCALE)
#define _mm512_mask_i64scatter_pd(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv8df ((double *)ADDR, (__mmask8)MASK, \
+ __builtin_ia32_scatterdiv8df ((void *)ADDR, (__mmask8)MASK, \
(__v8di)(__m512i)INDEX, \
(__v8df)(__m512d)V1, (int)SCALE)
#define _mm512_i32scatter_epi32(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv16si ((int *)ADDR, (__mmask16)0xFFFF, \
+ __builtin_ia32_scattersiv16si ((void *)ADDR, (__mmask16)0xFFFF, \
(__v16si)(__m512i)INDEX, \
(__v16si)(__m512i)V1, (int)SCALE)
#define _mm512_mask_i32scatter_epi32(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv16si ((int *)ADDR, (__mmask16)MASK, \
+ __builtin_ia32_scattersiv16si ((void *)ADDR, (__mmask16)MASK, \
(__v16si)(__m512i)INDEX, \
(__v16si)(__m512i)V1, (int)SCALE)
#define _mm512_i32scatter_epi64(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv8di ((long long *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scattersiv8di ((void *)ADDR, (__mmask8)0xFF, \
(__v8si)(__m256i)INDEX, \
(__v8di)(__m512i)V1, (int)SCALE)
#define _mm512_mask_i32scatter_epi64(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scattersiv8di ((long long *)ADDR, (__mmask8)MASK, \
+ __builtin_ia32_scattersiv8di ((void *)ADDR, (__mmask8)MASK, \
(__v8si)(__m256i)INDEX, \
(__v8di)(__m512i)V1, (int)SCALE)
#define _mm512_i64scatter_epi32(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv16si ((int *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scatterdiv16si ((void *)ADDR, (__mmask8)0xFF, \
(__v8di)(__m512i)INDEX, \
(__v8si)(__m256i)V1, (int)SCALE)
#define _mm512_mask_i64scatter_epi32(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv16si ((int *)ADDR, (__mmask8)MASK, \
+ __builtin_ia32_scatterdiv16si ((void *)ADDR, (__mmask8)MASK, \
(__v8di)(__m512i)INDEX, \
(__v8si)(__m256i)V1, (int)SCALE)
#define _mm512_i64scatter_epi64(ADDR, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv8di ((long long *)ADDR, (__mmask8)0xFF, \
+ __builtin_ia32_scatterdiv8di ((void *)ADDR, (__mmask8)0xFF, \
(__v8di)(__m512i)INDEX, \
(__v8di)(__m512i)V1, (int)SCALE)
#define _mm512_mask_i64scatter_epi64(ADDR, MASK, INDEX, V1, SCALE) \
- __builtin_ia32_scatterdiv8di ((long long *)ADDR, (__mmask8)MASK, \
+ __builtin_ia32_scatterdiv8di ((void *)ADDR, (__mmask8)MASK, \
(__v8di)(__m512i)INDEX, \
(__v8di)(__m512i)V1, (int)SCALE)
#endif
}
/* Mask arithmetic operations */
+#define _kand_mask16 _mm512_kand
+#define _kandn_mask16 _mm512_kandn
+#define _knot_mask16 _mm512_knot
+#define _kor_mask16 _mm512_kor
+#define _kxnor_mask16 _mm512_kxnor
+#define _kxor_mask16 _mm512_kxor
+
+extern __inline unsigned char
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kortest_mask16_u8 (__mmask16 __A, __mmask16 __B, unsigned char *__CF)
+{
+ *__CF = (unsigned char) __builtin_ia32_kortestchi (__A, __B);
+ return (unsigned char) __builtin_ia32_kortestzhi (__A, __B);
+}
+
+extern __inline unsigned char
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kortestz_mask16_u8 (__mmask16 __A, __mmask16 __B)
+{
+ return (unsigned char) __builtin_ia32_kortestzhi ((__mmask16) __A,
+ (__mmask16) __B);
+}
+
+extern __inline unsigned char
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kortestc_mask16_u8 (__mmask16 __A, __mmask16 __B)
+{
+ return (unsigned char) __builtin_ia32_kortestchi ((__mmask16) __A,
+ (__mmask16) __B);
+}
+
+extern __inline unsigned int
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_cvtmask16_u32 (__mmask16 __A)
+{
+ return (unsigned int) __builtin_ia32_kmovw ((__mmask16 ) __A);
+}
+
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_cvtu32_mask16 (unsigned int __A)
+{
+ return (__mmask16) __builtin_ia32_kmovw ((__mmask16 ) __A);
+}
+
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_load_mask16 (__mmask16 *__A)
+{
+ return (__mmask16) __builtin_ia32_kmovw (*(__mmask16 *) __A);
+}
+
+extern __inline void
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_store_mask16 (__mmask16 *__A, __mmask16 __B)
+{
+ *(__mmask16 *) __A = __builtin_ia32_kmovw (__B);
+}
+
extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_kand (__mmask16 __A, __mmask16 __B)
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_kandn (__mmask16 __A, __mmask16 __B)
{
- return (__mmask16) __builtin_ia32_kandnhi ((__mmask16) __A, (__mmask16) __B);
+ return (__mmask16) __builtin_ia32_kandnhi ((__mmask16) __A,
+ (__mmask16) __B);
}
extern __inline __mmask16
return (__mmask16) __builtin_ia32_kunpckhi ((__mmask16) __A, (__mmask16) __B);
}
+extern __inline __mmask16
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_kunpackb_mask16 (__mmask8 __A, __mmask8 __B)
+{
+ return (__mmask16) __builtin_ia32_kunpckhi ((__mmask16) __A, (__mmask16) __B);
+}
+
#ifdef __OPTIMIZE__
extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
#define _mm512_maskz_insertf32x4(A, X, Y, C) \
((__m512) __builtin_ia32_insertf32x4_mask ((__v16sf)(__m512) (X), \
(__v4sf)(__m128) (Y), (int) (C), (__v16sf)_mm512_setzero_ps(), \
- (__mmask8)(A)))
+ (__mmask16)(A)))
#define _mm512_maskz_inserti32x4(A, X, Y, C) \
((__m512i) __builtin_ia32_inserti32x4_mask ((__v16si)(__m512i) (X), \
(__v4si)(__m128i) (Y), (int) (C), (__v16si)_mm512_setzero_si512 (), \
- (__mmask8)(A)))
+ (__mmask16)(A)))
#define _mm512_mask_insertf32x4(A, B, X, Y, C) \
((__m512) __builtin_ia32_insertf32x4_mask ((__v16sf)(__m512) (X), \
(__v4sf)(__m128) (Y), (int) (C), (__v16sf)(__m512) (A), \
- (__mmask8)(B)))
+ (__mmask16)(B)))
#define _mm512_mask_inserti32x4(A, B, X, Y, C) \
((__m512i) __builtin_ia32_inserti32x4_mask ((__v16si)(__m512i) (X), \
(__v4si)(__m128i) (Y), (int) (C), (__v16si)(__m512i) (A), \
- (__mmask8)(B)))
+ (__mmask16)(B)))
#endif
extern __inline __m512i
return (__m512i) __builtin_ia32_pmaxsq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_pminsq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_pmaxuq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_pminuq512_mask ((__v8di) __A,
(__v8di) __B,
(__v8di)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask8) -1);
}
return (__m512i) __builtin_ia32_pmaxsd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_pminsd512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_pmaxud512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
return (__m512i) __builtin_ia32_pminud512_mask ((__v16si) __A,
(__v16si) __B,
(__v16si)
- _mm512_undefined_si512 (),
+ _mm512_undefined_epi32 (),
(__mmask16) -1);
}
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_max_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_maxsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_max_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_maxsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_max_round_ss (__m128 __A, __m128 __B, const int __R)
__R);
}
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_max_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_maxss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_max_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_maxss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_min_round_sd (__m128d __A, __m128d __B, const int __R)
__R);
}
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_min_round_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B, const int __R)
+{
+ return (__m128d) __builtin_ia32_minsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_min_round_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
+{
+ return (__m128d) __builtin_ia32_minsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U, __R);
+}
+
extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm_min_round_ss (__m128 __A, __m128 __B, const int __R)
__R);
}
-#else
-#define _mm_max_round_sd(A, B, C) \
- (__m128d)__builtin_ia32_addsd_round(A, B, C)
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_min_round_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B, const int __R)
+{
+ return (__m128) __builtin_ia32_minss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U, __R);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_min_round_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
+{
+ return (__m128) __builtin_ia32_minss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U, __R);
+}
+
+#else
+#define _mm_max_round_sd(A, B, C) \
+ (__m128d)__builtin_ia32_maxsd_round(A, B, C)
+
+#define _mm_mask_max_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_maxsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_max_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_maxsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
#define _mm_max_round_ss(A, B, C) \
- (__m128)__builtin_ia32_addss_round(A, B, C)
+ (__m128)__builtin_ia32_maxss_round(A, B, C)
+
+#define _mm_mask_max_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_maxss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_max_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_maxss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
#define _mm_min_round_sd(A, B, C) \
- (__m128d)__builtin_ia32_subsd_round(A, B, C)
+ (__m128d)__builtin_ia32_minsd_round(A, B, C)
+
+#define _mm_mask_min_round_sd(W, U, A, B, C) \
+ (__m128d)__builtin_ia32_minsd_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_min_round_sd(U, A, B, C) \
+ (__m128d)__builtin_ia32_minsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U, C)
#define _mm_min_round_ss(A, B, C) \
- (__m128)__builtin_ia32_subss_round(A, B, C)
+ (__m128)__builtin_ia32_minss_round(A, B, C)
+
+#define _mm_mask_min_round_ss(W, U, A, B, C) \
+ (__m128)__builtin_ia32_minss_mask_round(A, B, W, U, C)
+
+#define _mm_maskz_min_round_ss(U, A, B, C) \
+ (__m128)__builtin_ia32_minss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U, C)
+
#endif
extern __inline __m512d
(__m128)__builtin_ia32_vfmaddss3_round(A, -(B), -(C), R)
#endif
-#ifdef __OPTIMIZE__
-extern __inline int
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_comi_round_ss (__m128 __A, __m128 __B, const int __P, const int __R)
+_mm_mask_fmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return __builtin_ia32_vcomiss ((__v4sf) __A, (__v4sf) __B, __P, __R);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline int
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_comi_round_sd (__m128d __A, __m128d __B, const int __P, const int __R)
+_mm_mask_fmadd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
{
- return __builtin_ia32_vcomisd ((__v2df) __A, (__v2df) __B, __P, __R);
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-#else
-#define _mm_comi_round_ss(A, B, C, D)\
-__builtin_ia32_vcomiss(A, B, C, D)
-#define _mm_comi_round_sd(A, B, C, D)\
-__builtin_ia32_vcomisd(A, B, C, D)
-#endif
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_sqrt_pd (__m512d __A)
+_mm_mask3_fmadd_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U)
{
- return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask3 ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask3_fmadd_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U)
+{
+ return (__m128) __builtin_ia32_vfmaddss3_mask3 ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_sqrt_pd (__m512d __W, __mmask8 __U, __m512d __A)
+_mm_maskz_fmadd_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
- (__v8df) __W,
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_fmadd_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
(__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_sqrt_pd (__mmask8 __U, __m512d __A)
+_mm_mask_fmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
- (__v8df)
- _mm512_setzero_pd (),
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ (__v2df) __A,
+ -(__v2df) __B,
(__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_sqrt_ps (__m512 __A)
+_mm_mask_fmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ (__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_sqrt_ps (__m512 __W, __mmask16 __U, __m512 __A)
+_mm_mask3_fmsub_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmsubsd3_mask3 ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_sqrt_ps (__mmask16 __U, __m512 __A)
+_mm_mask3_fmsub_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmsubss3_mask3 ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_add_pd (__m512d __A, __m512d __B)
+_mm_maskz_fmsub_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_addpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ (__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_add_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_maskz_fmsub_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_addpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ (__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_add_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_addpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_fnmadd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
(__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_add_ps (__m512 __A, __m512 __B)
+_mm_mask3_fnmadd_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_addps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask3 ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_add_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fnmadd_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_addps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask3 ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_add_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fnmadd_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B)
{
- return (__m512) __builtin_ia32_addps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_sub_pd (__m512d __A, __m512d __B)
+_mm_maskz_fnmadd_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_subpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_sub_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_subpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ -(__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_sub_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_subpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ -(__v4sf) __A,
+ -(__v4sf) __B,
(__mmask8) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_sub_ps (__m512 __A, __m512 __B)
+_mm_mask3_fnmsub_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_subps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmsubsd3_mask3 ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_sub_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fnmsub_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U)
{
- return (__m512) __builtin_ia32_subps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmsubss3_mask3 ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_sub_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fnmsub_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B)
{
- return (__m512) __builtin_ia32_subps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ -(__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mul_pd (__m512d __A, __m512d __B)
+_mm_maskz_fnmsub_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_mulpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ -(__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+#ifdef __OPTIMIZE__
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_mul_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fmadd_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_mulpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_mul_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fmadd_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_mulpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mul_ps (__m512 __A, __m512 __B)
+_mm_mask3_fmadd_round_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_mulps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask3 ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_mul_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fmadd_round_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_mulps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask3 ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_mul_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fmadd_round_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512) __builtin_ia32_mulps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_div_pd (__m512d __M, __m512d __V)
+_mm_maskz_fmadd_round_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_divpd512_mask ((__v8df) __M,
- (__v8df) __V,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_div_pd (__m512d __W, __mmask8 __U, __m512d __M, __m512d __V)
+_mm_mask_fmsub_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_divpd512_mask ((__v8df) __M,
- (__v8df) __V,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ (__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_div_pd (__mmask8 __U, __m512d __M, __m512d __V)
+_mm_mask_fmsub_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_divpd512_mask ((__v8df) __M,
- (__v8df) __V,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ (__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_div_ps (__m512 __A, __m512 __B)
+_mm_mask3_fmsub_round_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_divps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmsubsd3_mask3 ((__v2df) __W,
+ (__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_div_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fmsub_round_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_divps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmsubss3_mask3 ((__v4sf) __W,
+ (__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_div_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fmsub_round_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512) __builtin_ia32_divps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ (__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_max_pd (__m512d __A, __m512d __B)
+_mm_maskz_fmsub_round_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ (__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_max_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmadd_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_max_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmadd_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_max_ps (__m512 __A, __m512 __B)
+_mm_mask3_fnmadd_round_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask3 ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_max_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fnmadd_round_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask3 ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_max_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fnmadd_round_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_min_pd (__m512d __A, __m512d __B)
+_mm_maskz_fnmadd_round_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_min_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmsub_round_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_mask ((__v2df) __W,
+ -(__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_min_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_mask_fnmsub_round_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_mask ((__v4sf) __W,
+ -(__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_min_ps (__m512 __A, __m512 __B)
+_mm_mask3_fnmsub_round_sd (__m128d __W, __m128d __A, __m128d __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmsubsd3_mask3 ((__v2df) __W,
+ -(__v2df) __A,
+ (__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_min_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm_mask3_fnmsub_round_ss (__m128 __W, __m128 __A, __m128 __B, __mmask8 __U,
+ const int __R)
{
- return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmsubss3_mask3 ((__v4sf) __W,
+ -(__v4sf) __A,
+ (__v4sf) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_min_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm_maskz_fnmsub_round_sd (__mmask8 __U, __m128d __W, __m128d __A, __m128d __B,
+ const int __R)
{
- return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_vfmaddsd3_maskz ((__v2df) __W,
+ -(__v2df) __A,
+ -(__v2df) __B,
+ (__mmask8) __U, __R);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_scalef_pd (__m512d __A, __m512d __B)
+_mm_maskz_fnmsub_round_ss (__mmask8 __U, __m128 __W, __m128 __A, __m128 __B,
+ const int __R)
{
- return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_vfmaddss3_maskz ((__v4sf) __W,
+ -(__v4sf) __A,
+ -(__v4sf) __B,
+ (__mmask8) __U, __R);
}
+#else
+#define _mm_mask_fmadd_round_sd(A, U, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask (A, B, C, U, R)
-extern __inline __m512d
+#define _mm_mask_fmadd_round_ss(A, U, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask (A, B, C, U, R)
+
+#define _mm_mask3_fmadd_round_sd(A, B, C, U, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask3 (A, B, C, U, R)
+
+#define _mm_mask3_fmadd_round_ss(A, B, C, U, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask3 (A, B, C, U, R)
+
+#define _mm_maskz_fmadd_round_sd(U, A, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_maskz (A, B, C, U, R)
+
+#define _mm_maskz_fmadd_round_ss(U, A, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_maskz (A, B, C, U, R)
+
+#define _mm_mask_fmsub_round_sd(A, U, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask (A, B, -(C), U, R)
+
+#define _mm_mask_fmsub_round_ss(A, U, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask (A, B, -(C), U, R)
+
+#define _mm_mask3_fmsub_round_sd(A, B, C, U, R) \
+ (__m128d) __builtin_ia32_vfmsubsd3_mask3 (A, B, C, U, R)
+
+#define _mm_mask3_fmsub_round_ss(A, B, C, U, R) \
+ (__m128) __builtin_ia32_vfmsubss3_mask3 (A, B, C, U, R)
+
+#define _mm_maskz_fmsub_round_sd(U, A, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_maskz (A, B, -(C), U, R)
+
+#define _mm_maskz_fmsub_round_ss(U, A, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_maskz (A, B, -(C), U, R)
+
+#define _mm_mask_fnmadd_round_sd(A, U, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask (A, -(B), C, U, R)
+
+#define _mm_mask_fnmadd_round_ss(A, U, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask (A, -(B), C, U, R)
+
+#define _mm_mask3_fnmadd_round_sd(A, B, C, U, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask3 (A, -(B), C, U, R)
+
+#define _mm_mask3_fnmadd_round_ss(A, B, C, U, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask3 (A, -(B), C, U, R)
+
+#define _mm_maskz_fnmadd_round_sd(U, A, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_maskz (A, -(B), C, U, R)
+
+#define _mm_maskz_fnmadd_round_ss(U, A, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_maskz (A, -(B), C, U, R)
+
+#define _mm_mask_fnmsub_round_sd(A, U, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_mask (A, -(B), -(C), U, R)
+
+#define _mm_mask_fnmsub_round_ss(A, U, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_mask (A, -(B), -(C), U, R)
+
+#define _mm_mask3_fnmsub_round_sd(A, B, C, U, R) \
+ (__m128d) __builtin_ia32_vfmsubsd3_mask3 (A, -(B), C, U, R)
+
+#define _mm_mask3_fnmsub_round_ss(A, B, C, U, R) \
+ (__m128) __builtin_ia32_vfmsubss3_mask3 (A, -(B), C, U, R)
+
+#define _mm_maskz_fnmsub_round_sd(U, A, B, C, R) \
+ (__m128d) __builtin_ia32_vfmaddsd3_maskz (A, -(B), -(C), U, R)
+
+#define _mm_maskz_fnmsub_round_ss(U, A, B, C, R) \
+ (__m128) __builtin_ia32_vfmaddss3_maskz (A, -(B), -(C), U, R)
+#endif
+
+#ifdef __OPTIMIZE__
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_scalef_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+_mm_comi_round_ss (__m128 __A, __m128 __B, const int __P, const int __R)
{
- return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return __builtin_ia32_vcomiss ((__v4sf) __A, (__v4sf) __B, __P, __R);
}
-extern __inline __m512d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_scalef_pd (__mmask8 __U, __m512d __A, __m512d __B)
+_mm_comi_round_sd (__m128d __A, __m128d __B, const int __P, const int __R)
{
- return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return __builtin_ia32_vcomisd ((__v2df) __A, (__v2df) __B, __P, __R);
}
+#else
+#define _mm_comi_round_ss(A, B, C, D)\
+__builtin_ia32_vcomiss(A, B, C, D)
+#define _mm_comi_round_sd(A, B, C, D)\
+__builtin_ia32_vcomisd(A, B, C, D)
+#endif
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_scalef_ps (__m512 __A, __m512 __B)
+_mm512_sqrt_pd (__m512d __A)
{
- return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_scalef_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+_mm512_mask_sqrt_pd (__m512d __W, __mmask8 __U, __m512d __A)
{
- return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_scalef_ps (__mmask16 __U, __m512 __A, __m512 __B)
+_mm512_maskz_sqrt_pd (__mmask8 __U, __m512d __A)
{
- return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_sqrtpd512_mask ((__v8df) __A,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m128d
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_scalef_sd (__m128d __A, __m128d __B)
+_mm512_sqrt_ps (__m512 __A)
{
- return (__m128d) __builtin_ia32_scalefsd_round ((__v2df) __A,
- (__v2df) __B,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m128
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_scalef_ss (__m128 __A, __m128 __B)
+_mm512_mask_sqrt_ps (__m512 __W, __mmask16 __U, __m512 __A)
{
- return (__m128) __builtin_ia32_scalefss_round ((__v4sf) __A,
- (__v4sf) __B,
+ return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
+ (__v16sf) __W,
+ (__mmask16) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmadd_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm512_maskz_sqrt_ps (__mmask16 __U, __m512 __A)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_sqrtps512_mask ((__v16sf) __A,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm512_add_pd (__m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) ((__v8df)__A + (__v8df)__B);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm512_mask_add_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask3 ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_addpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm512_maskz_add_pd (__mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_addpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmadd_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm512_add_ps (__m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) ((__v16sf)__A + (__v16sf)__B);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm512_mask_add_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_addps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmadd_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+_mm512_maskz_add_ps (__mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask3 ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_addps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm_mask_add_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_addsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmsub_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm_maskz_add_sd (__mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_addsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_add_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_addss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_add_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_addss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm512_sub_pd (__m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) ((__v8df)__A - (__v8df)__B);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm512_mask_sub_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmsubpd512_mask3 ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_subpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm512_maskz_sub_pd (__mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_subpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmsub_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm512_sub_ps (__m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) ((__v16sf)__A - (__v16sf)__B);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm512_mask_sub_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_subps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+_mm512_maskz_sub_ps (__mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmsubps512_mask3 ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_subps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm_mask_sub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_subsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmaddsub_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm_maskz_sub_sd (__mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_subsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmaddsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm_mask_sub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_subss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_sub_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_subss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmaddsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm512_mul_pd (__m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_mask3 ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) ((__v8df)__A * (__v8df)__B);
}
extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmaddsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm512_mask_mul_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_mulpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmaddsub_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm512_maskz_mul_pd (__mmask8 __U, __m512d __A, __m512d __B)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_mulpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmaddsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm512_mul_ps (__m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) ((__v16sf)__A * (__v16sf)__B);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmaddsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+_mm512_mask_mul_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_mask3 ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_mulps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmaddsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm512_maskz_mul_ps (__mmask16 __U, __m512 __A, __m512 __B)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_mulps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmsubadd_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm_mask_mul_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_mulsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmsubadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm_maskz_mul_sd (__mmask8 __U, __m128d __A, __m128d __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d) __builtin_ia32_mulsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fmsubadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm_mask_mul_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B)
{
- return (__m512d) __builtin_ia32_vfmsubaddpd512_mask3 ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_mulss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmsubadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm_maskz_mul_ss (__mmask8 __U, __m128 __A, __m128 __B)
{
- return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128) __builtin_ia32_mulss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fmsubadd_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm512_div_pd (__m512d __M, __m512d __V)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) ((__v8df)__M / (__v8df)__V);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fmsubadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm512_mask_div_pd (__m512d __W, __mmask8 __U, __m512d __M, __m512d __V)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_divpd512_mask ((__v8df) __M,
+ (__v8df) __V,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_div_pd (__mmask8 __U, __m512d __M, __m512d __V)
+{
+ return (__m512d) __builtin_ia32_divpd512_mask ((__v8df) __M,
+ (__v8df) __V,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_div_ps (__m512 __A, __m512 __B)
+{
+ return (__m512) ((__v16sf)__A / (__v16sf)__B);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_div_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_divps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_div_ps (__mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_divps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_div_sd (__m128d __W, __mmask8 __U, __m128d __A,
+ __m128d __B)
+{
+ return (__m128d) __builtin_ia32_divsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_div_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_divsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_div_ss (__m128 __W, __mmask8 __U, __m128 __A,
+ __m128 __B)
+{
+ return (__m128) __builtin_ia32_divss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_div_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_divss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_max_pd (__m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_max_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_max_pd (__mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_maxpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_max_ps (__m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_max_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_max_ps (__mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_maxps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_max_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_maxsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_max_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_maxsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_max_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_maxss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_max_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_maxss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_min_pd (__m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_min_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_min_pd (__mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_minpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_min_ps (__m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_min_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_min_ps (__mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_minps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_min_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_minsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_min_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_minsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_min_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_minss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_min_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_minss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_scalef_pd (__m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_scalef_pd (__m512d __W, __mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_scalef_pd (__mmask8 __U, __m512d __A, __m512d __B)
+{
+ return (__m512d) __builtin_ia32_scalefpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_scalef_ps (__m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_scalef_ps (__m512 __W, __mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_scalef_ps (__mmask16 __U, __m512 __A, __m512 __B)
+{
+ return (__m512) __builtin_ia32_scalefps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_scalef_sd (__m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_scalefsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_scalef_ss (__m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_scalefss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmadd_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfmaddpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmadd_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmadd_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+{
+ return (__m512) __builtin_ia32_vfmaddps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmsub_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfmsubpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmsubpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmsub_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+{
+ return (__m512) __builtin_ia32_vfmsubps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmsubps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmaddsub_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmaddsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmaddsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmaddsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmaddsub_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmaddsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmaddsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmaddsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmsubadd_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ -(__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmsubadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ -(__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fmsubadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfmsubaddpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmsubadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfmaddsubpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ -(__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fmsubadd_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ -(__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fmsubadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ -(__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_mask3_fmsubadd_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
{
- return (__m512) __builtin_ia32_vfmsubaddps512_mask3 ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_vfmsubaddps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fmsubadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ -(__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fnmadd_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmaddpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fnmadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmaddpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fnmadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfnmaddpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fnmadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmaddpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fnmadd_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmaddps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fnmadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmaddps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fnmadd_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+{
+ return (__m512) __builtin_ia32_vfnmaddps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fnmadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmaddps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fnmsub_pd (__m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fnmsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmsubpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fnmsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+{
+ return (__m512d) __builtin_ia32_vfnmsubpd512_mask3 ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fnmsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+{
+ return (__m512d) __builtin_ia32_vfnmsubpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8df) __C,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fnmsub_ps (__m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fnmsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmsubps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask3_fnmsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+{
+ return (__m512) __builtin_ia32_vfnmsubps512_mask3 ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fnmsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+{
+ return (__m512) __builtin_ia32_vfnmsubps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16sf) __C,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvttpd_epi32 (__m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_undefined_si256 (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvttpd_epi32 (__m256i __W, __mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
+ (__v8si) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvttpd_epi32 (__mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_setzero_si256 (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvttpd_epu32 (__m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_undefined_si256 (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvttpd_epu32 (__m256i __W, __mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
+ (__v8si) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvttpd_epu32 (__mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_setzero_si256 (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtpd_epi32 (__m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_undefined_si256 (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtpd_epi32 (__m256i __W, __mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
+ (__v8si) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtpd_epi32 (__mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_setzero_si256 (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtpd_epu32 (__m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_undefined_si256 (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtpd_epu32 (__m256i __W, __mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
+ (__v8si) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtpd_epu32 (__mmask8 __U, __m512d __A)
+{
+ return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
+ (__v8si)
+ _mm256_setzero_si256 (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvttps_epi32 (__m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_undefined_epi32 (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvttps_epi32 (__m512i __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
+ (__v16si) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvttps_epi32 (__mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_setzero_si512 (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvttps_epu32 (__m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_undefined_epi32 (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvttps_epu32 (__m512i __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
+ (__v16si) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvttps_epu32 (__mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_setzero_si512 (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtps_epi32 (__m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_undefined_epi32 (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtps_epi32 (__m512i __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
+ (__v16si) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtps_epi32 (__mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_setzero_si512 (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtps_epu32 (__m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_undefined_epi32 (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtps_epu32 (__m512i __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
+ (__v16si) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtps_epu32 (__mmask16 __U, __m512 __A)
+{
+ return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
+ (__v16si)
+ _mm512_setzero_si512 (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtsd_f64 (__m512d __A)
+{
+ return __A[0];
+}
+
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtss_f32 (__m512 __A)
+{
+ return __A[0];
+}
+
+#ifdef __x86_64__
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtu64_ss (__m128 __A, unsigned long long __B)
+{
+ return (__m128) __builtin_ia32_cvtusi2ss64 ((__v4sf) __A, __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtu64_sd (__m128d __A, unsigned long long __B)
+{
+ return (__m128d) __builtin_ia32_cvtusi2sd64 ((__v2df) __A, __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+#endif
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtu32_ss (__m128 __A, unsigned __B)
+{
+ return (__m128) __builtin_ia32_cvtusi2ss32 ((__v4sf) __A, __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtepi32_ps (__m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtepi32_ps (__m512 __W, __mmask16 __U, __m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtepi32_ps (__mmask16 __U, __m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtepu32_ps (__m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtepu32_ps (__m512 __W, __mmask16 __U, __m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtepu32_ps (__mmask16 __U, __m512i __A)
+{
+ return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+#ifdef __OPTIMIZE__
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fixupimm_pd (__m512d __A, __m512d __B, __m512i __C, const int __imm)
+{
+ return (__m512d) __builtin_ia32_fixupimmpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8di) __C,
+ __imm,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fixupimm_pd (__m512d __A, __mmask8 __U, __m512d __B,
+ __m512i __C, const int __imm)
+{
+ return (__m512d) __builtin_ia32_fixupimmpd512_mask ((__v8df) __A,
+ (__v8df) __B,
+ (__v8di) __C,
+ __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fixupimm_pd (__mmask8 __U, __m512d __A, __m512d __B,
+ __m512i __C, const int __imm)
+{
+ return (__m512d) __builtin_ia32_fixupimmpd512_maskz ((__v8df) __A,
+ (__v8df) __B,
+ (__v8di) __C,
+ __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_fixupimm_ps (__m512 __A, __m512 __B, __m512i __C, const int __imm)
+{
+ return (__m512) __builtin_ia32_fixupimmps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16si) __C,
+ __imm,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_fixupimm_ps (__m512 __A, __mmask16 __U, __m512 __B,
+ __m512i __C, const int __imm)
+{
+ return (__m512) __builtin_ia32_fixupimmps512_mask ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16si) __C,
+ __imm,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_fixupimm_ps (__mmask16 __U, __m512 __A, __m512 __B,
+ __m512i __C, const int __imm)
+{
+ return (__m512) __builtin_ia32_fixupimmps512_maskz ((__v16sf) __A,
+ (__v16sf) __B,
+ (__v16si) __C,
+ __imm,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_fixupimm_sd (__m128d __A, __m128d __B, __m128i __C, const int __imm)
+{
+ return (__m128d) __builtin_ia32_fixupimmsd_mask ((__v2df) __A,
+ (__v2df) __B,
+ (__v2di) __C, __imm,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_fixupimm_sd (__m128d __A, __mmask8 __U, __m128d __B,
+ __m128i __C, const int __imm)
+{
+ return (__m128d) __builtin_ia32_fixupimmsd_mask ((__v2df) __A,
+ (__v2df) __B,
+ (__v2di) __C, __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_fixupimm_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ __m128i __C, const int __imm)
+{
+ return (__m128d) __builtin_ia32_fixupimmsd_maskz ((__v2df) __A,
+ (__v2df) __B,
+ (__v2di) __C,
+ __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_fixupimm_ss (__m128 __A, __m128 __B, __m128i __C, const int __imm)
+{
+ return (__m128) __builtin_ia32_fixupimmss_mask ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4si) __C, __imm,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_fixupimm_ss (__m128 __A, __mmask8 __U, __m128 __B,
+ __m128i __C, const int __imm)
+{
+ return (__m128) __builtin_ia32_fixupimmss_mask ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4si) __C, __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_fixupimm_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ __m128i __C, const int __imm)
+{
+ return (__m128) __builtin_ia32_fixupimmss_maskz ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4si) __C, __imm,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+#else
+#define _mm512_fixupimm_pd(X, Y, Z, C) \
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask ((__v8df)(__m512d)(X), \
+ (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
+ (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_fixupimm_pd(X, U, Y, Z, C) \
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask ((__v8df)(__m512d)(X), \
+ (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_fixupimm_pd(U, X, Y, Z, C) \
+ ((__m512d)__builtin_ia32_fixupimmpd512_maskz ((__v8df)(__m512d)(X), \
+ (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_fixupimm_ps(X, Y, Z, C) \
+ ((__m512)__builtin_ia32_fixupimmps512_mask ((__v16sf)(__m512)(X), \
+ (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
+ (__mmask16)(-1), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_fixupimm_ps(X, U, Y, Z, C) \
+ ((__m512)__builtin_ia32_fixupimmps512_mask ((__v16sf)(__m512)(X), \
+ (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
+ (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_fixupimm_ps(U, X, Y, Z, C) \
+ ((__m512)__builtin_ia32_fixupimmps512_maskz ((__v16sf)(__m512)(X), \
+ (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
+ (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_fixupimm_sd(X, Y, Z, C) \
+ ((__m128d)__builtin_ia32_fixupimmsd_mask ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
+ (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_fixupimm_sd(X, U, Y, Z, C) \
+ ((__m128d)__builtin_ia32_fixupimmsd_mask ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_maskz_fixupimm_sd(U, X, Y, Z, C) \
+ ((__m128d)__builtin_ia32_fixupimmsd_maskz ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_fixupimm_ss(X, Y, Z, C) \
+ ((__m128)__builtin_ia32_fixupimmss_mask ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
+ (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_fixupimm_ss(X, U, Y, Z, C) \
+ ((__m128)__builtin_ia32_fixupimmss_mask ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_maskz_fixupimm_ss(U, X, Y, Z, C) \
+ ((__m128)__builtin_ia32_fixupimmss_maskz ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
+ (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+#endif
+
+#ifdef __x86_64__
+extern __inline unsigned long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtss_u64 (__m128 __A)
+{
+ return (unsigned long long) __builtin_ia32_vcvtss2usi64 ((__v4sf)
+ __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline unsigned long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttss_u64 (__m128 __A)
+{
+ return (unsigned long long) __builtin_ia32_vcvttss2usi64 ((__v4sf)
+ __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttss_i64 (__m128 __A)
+{
+ return (long long) __builtin_ia32_vcvttss2si64 ((__v4sf) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+#endif /* __x86_64__ */
+
+extern __inline unsigned
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtss_u32 (__m128 __A)
+{
+ return (unsigned) __builtin_ia32_vcvtss2usi32 ((__v4sf) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline unsigned
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttss_u32 (__m128 __A)
+{
+ return (unsigned) __builtin_ia32_vcvttss2usi32 ((__v4sf) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline int
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttss_i32 (__m128 __A)
+{
+ return (int) __builtin_ia32_vcvttss2si32 ((__v4sf) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+#ifdef __x86_64__
+extern __inline unsigned long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtsd_u64 (__m128d __A)
+{
+ return (unsigned long long) __builtin_ia32_vcvtsd2usi64 ((__v2df)
+ __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline unsigned long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttsd_u64 (__m128d __A)
+{
+ return (unsigned long long) __builtin_ia32_vcvttsd2usi64 ((__v2df)
+ __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline long long
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttsd_i64 (__m128d __A)
+{
+ return (long long) __builtin_ia32_vcvttsd2si64 ((__v2df) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+#endif /* __x86_64__ */
+
+extern __inline unsigned
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvtsd_u32 (__m128d __A)
+{
+ return (unsigned) __builtin_ia32_vcvtsd2usi32 ((__v2df) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline unsigned
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttsd_u32 (__m128d __A)
+{
+ return (unsigned) __builtin_ia32_vcvttsd2usi32 ((__v2df) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline int
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cvttsd_i32 (__m128d __A)
+{
+ return (int) __builtin_ia32_vcvttsd2si32 ((__v2df) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtps_pd (__m256 __A)
+{
+ return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtps_pd (__m512d __W, __mmask8 __U, __m256 __A)
+{
+ return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtps_pd (__mmask8 __U, __m256 __A)
+{
+ return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtph_ps (__m256i __A)
+{
+ return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtph_ps (__m512 __W, __mmask16 __U, __m256i __A)
+{
+ return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtph_ps (__mmask16 __U, __m256i __A)
+{
+ return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cvtpd_ps (__m512d __A)
+{
+ return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
+ (__v8sf)
+ _mm256_undefined_ps (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_cvtpd_ps (__m256 __W, __mmask8 __U, __m512d __A)
+{
+ return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
+ (__v8sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m256
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_cvtpd_ps (__mmask8 __U, __m512d __A)
+{
+ return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
+ (__v8sf)
+ _mm256_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+#ifdef __OPTIMIZE__
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_getexp_ps (__m512 __A)
+{
+ return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_getexp_ps (__m512 __W, __mmask16 __U, __m512 __A)
+{
+ return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
+ (__v16sf) __W,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_getexp_ps (__mmask16 __U, __m512 __A)
+{
+ return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_getexp_pd (__m512d __A)
+{
+ return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
+ (__v8df)
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_getexp_pd (__m512d __W, __mmask8 __U, __m512d __A)
+{
+ return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
+ (__v8df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_getexp_pd (__mmask8 __U, __m512d __A)
+{
+ return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_getexp_ss (__m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_getexpss128_round ((__v4sf) __A,
+ (__v4sf) __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getexp_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_getexpss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getexp_ss (__mmask8 __U, __m128 __A, __m128 __B)
+{
+ return (__m128) __builtin_ia32_getexpss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_getexp_sd (__m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_getexpsd128_round ((__v2df) __A,
+ (__v2df) __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getexp_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_getexpsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df) __W,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getexp_sd (__mmask8 __U, __m128d __A, __m128d __B)
+{
+ return (__m128d) __builtin_ia32_getexpsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_getmant_pd (__m512d __A, _MM_MANTISSA_NORM_ENUM __B,
+ _MM_MANTISSA_SIGN_ENUM __C)
+{
+ return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
+ (__C << 2) | __B,
+ _mm512_undefined_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_getmant_pd (__m512d __W, __mmask8 __U, __m512d __A,
+ _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+{
+ return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
+ (__C << 2) | __B,
+ (__v8df) __W, __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_getmant_pd (__mmask8 __U, __m512d __A,
+ _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+{
+ return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
+ (__C << 2) | __B,
+ (__v8df)
+ _mm512_setzero_pd (),
+ __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_getmant_ps (__m512 __A, _MM_MANTISSA_NORM_ENUM __B,
+ _MM_MANTISSA_SIGN_ENUM __C)
+{
+ return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
+ (__C << 2) | __B,
+ _mm512_undefined_ps (),
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fmsubadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm512_mask_getmant_ps (__m512 __W, __mmask16 __U, __m512 __A,
+ _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
{
- return (__m512) __builtin_ia32_vfmaddsubps512_maskz ((__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
+ (__C << 2) | __B,
+ (__v16sf) __W, __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fnmadd_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm512_maskz_getmant_ps (__mmask16 __U, __m512 __A,
+ _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) -1,
+ return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
+ (__C << 2) | __B,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fnmadd_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm_getmant_sd (__m128d __A, __m128d __B, _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D)
{
- return (__m512d) __builtin_ia32_vfnmaddpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
+ return (__m128d) __builtin_ia32_getmantsd_round ((__v2df) __A,
+ (__v2df) __B,
+ (__D << 2) | __C,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getmant_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B,
+ _MM_MANTISSA_NORM_ENUM __C, _MM_MANTISSA_SIGN_ENUM __D)
+{
+ return (__m128d) __builtin_ia32_getmantsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__D << 2) | __C,
+ (__v2df) __W,
+ __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fnmadd_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm_maskz_getmant_sd (__mmask8 __U, __m128d __A, __m128d __B,
+ _MM_MANTISSA_NORM_ENUM __C, _MM_MANTISSA_SIGN_ENUM __D)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask3 (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
+ return (__m128d) __builtin_ia32_getmantsd_mask_round ((__v2df) __A,
+ (__v2df) __B,
+ (__D << 2) | __C,
+ (__v2df)
+ _mm_setzero_pd(),
+ __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fnmadd_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm_getmant_ss (__m128 __A, __m128 __B, _MM_MANTISSA_NORM_ENUM __C,
+ _MM_MANTISSA_SIGN_ENUM __D)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
+ return (__m128) __builtin_ia32_getmantss_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__D << 2) | __C,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_getmant_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B,
+ _MM_MANTISSA_NORM_ENUM __C, _MM_MANTISSA_SIGN_ENUM __D)
+{
+ return (__m128) __builtin_ia32_getmantss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__D << 2) | __C,
+ (__v4sf) __W,
+ __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_maskz_getmant_ss (__mmask8 __U, __m128 __A, __m128 __B,
+ _MM_MANTISSA_NORM_ENUM __C, _MM_MANTISSA_SIGN_ENUM __D)
+{
+ return (__m128) __builtin_ia32_getmantss_mask_round ((__v4sf) __A,
+ (__v4sf) __B,
+ (__D << 2) | __C,
+ (__v4sf)
+ _mm_setzero_ps(),
+ __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+#else
+#define _mm512_getmant_pd(X, B, C) \
+ ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v8df)_mm512_undefined_pd(), \
+ (__mmask8)-1,\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_getmant_pd(W, U, X, B, C) \
+ ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v8df)(__m512d)(W), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_getmant_pd(U, X, B, C) \
+ ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v8df)_mm512_setzero_pd(), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm512_getmant_ps(X, B, C) \
+ ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v16sf)_mm512_undefined_ps(), \
+ (__mmask16)-1,\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_getmant_ps(W, U, X, B, C) \
+ ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v16sf)(__m512)(W), \
+ (__mmask16)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_getmant_ps(U, X, B, C) \
+ ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
+ (int)(((C)<<2) | (B)), \
+ (__v16sf)_mm512_setzero_ps(), \
+ (__mmask16)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_getmant_sd(X, Y, C, D) \
+ ((__m128d)__builtin_ia32_getmantsd_round ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), \
+ (int)(((D)<<2) | (C)), \
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_getmant_sd(W, U, X, Y, C, D) \
+ ((__m128d)__builtin_ia32_getmantsd_mask_round ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v2df)(__m128d)(W), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_maskz_getmant_sd(U, X, Y, C, D) \
+ ((__m128d)__builtin_ia32_getmantsd_mask_round ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v2df)_mm_setzero_pd(), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_getmant_ss(X, Y, C, D) \
+ ((__m128)__builtin_ia32_getmantss_round ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), \
+ (int)(((D)<<2) | (C)), \
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_getmant_ss(W, U, X, Y, C, D) \
+ ((__m128)__builtin_ia32_getmantss_mask_round ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v4sf)(__m128)(W), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_maskz_getmant_ss(U, X, Y, C, D) \
+ ((__m128)__builtin_ia32_getmantss_mask_round ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), \
+ (int)(((D)<<2) | (C)), \
+ (__v4sf)_mm_setzero_ps(), \
+ (__mmask8)(U),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_getexp_ss(A, B) \
+ ((__m128)__builtin_ia32_getexpss128_round((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), \
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_getexp_ss(W, U, A, B) \
+ (__m128)__builtin_ia32_getexpss_mask_round(A, B, W, U,\
+ _MM_FROUND_CUR_DIRECTION)
+
+#define _mm_maskz_getexp_ss(U, A, B) \
+ (__m128)__builtin_ia32_getexpss_mask_round(A, B, (__v4sf)_mm_setzero_ps(), U,\
+ _MM_FROUND_CUR_DIRECTION)
+
+#define _mm_getexp_sd(A, B) \
+ ((__m128d)__builtin_ia32_getexpsd128_round((__v2df)(__m128d)(A), (__v2df)(__m128d)(B),\
+ _MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_getexp_sd(W, U, A, B) \
+ (__m128d)__builtin_ia32_getexpsd_mask_round(A, B, W, U,\
+ _MM_FROUND_CUR_DIRECTION)
+
+#define _mm_maskz_getexp_sd(U, A, B) \
+ (__m128d)__builtin_ia32_getexpsd_mask_round(A, B, (__v2df)_mm_setzero_pd(), U,\
+ _MM_FROUND_CUR_DIRECTION)
+
+#define _mm512_getexp_ps(A) \
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
+ (__v16sf)_mm512_undefined_ps(), (__mmask16)-1, _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_getexp_ps(W, U, A) \
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
+ (__v16sf)(__m512)(W), (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_getexp_ps(U, A) \
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
+ (__v16sf)_mm512_setzero_ps(), (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_getexp_pd(A) \
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
+ (__v8df)_mm512_undefined_pd(), (__mmask8)-1, _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_getexp_pd(W, U, A) \
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
+ (__v8df)(__m512d)(W), (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_maskz_getexp_pd(U, A) \
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
+ (__v8df)_mm512_setzero_pd(), (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+#endif
+
+#ifdef __OPTIMIZE__
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_roundscale_ps (__m512 __A, const int __imm)
+{
+ return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A, __imm,
+ (__v16sf)
+ _mm512_undefined_ps (),
+ -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_roundscale_ps (__m512 __A, __mmask16 __B, __m512 __C,
+ const int __imm)
+{
+ return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __C, __imm,
+ (__v16sf) __A,
+ (__mmask16) __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_roundscale_ps (__mmask16 __A, __m512 __B, const int __imm)
+{
+ return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __B,
+ __imm,
+ (__v16sf)
+ _mm512_setzero_ps (),
+ (__mmask16) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_roundscale_pd (__m512d __A, const int __imm)
+{
+ return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A, __imm,
+ (__v8df)
+ _mm512_undefined_pd (),
+ -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_roundscale_pd (__m512d __A, __mmask8 __B, __m512d __C,
+ const int __imm)
+{
+ return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __C, __imm,
+ (__v8df) __A,
+ (__mmask8) __B,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m512d
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_maskz_roundscale_pd (__mmask8 __A, __m512d __B, const int __imm)
+{
+ return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __B,
+ __imm,
+ (__v8df)
+ _mm512_setzero_pd (),
+ (__mmask8) __A,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fnmadd_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm_roundscale_ss (__m128 __A, __m128 __B, const int __imm)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __A,
+ (__v4sf) __B, __imm,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fnmadd_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm_mask_roundscale_ss (__m128 __A, __mmask8 __B, __m128 __C, __m128 __D,
+ const int __imm)
{
- return (__m512) __builtin_ia32_vfnmaddps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __C,
+ (__v4sf) __D, __imm,
+ (__v4sf) __A,
+ (__mmask8) __B,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fnmadd_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+_mm_maskz_roundscale_ss (__mmask8 __A, __m128 __B, __m128 __C,
+ const int __imm)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask3 (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128)
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) __B,
+ (__v4sf) __C, __imm,
+ (__v4sf)
+ _mm_setzero_ps (),
+ (__mmask8) __A,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fnmadd_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm_roundscale_sd (__m128d __A, __m128d __B, const int __imm)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __A,
+ (__v2df) __B, __imm,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fnmsub_pd (__m512d __A, __m512d __B, __m512d __C)
+_mm_mask_roundscale_sd (__m128d __A, __mmask8 __B, __m128d __C, __m128d __D,
+ const int __imm)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_mask (-(__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __C,
+ (__v2df) __D, __imm,
+ (__v2df) __A,
+ (__mmask8) __B,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fnmsub_pd (__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
+_mm_maskz_roundscale_sd (__mmask8 __A, __m128d __B, __m128d __C,
+ const int __imm)
{
- return (__m512d) __builtin_ia32_vfnmsubpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d)
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) __B,
+ (__v2df) __C, __imm,
+ (__v2df)
+ _mm_setzero_pd (),
+ (__mmask8) __A,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+#else
+#define _mm512_roundscale_ps(A, B) \
+ ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(A), (int)(B),\
+ (__v16sf)_mm512_undefined_ps(), (__mmask16)(-1), _MM_FROUND_CUR_DIRECTION))
+#define _mm512_mask_roundscale_ps(A, B, C, D) \
+ ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(C), \
+ (int)(D), \
+ (__v16sf)(__m512)(A), \
+ (__mmask16)(B), _MM_FROUND_CUR_DIRECTION))
+#define _mm512_maskz_roundscale_ps(A, B, C) \
+ ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(B), \
+ (int)(C), \
+ (__v16sf)_mm512_setzero_ps(),\
+ (__mmask16)(A), _MM_FROUND_CUR_DIRECTION))
+#define _mm512_roundscale_pd(A, B) \
+ ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(A), (int)(B),\
+ (__v8df)_mm512_undefined_pd(), (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
+#define _mm512_mask_roundscale_pd(A, B, C, D) \
+ ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(C), \
+ (int)(D), \
+ (__v8df)(__m512d)(A), \
+ (__mmask8)(B), _MM_FROUND_CUR_DIRECTION))
+#define _mm512_maskz_roundscale_pd(A, B, C) \
+ ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(B), \
+ (int)(C), \
+ (__v8df)_mm512_setzero_pd(),\
+ (__mmask8)(A), _MM_FROUND_CUR_DIRECTION))
+#define _mm_roundscale_ss(A, B, I) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (A), \
+ (__v4sf) (__m128) (B), \
+ (int) (I), \
+ (__v4sf) _mm_setzero_ps (), \
+ (__mmask8) (-1), \
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_mask_roundscale_ss(A, U, B, C, I) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (B), \
+ (__v4sf) (__m128) (C), \
+ (int) (I), \
+ (__v4sf) (__m128) (A), \
+ (__mmask8) (U), \
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_maskz_roundscale_ss(U, A, B, I) \
+ ((__m128) \
+ __builtin_ia32_rndscaless_mask_round ((__v4sf) (__m128) (A), \
+ (__v4sf) (__m128) (B), \
+ (int) (I), \
+ (__v4sf) _mm_setzero_ps (), \
+ (__mmask8) (U), \
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_roundscale_sd(A, B, I) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (A), \
+ (__v2df) (__m128d) (B), \
+ (int) (I), \
+ (__v2df) _mm_setzero_pd (), \
+ (__mmask8) (-1), \
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_mask_roundscale_sd(A, U, B, C, I) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (B), \
+ (__v2df) (__m128d) (C), \
+ (int) (I), \
+ (__v2df) (__m128d) (A), \
+ (__mmask8) (U), \
+ _MM_FROUND_CUR_DIRECTION))
+#define _mm_maskz_roundscale_sd(U, A, B, I) \
+ ((__m128d) \
+ __builtin_ia32_rndscalesd_mask_round ((__v2df) (__m128d) (A), \
+ (__v2df) (__m128d) (B), \
+ (int) (I), \
+ (__v2df) _mm_setzero_pd (), \
+ (__mmask8) (U), \
+ _MM_FROUND_CUR_DIRECTION))
+#endif
+
+#ifdef __OPTIMIZE__
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fnmsub_pd (__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
+_mm512_cmp_pd_mask (__m512d __X, __m512d __Y, const int __P)
{
- return (__m512d) __builtin_ia32_vfnmsubpd512_mask3 ((__v8df) __A,
- (__v8df) __B,
- (__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, __P,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512d
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fnmsub_pd (__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
+_mm512_cmp_ps_mask (__m512 __X, __m512 __Y, const int __P)
{
- return (__m512d) __builtin_ia32_vfmaddpd512_maskz (-(__v8df) __A,
- (__v8df) __B,
- -(__v8df) __C,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, __P,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fnmsub_ps (__m512 __A, __m512 __B, __m512 __C)
+_mm512_mask_cmp_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y, const int __P)
{
- return (__m512) __builtin_ia32_vfmaddps512_mask (-(__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) -1,
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, __P,
+ (__mmask16) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fnmsub_ps (__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
+_mm512_mask_cmp_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y, const int __P)
{
- return (__m512) __builtin_ia32_vfnmsubps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, __P,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask3_fnmsub_ps (__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
+_mm512_cmpeq_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m512) __builtin_ia32_vfnmsubps512_mask3 ((__v16sf) __A,
- (__v16sf) __B,
- (__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_EQ_OQ,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fnmsub_ps (__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
+_mm512_mask_cmpeq_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m512) __builtin_ia32_vfmaddps512_maskz (-(__v16sf) __A,
- (__v16sf) __B,
- -(__v16sf) __C,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_EQ_OQ,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvttpd_epi32 (__m512d __A)
+_mm512_cmplt_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_undefined_si256 (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_LT_OS,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvttpd_epi32 (__m256i __W, __mmask8 __U, __m512d __A)
+_mm512_mask_cmplt_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
- (__v8si) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_LT_OS,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvttpd_epi32 (__mmask8 __U, __m512d __A)
+_mm512_cmple_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2dq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_setzero_si256 (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_LE_OS,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvttpd_epu32 (__m512d __A)
+_mm512_mask_cmple_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_undefined_si256 (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_LE_OS,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvttpd_epu32 (__m256i __W, __mmask8 __U, __m512d __A)
+_mm512_cmpunord_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
- (__v8si) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_UNORD_Q,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvttpd_epu32 (__mmask8 __U, __m512d __A)
+_mm512_mask_cmpunord_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvttpd2udq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_setzero_si256 (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_UNORD_Q,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtpd_epi32 (__m512d __A)
+_mm512_cmpneq_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_undefined_si256 (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NEQ_UQ,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtpd_epi32 (__m256i __W, __mmask8 __U, __m512d __A)
+_mm512_mask_cmpneq_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
- (__v8si) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NEQ_UQ,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtpd_epi32 (__mmask8 __U, __m512d __A)
+_mm512_cmpnlt_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2dq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_setzero_si256 (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NLT_US,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtpd_epu32 (__m512d __A)
+_mm512_mask_cmpnlt_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_undefined_si256 (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NLT_US,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_cmpnle_pd_mask (__m512d __X, __m512d __Y)
+{
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NLE_US,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtpd_epu32 (__m256i __W, __mmask8 __U, __m512d __A)
+_mm512_mask_cmpnle_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
- (__v8si) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_NLE_US,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m256i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtpd_epu32 (__mmask8 __U, __m512d __A)
+_mm512_cmpord_pd_mask (__m512d __X, __m512d __Y)
{
- return (__m256i) __builtin_ia32_cvtpd2udq512_mask ((__v8df) __A,
- (__v8si)
- _mm256_setzero_si256 (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_ORD_Q,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvttps_epi32 (__m512 __A)
+_mm512_mask_cmpord_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y)
{
- return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
+ (__v8df) __Y, _CMP_ORD_Q,
+ (__mmask8) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvttps_epi32 (__m512i __W, __mmask16 __U, __m512 __A)
+_mm512_cmpeq_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
- (__v16si) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_EQ_OQ,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvttps_epi32 (__mmask16 __U, __m512 __A)
+_mm512_mask_cmpeq_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvttps2dq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_EQ_OQ,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvttps_epu32 (__m512 __A)
+_mm512_cmplt_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_LT_OS,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvttps_epu32 (__m512i __W, __mmask16 __U, __m512 __A)
+_mm512_mask_cmplt_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
- (__v16si) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_LT_OS,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvttps_epu32 (__mmask16 __U, __m512 __A)
+_mm512_cmple_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvttps2udq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_LE_OS,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtps_epi32 (__m512 __A)
+_mm512_mask_cmple_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_LE_OS,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtps_epi32 (__m512i __W, __mmask16 __U, __m512 __A)
+_mm512_cmpunord_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
- (__v16si) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_UNORD_Q,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtps_epi32 (__mmask16 __U, __m512 __A)
+_mm512_mask_cmpunord_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2dq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_UNORD_Q,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtps_epu32 (__m512 __A)
+_mm512_cmpneq_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_undefined_si512 (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NEQ_UQ,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtps_epu32 (__m512i __W, __mmask16 __U, __m512 __A)
+_mm512_mask_cmpneq_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
- (__v16si) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NEQ_UQ,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512i
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtps_epu32 (__mmask16 __U, __m512 __A)
+_mm512_cmpnlt_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512i) __builtin_ia32_cvtps2udq512_mask ((__v16sf) __A,
- (__v16si)
- _mm512_setzero_si512 (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NLT_US,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-#ifdef __x86_64__
-extern __inline __m128
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtu64_ss (__m128 __A, unsigned long long __B)
+_mm512_mask_cmpnlt_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m128) __builtin_ia32_cvtusi2ss64 ((__v4sf) __A, __B,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NLT_US,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m128d
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtu64_sd (__m128d __A, unsigned long long __B)
+_mm512_cmpnle_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m128d) __builtin_ia32_cvtusi2sd64 ((__v2df) __A, __B,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NLE_US,
+ (__mmask16) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-#endif
-extern __inline __m128
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtu32_ss (__m128 __A, unsigned __B)
+_mm512_mask_cmpnle_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m128) __builtin_ia32_cvtusi2ss32 ((__v4sf) __A, __B,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_NLE_US,
+ (__mmask16) __U,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtepi32_ps (__m512i __A)
+_mm512_cmpord_ps_mask (__m512 __X, __m512 __Y)
{
- return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
- (__v16sf)
- _mm512_undefined_ps (),
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_ORD_Q,
(__mmask16) -1,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtepi32_ps (__m512 __W, __mmask16 __U, __m512i __A)
+_mm512_mask_cmpord_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y)
{
- return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
- (__v16sf) __W,
+ return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
+ (__v16sf) __Y, _CMP_ORD_Q,
(__mmask16) __U,
_MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtepi32_ps (__mmask16 __U, __m512i __A)
+_mm_cmp_sd_mask (__m128d __X, __m128d __Y, const int __P)
{
- return (__m512) __builtin_ia32_cvtdq2ps512_mask ((__v16si) __A,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmpsd_mask ((__v2df) __X,
+ (__v2df) __Y, __P,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtepu32_ps (__m512i __A)
+_mm_mask_cmp_sd_mask (__mmask8 __M, __m128d __X, __m128d __Y, const int __P)
{
- return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_cmpsd_mask ((__v2df) __X,
+ (__v2df) __Y, __P,
+ (__mmask8) __M,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_cmp_ss_mask (__m128 __X, __m128 __Y, const int __P)
+{
+ return (__mmask8) __builtin_ia32_cmpss_mask ((__v4sf) __X,
+ (__v4sf) __Y, __P,
+ (__mmask8) -1,
+ _MM_FROUND_CUR_DIRECTION);
+}
+
+extern __inline __mmask8
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_mask_cmp_ss_mask (__mmask8 __M, __m128 __X, __m128 __Y, const int __P)
+{
+ return (__mmask8) __builtin_ia32_cmpss_mask ((__v4sf) __X,
+ (__v4sf) __Y, __P,
+ (__mmask8) __M,
+ _MM_FROUND_CUR_DIRECTION);
}
-extern __inline __m512
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtepu32_ps (__m512 __W, __mmask16 __U, __m512i __A)
-{
- return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
-}
+#else
+#define _mm512_cmp_pd_mask(X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
+ (__v8df)(__m512d)(Y), (int)(P),\
+ (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_cmp_ps_mask(X, Y, P) \
+ ((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
+ (__v16sf)(__m512)(Y), (int)(P),\
+ (__mmask16)-1,_MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_cmp_pd_mask(M, X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
+ (__v8df)(__m512d)(Y), (int)(P),\
+ (__mmask8)M, _MM_FROUND_CUR_DIRECTION))
+
+#define _mm512_mask_cmp_ps_mask(M, X, Y, P) \
+ ((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
+ (__v16sf)(__m512)(Y), (int)(P),\
+ (__mmask16)M,_MM_FROUND_CUR_DIRECTION))
+
+#define _mm_cmp_sd_mask(X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), (int)(P),\
+ (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
+
+#define _mm_mask_cmp_sd_mask(M, X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
+ (__v2df)(__m128d)(Y), (int)(P),\
+ M,_MM_FROUND_CUR_DIRECTION))
+
+#define _mm_cmp_ss_mask(X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), (int)(P), \
+ (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
-extern __inline __m512
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtepu32_ps (__mmask16 __U, __m512i __A)
-{
- return (__m512) __builtin_ia32_cvtudq2ps512_mask ((__v16si) __A,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
-}
+#define _mm_mask_cmp_ss_mask(M, X, Y, P) \
+ ((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
+ (__v4sf)(__m128)(Y), (int)(P), \
+ M,_MM_FROUND_CUR_DIRECTION))
+#endif
-#ifdef __OPTIMIZE__
-extern __inline __m512d
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fixupimm_pd (__m512d __A, __m512d __B, __m512i __C, const int __imm)
+_mm512_kmov (__mmask16 __A)
{
- return (__m512d) __builtin_ia32_fixupimmpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8di) __C,
- __imm,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return __builtin_ia32_kmovw (__A);
}
-extern __inline __m512d
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fixupimm_pd (__m512d __A, __mmask8 __U, __m512d __B,
- __m512i __C, const int __imm)
+_mm512_castpd_ps (__m512d __A)
{
- return (__m512d) __builtin_ia32_fixupimmpd512_mask ((__v8df) __A,
- (__v8df) __B,
- (__v8di) __C,
- __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) (__A);
}
-extern __inline __m512d
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fixupimm_pd (__mmask8 __U, __m512d __A, __m512d __B,
- __m512i __C, const int __imm)
+_mm512_castpd_si512 (__m512d __A)
{
- return (__m512d) __builtin_ia32_fixupimmpd512_maskz ((__v8df) __A,
- (__v8df) __B,
- (__v8di) __C,
- __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512i) (__A);
}
-extern __inline __m512
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_fixupimm_ps (__m512 __A, __m512 __B, __m512i __C, const int __imm)
+_mm512_castps_pd (__m512 __A)
{
- return (__m512) __builtin_ia32_fixupimmps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16si) __C,
- __imm,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) (__A);
}
-extern __inline __m512
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_fixupimm_ps (__m512 __A, __mmask16 __U, __m512 __B,
- __m512i __C, const int __imm)
+_mm512_castps_si512 (__m512 __A)
{
- return (__m512) __builtin_ia32_fixupimmps512_mask ((__v16sf) __A,
- (__v16sf) __B,
- (__v16si) __C,
- __imm,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512i) (__A);
}
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_fixupimm_ps (__mmask16 __U, __m512 __A, __m512 __B,
- __m512i __C, const int __imm)
+_mm512_castsi512_ps (__m512i __A)
{
- return (__m512) __builtin_ia32_fixupimmps512_maskz ((__v16sf) __A,
- (__v16sf) __B,
- (__v16si) __C,
- __imm,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) (__A);
}
-extern __inline __m128d
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_fixupimm_sd (__m128d __A, __m128d __B, __m128i __C, const int __imm)
+_mm512_castsi512_pd (__m512i __A)
{
- return (__m128d) __builtin_ia32_fixupimmsd_mask ((__v2df) __A,
- (__v2df) __B,
- (__v2di) __C, __imm,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) (__A);
}
extern __inline __m128d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_mask_fixupimm_sd (__m128d __A, __mmask8 __U, __m128d __B,
- __m128i __C, const int __imm)
+_mm512_castpd512_pd128 (__m512d __A)
{
- return (__m128d) __builtin_ia32_fixupimmsd_mask ((__v2df) __A,
- (__v2df) __B,
- (__v2di) __C, __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128d)_mm512_extractf32x4_ps((__m512)__A, 0);
}
-extern __inline __m128d
+extern __inline __m128
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_maskz_fixupimm_sd (__mmask8 __U, __m128d __A, __m128d __B,
- __m128i __C, const int __imm)
+_mm512_castps512_ps128 (__m512 __A)
{
- return (__m128d) __builtin_ia32_fixupimmsd_maskz ((__v2df) __A,
- (__v2df) __B,
- (__v2di) __C,
- __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_extractf32x4_ps(__A, 0);
}
-extern __inline __m128
+extern __inline __m128i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_fixupimm_ss (__m128 __A, __m128 __B, __m128i __C, const int __imm)
+_mm512_castsi512_si128 (__m512i __A)
{
- return (__m128) __builtin_ia32_fixupimmss_mask ((__v4sf) __A,
- (__v4sf) __B,
- (__v4si) __C, __imm,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m128i)_mm512_extracti32x4_epi32((__m512i)__A, 0);
}
-extern __inline __m128
+extern __inline __m256d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_mask_fixupimm_ss (__m128 __A, __mmask8 __U, __m128 __B,
- __m128i __C, const int __imm)
+_mm512_castpd512_pd256 (__m512d __A)
{
- return (__m128) __builtin_ia32_fixupimmss_mask ((__v4sf) __A,
- (__v4sf) __B,
- (__v4si) __C, __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_extractf64x4_pd(__A, 0);
}
-extern __inline __m128
+extern __inline __m256
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_maskz_fixupimm_ss (__mmask8 __U, __m128 __A, __m128 __B,
- __m128i __C, const int __imm)
+_mm512_castps512_ps256 (__m512 __A)
{
- return (__m128) __builtin_ia32_fixupimmss_maskz ((__v4sf) __A,
- (__v4sf) __B,
- (__v4si) __C, __imm,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m256)_mm512_extractf64x4_pd((__m512d)__A, 0);
}
-#else
-#define _mm512_fixupimm_pd(X, Y, Z, C) \
- ((__m512d)__builtin_ia32_fixupimmpd512_mask ((__v8df)(__m512d)(X), \
- (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
- (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_mask_fixupimm_pd(X, U, Y, Z, C) \
- ((__m512d)__builtin_ia32_fixupimmpd512_mask ((__v8df)(__m512d)(X), \
- (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_maskz_fixupimm_pd(U, X, Y, Z, C) \
- ((__m512d)__builtin_ia32_fixupimmpd512_maskz ((__v8df)(__m512d)(X), \
- (__v8df)(__m512d)(Y), (__v8di)(__m512i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_fixupimm_ps(X, Y, Z, C) \
- ((__m512)__builtin_ia32_fixupimmps512_mask ((__v16sf)(__m512)(X), \
- (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
- (__mmask16)(-1), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_mask_fixupimm_ps(X, U, Y, Z, C) \
- ((__m512)__builtin_ia32_fixupimmps512_mask ((__v16sf)(__m512)(X), \
- (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
- (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_maskz_fixupimm_ps(U, X, Y, Z, C) \
- ((__m512)__builtin_ia32_fixupimmps512_maskz ((__v16sf)(__m512)(X), \
- (__v16sf)(__m512)(Y), (__v16si)(__m512i)(Z), (int)(C), \
- (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm_fixupimm_sd(X, Y, Z, C) \
- ((__m128d)__builtin_ia32_fixupimmsd_mask ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
- (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm_mask_fixupimm_sd(X, U, Y, Z, C) \
- ((__m128d)__builtin_ia32_fixupimmsd_mask ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm_maskz_fixupimm_sd(U, X, Y, Z, C) \
- ((__m128d)__builtin_ia32_fixupimmsd_maskz ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), (__v2di)(__m128i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm_fixupimm_ss(X, Y, Z, C) \
- ((__m128)__builtin_ia32_fixupimmss_mask ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
- (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
-
-#define _mm_mask_fixupimm_ss(X, U, Y, Z, C) \
- ((__m128)__builtin_ia32_fixupimmss_mask ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-#define _mm_maskz_fixupimm_ss(U, X, Y, Z, C) \
- ((__m128)__builtin_ia32_fixupimmss_maskz ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), (__v4si)(__m128i)(Z), (int)(C), \
- (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-#endif
-
-#ifdef __x86_64__
-extern __inline unsigned long long
+extern __inline __m256i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtss_u64 (__m128 __A)
+_mm512_castsi512_si256 (__m512i __A)
{
- return (unsigned long long) __builtin_ia32_vcvtss2usi64 ((__v4sf)
- __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m256i)_mm512_extractf64x4_pd((__m512d)__A, 0);
}
-extern __inline unsigned long long
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttss_u64 (__m128 __A)
+_mm512_castpd128_pd512 (__m128d __A)
{
- return (unsigned long long) __builtin_ia32_vcvttss2usi64 ((__v4sf)
- __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) __builtin_ia32_pd512_pd((__m128d)__A);
}
-extern __inline long long
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttss_i64 (__m128 __A)
+_mm512_castps128_ps512 (__m128 __A)
{
- return (long long) __builtin_ia32_vcvttss2si64 ((__v4sf) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) __builtin_ia32_ps512_ps((__m128)__A);
}
-#endif /* __x86_64__ */
-extern __inline unsigned
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtss_u32 (__m128 __A)
+_mm512_castsi128_si512 (__m128i __A)
{
- return (unsigned) __builtin_ia32_vcvtss2usi32 ((__v4sf) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512i) __builtin_ia32_si512_si((__v4si)__A);
}
-extern __inline unsigned
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttss_u32 (__m128 __A)
+_mm512_castpd256_pd512 (__m256d __A)
{
- return (unsigned) __builtin_ia32_vcvttss2usi32 ((__v4sf) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return __builtin_ia32_pd512_256pd (__A);
}
-extern __inline int
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttss_i32 (__m128 __A)
+_mm512_castps256_ps512 (__m256 __A)
{
- return (int) __builtin_ia32_vcvttss2si32 ((__v4sf) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return __builtin_ia32_ps512_256ps (__A);
}
-#ifdef __x86_64__
-extern __inline unsigned long long
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtsd_u64 (__m128d __A)
+_mm512_castsi256_si512 (__m256i __A)
{
- return (unsigned long long) __builtin_ia32_vcvtsd2usi64 ((__v2df)
- __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512i)__builtin_ia32_si512_256si ((__v8si)__A);
}
-extern __inline unsigned long long
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttsd_u64 (__m128d __A)
+_mm512_zextpd128_pd512 (__m128d __A)
{
- return (unsigned long long) __builtin_ia32_vcvttsd2usi64 ((__v2df)
- __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512d) _mm512_insertf32x4 (_mm512_setzero_ps (), (__m128) __A, 0);
}
-extern __inline long long
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttsd_i64 (__m128d __A)
+_mm512_zextps128_ps512 (__m128 __A)
{
- return (long long) __builtin_ia32_vcvttsd2si64 ((__v2df) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_insertf32x4 (_mm512_setzero_ps (), __A, 0);
}
-#endif /* __x86_64__ */
-extern __inline unsigned
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvtsd_u32 (__m128d __A)
+_mm512_zextsi128_si512 (__m128i __A)
{
- return (unsigned) __builtin_ia32_vcvtsd2usi32 ((__v2df) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_inserti32x4 (_mm512_setzero_si512 (), __A, 0);
}
-extern __inline unsigned
+extern __inline __m512d
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttsd_u32 (__m128d __A)
+_mm512_zextpd256_pd512 (__m256d __A)
{
- return (unsigned) __builtin_ia32_vcvttsd2usi32 ((__v2df) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_insertf64x4 (_mm512_setzero_pd (), __A, 0);
}
-extern __inline int
+extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cvttsd_i32 (__m128d __A)
+_mm512_zextps256_ps512 (__m256 __A)
{
- return (int) __builtin_ia32_vcvttsd2si32 ((__v2df) __A,
- _MM_FROUND_CUR_DIRECTION);
+ return (__m512) _mm512_insertf64x4 (_mm512_setzero_pd (), (__m256d) __A, 0);
}
-extern __inline __m512d
+extern __inline __m512i
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtps_pd (__m256 __A)
+_mm512_zextsi256_si512 (__m256i __A)
{
- return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return _mm512_inserti64x4 (_mm512_setzero_si512 (), __A, 0);
}
-extern __inline __m512d
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtps_pd (__m512d __W, __mmask8 __U, __m256 __A)
+_mm512_cmpeq_epu32_mask (__m512i __A, __m512i __B)
{
- return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __A,
+ (__v16si) __B, 0,
+ (__mmask16) -1);
}
-extern __inline __m512d
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtps_pd (__mmask8 __U, __m256 __A)
+_mm512_mask_cmpeq_epu32_mask (__mmask16 __U, __m512i __A, __m512i __B)
{
- return (__m512d) __builtin_ia32_cvtps2pd512_mask ((__v8sf) __A,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __A,
+ (__v16si) __B, 0, __U);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtph_ps (__m256i __A)
+_mm512_mask_cmpeq_epu64_mask (__mmask8 __U, __m512i __A, __m512i __B)
{
- return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __A,
+ (__v8di) __B, 0, __U);
}
-extern __inline __m512
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtph_ps (__m512 __W, __mmask16 __U, __m256i __A)
+_mm512_cmpeq_epu64_mask (__m512i __A, __m512i __B)
{
- return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __A,
+ (__v8di) __B, 0,
+ (__mmask8) -1);
}
-extern __inline __m512
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtph_ps (__mmask16 __U, __m256i __A)
+_mm512_cmpgt_epu32_mask (__m512i __A, __m512i __B)
{
- return (__m512) __builtin_ia32_vcvtph2ps512_mask ((__v16hi) __A,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __A,
+ (__v16si) __B, 6,
+ (__mmask16) -1);
}
-extern __inline __m256
+extern __inline __mmask16
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cvtpd_ps (__m512d __A)
+_mm512_mask_cmpgt_epu32_mask (__mmask16 __U, __m512i __A, __m512i __B)
{
- return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
- (__v8sf)
- _mm256_undefined_ps (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask16) __builtin_ia32_ucmpd512_mask ((__v16si) __A,
+ (__v16si) __B, 6, __U);
}
-extern __inline __m256
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cvtpd_ps (__m256 __W, __mmask8 __U, __m512d __A)
+_mm512_mask_cmpgt_epu64_mask (__mmask8 __U, __m512i __A, __m512i __B)
{
- return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
- (__v8sf) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __A,
+ (__v8di) __B, 6, __U);
}
-extern __inline __m256
+extern __inline __mmask8
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_cvtpd_ps (__mmask8 __U, __m512d __A)
+_mm512_cmpgt_epu64_mask (__m512i __A, __m512i __B)
{
- return (__m256) __builtin_ia32_cvtpd2ps512_mask ((__v8df) __A,
- (__v8sf)
- _mm256_setzero_ps (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ return (__mmask8) __builtin_ia32_ucmpq512_mask ((__v8di) __A,
+ (__v8di) __B, 6,
+ (__mmask8) -1);
}
-#ifdef __OPTIMIZE__
-extern __inline __m512
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __v8si __T1 = (__v8si) _mm512_extracti64x4_epi64 (__A, 1); \
+ __v8si __T2 = (__v8si) _mm512_extracti64x4_epi64 (__A, 0); \
+ __m256i __T3 = (__m256i) (__T1 op __T2); \
+ __v4si __T4 = (__v4si) _mm256_extracti128_si256 (__T3, 1); \
+ __v4si __T5 = (__v4si) _mm256_extracti128_si256 (__T3, 0); \
+ __v4si __T6 = __T4 op __T5; \
+ __v4si __T7 = __builtin_shuffle (__T6, (__v4si) { 2, 3, 0, 1 }); \
+ __v4si __T8 = __T6 op __T7; \
+ return __T8[0] op __T8[1]
+
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_getexp_ps (__m512 __A)
+_mm512_reduce_add_epi32 (__m512i __A)
{
- return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
- (__v16sf)
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (+);
}
-extern __inline __m512
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_getexp_ps (__m512 __W, __mmask16 __U, __m512 __A)
+_mm512_reduce_mul_epi32 (__m512i __A)
{
- return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
- (__v16sf) __W,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (*);
}
-extern __inline __m512
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_getexp_ps (__mmask16 __U, __m512 __A)
+_mm512_reduce_and_epi32 (__m512i __A)
{
- return (__m512) __builtin_ia32_getexpps512_mask ((__v16sf) __A,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (&);
}
-extern __inline __m512d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_getexp_pd (__m512d __A)
+_mm512_reduce_or_epi32 (__m512i __A)
{
- return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
- (__v8df)
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (|);
}
-extern __inline __m512d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_getexp_pd (__m512d __W, __mmask8 __U, __m512d __A)
+_mm512_mask_reduce_add_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
- (__v8df) __W,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi32 (__U, __A);
+ __MM512_REDUCE_OP (+);
}
-extern __inline __m512d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_getexp_pd (__mmask8 __U, __m512d __A)
+_mm512_mask_reduce_mul_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m512d) __builtin_ia32_getexppd512_mask ((__v8df) __A,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi32 (_mm512_set1_epi32 (1), __U, __A);
+ __MM512_REDUCE_OP (*);
}
-extern __inline __m128
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_getexp_ss (__m128 __A, __m128 __B)
+_mm512_mask_reduce_and_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m128) __builtin_ia32_getexpss128_round ((__v4sf) __A,
- (__v4sf) __B,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi32 (_mm512_set1_epi32 (~0), __U, __A);
+ __MM512_REDUCE_OP (&);
}
-extern __inline __m128d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_getexp_sd (__m128d __A, __m128d __B)
+_mm512_mask_reduce_or_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m128d) __builtin_ia32_getexpsd128_round ((__v2df) __A,
- (__v2df) __B,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi32 (__U, __A);
+ __MM512_REDUCE_OP (|);
}
-extern __inline __m512d
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m256i __T1 = (__m256i) _mm512_extracti64x4_epi64 (__A, 1); \
+ __m256i __T2 = (__m256i) _mm512_extracti64x4_epi64 (__A, 0); \
+ __m256i __T3 = _mm256_##op (__T1, __T2); \
+ __m128i __T4 = (__m128i) _mm256_extracti128_si256 (__T3, 1); \
+ __m128i __T5 = (__m128i) _mm256_extracti128_si256 (__T3, 0); \
+ __m128i __T6 = _mm_##op (__T4, __T5); \
+ __m128i __T7 = (__m128i) __builtin_shuffle ((__v4si) __T6, \
+ (__v4si) { 2, 3, 0, 1 }); \
+ __m128i __T8 = _mm_##op (__T6, __T7); \
+ __m128i __T9 = (__m128i) __builtin_shuffle ((__v4si) __T8, \
+ (__v4si) { 1, 0, 1, 0 }); \
+ __v4si __T10 = (__v4si) _mm_##op (__T8, __T9); \
+ return __T10[0]
+
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_getmant_pd (__m512d __A, _MM_MANTISSA_NORM_ENUM __B,
- _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_reduce_min_epi32 (__m512i __A)
{
- return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
- (__C << 2) | __B,
- _mm512_undefined_pd (),
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (min_epi32);
}
-extern __inline __m512d
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_getmant_pd (__m512d __W, __mmask8 __U, __m512d __A,
- _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_reduce_max_epi32 (__m512i __A)
{
- return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
- (__C << 2) | __B,
- (__v8df) __W, __U,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (max_epi32);
}
-extern __inline __m512d
+extern __inline unsigned int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_getmant_pd (__mmask8 __U, __m512d __A,
- _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_reduce_min_epu32 (__m512i __A)
{
- return (__m512d) __builtin_ia32_getmantpd512_mask ((__v8df) __A,
- (__C << 2) | __B,
- (__v8df)
- _mm512_setzero_pd (),
- __U,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (min_epu32);
}
-extern __inline __m512
+extern __inline unsigned int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_getmant_ps (__m512 __A, _MM_MANTISSA_NORM_ENUM __B,
- _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_reduce_max_epu32 (__m512i __A)
{
- return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
- (__C << 2) | __B,
- _mm512_undefined_ps (),
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (max_epu32);
}
-extern __inline __m512
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_getmant_ps (__m512 __W, __mmask16 __U, __m512 __A,
- _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_mask_reduce_min_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
- (__C << 2) | __B,
- (__v16sf) __W, __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi32 (_mm512_set1_epi32 (__INT_MAX__), __U, __A);
+ __MM512_REDUCE_OP (min_epi32);
}
-extern __inline __m512
+extern __inline int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_getmant_ps (__mmask16 __U, __m512 __A,
- _MM_MANTISSA_NORM_ENUM __B, _MM_MANTISSA_SIGN_ENUM __C)
+_mm512_mask_reduce_max_epi32 (__mmask16 __U, __m512i __A)
{
- return (__m512) __builtin_ia32_getmantps512_mask ((__v16sf) __A,
- (__C << 2) | __B,
- (__v16sf)
- _mm512_setzero_ps (),
- __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi32 (_mm512_set1_epi32 (-__INT_MAX__ - 1), __U, __A);
+ __MM512_REDUCE_OP (max_epi32);
}
-extern __inline __m128d
+extern __inline unsigned int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_getmant_sd (__m128d __A, __m128d __B, _MM_MANTISSA_NORM_ENUM __C,
- _MM_MANTISSA_SIGN_ENUM __D)
+_mm512_mask_reduce_min_epu32 (__mmask16 __U, __m512i __A)
{
- return (__m128d) __builtin_ia32_getmantsd_round ((__v2df) __A,
- (__v2df) __B,
- (__D << 2) | __C,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi32 (_mm512_set1_epi32 (~0), __U, __A);
+ __MM512_REDUCE_OP (min_epu32);
}
-extern __inline __m128
+extern __inline unsigned int
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_getmant_ss (__m128 __A, __m128 __B, _MM_MANTISSA_NORM_ENUM __C,
- _MM_MANTISSA_SIGN_ENUM __D)
+_mm512_mask_reduce_max_epu32 (__mmask16 __U, __m512i __A)
{
- return (__m128) __builtin_ia32_getmantss_round ((__v4sf) __A,
- (__v4sf) __B,
- (__D << 2) | __C,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi32 (__U, __A);
+ __MM512_REDUCE_OP (max_epu32);
}
-#else
-#define _mm512_getmant_pd(X, B, C) \
- ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
- (int)(((C)<<2) | (B)), \
- (__v8df)_mm512_undefined_pd(), \
- (__mmask8)-1,\
- _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_mask_getmant_pd(W, U, X, B, C) \
- ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
- (int)(((C)<<2) | (B)), \
- (__v8df)(__m512d)(W), \
- (__mmask8)(U),\
- _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_maskz_getmant_pd(U, X, B, C) \
- ((__m512d)__builtin_ia32_getmantpd512_mask ((__v8df)(__m512d)(X), \
- (int)(((C)<<2) | (B)), \
- (__v8df)_mm512_setzero_pd(), \
- (__mmask8)(U),\
- _MM_FROUND_CUR_DIRECTION))
-#define _mm512_getmant_ps(X, B, C) \
- ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
- (int)(((C)<<2) | (B)), \
- (__v16sf)_mm512_undefined_ps(), \
- (__mmask16)-1,\
- _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_mask_getmant_ps(W, U, X, B, C) \
- ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
- (int)(((C)<<2) | (B)), \
- (__v16sf)(__m512)(W), \
- (__mmask16)(U),\
- _MM_FROUND_CUR_DIRECTION))
-
-#define _mm512_maskz_getmant_ps(U, X, B, C) \
- ((__m512)__builtin_ia32_getmantps512_mask ((__v16sf)(__m512)(X), \
- (int)(((C)<<2) | (B)), \
- (__v16sf)_mm512_setzero_ps(), \
- (__mmask16)(U),\
- _MM_FROUND_CUR_DIRECTION))
-#define _mm_getmant_sd(X, Y, C, D) \
- ((__m128d)__builtin_ia32_getmantsd_round ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), \
- (int)(((D)<<2) | (C)), \
- _MM_FROUND_CUR_DIRECTION))
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m256 __T1 = (__m256) _mm512_extractf64x4_pd ((__m512d) __A, 1); \
+ __m256 __T2 = (__m256) _mm512_extractf64x4_pd ((__m512d) __A, 0); \
+ __m256 __T3 = __T1 op __T2; \
+ __m128 __T4 = _mm256_extractf128_ps (__T3, 1); \
+ __m128 __T5 = _mm256_extractf128_ps (__T3, 0); \
+ __m128 __T6 = __T4 op __T5; \
+ __m128 __T7 = __builtin_shuffle (__T6, (__v4si) { 2, 3, 0, 1 }); \
+ __m128 __T8 = __T6 op __T7; \
+ return __T8[0] op __T8[1]
-#define _mm_getmant_ss(X, Y, C, D) \
- ((__m128)__builtin_ia32_getmantss_round ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), \
- (int)(((D)<<2) | (C)), \
- _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_add_ps (__m512 __A)
+{
+ __MM512_REDUCE_OP (+);
+}
-#define _mm_getexp_ss(A, B) \
- ((__m128)__builtin_ia32_getexpss128_mask((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), \
- _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_mul_ps (__m512 __A)
+{
+ __MM512_REDUCE_OP (*);
+}
-#define _mm_getexp_sd(A, B) \
- ((__m128d)__builtin_ia32_getexpsd128_mask((__v2df)(__m128d)(A), (__v2df)(__m128d)(B),\
- _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_add_ps (__mmask16 __U, __m512 __A)
+{
+ __A = _mm512_maskz_mov_ps (__U, __A);
+ __MM512_REDUCE_OP (+);
+}
-#define _mm512_getexp_ps(A) \
- ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
- (__v16sf)_mm512_undefined_ps(), (__mmask16)-1, _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_mul_ps (__mmask16 __U, __m512 __A)
+{
+ __A = _mm512_mask_mov_ps (_mm512_set1_ps (1.0f), __U, __A);
+ __MM512_REDUCE_OP (*);
+}
-#define _mm512_mask_getexp_ps(W, U, A) \
- ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
- (__v16sf)(__m512)(W), (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m256 __T1 = (__m256) _mm512_extractf64x4_pd ((__m512d) __A, 1); \
+ __m256 __T2 = (__m256) _mm512_extractf64x4_pd ((__m512d) __A, 0); \
+ __m256 __T3 = _mm256_##op (__T1, __T2); \
+ __m128 __T4 = _mm256_extractf128_ps (__T3, 1); \
+ __m128 __T5 = _mm256_extractf128_ps (__T3, 0); \
+ __m128 __T6 = _mm_##op (__T4, __T5); \
+ __m128 __T7 = __builtin_shuffle (__T6, (__v4si) { 2, 3, 0, 1 }); \
+ __m128 __T8 = _mm_##op (__T6, __T7); \
+ __m128 __T9 = __builtin_shuffle (__T8, (__v4si) { 1, 0, 1, 0 }); \
+ __m128 __T10 = _mm_##op (__T8, __T9); \
+ return __T10[0]
-#define _mm512_maskz_getexp_ps(U, A) \
- ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
- (__v16sf)_mm512_setzero_ps(), (__mmask16)(U), _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_min_ps (__m512 __A)
+{
+ __MM512_REDUCE_OP (min_ps);
+}
-#define _mm512_getexp_pd(A) \
- ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
- (__v8df)_mm512_undefined_pd(), (__mmask8)-1, _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_max_ps (__m512 __A)
+{
+ __MM512_REDUCE_OP (max_ps);
+}
-#define _mm512_mask_getexp_pd(W, U, A) \
- ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
- (__v8df)(__m512d)(W), (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_min_ps (__mmask16 __U, __m512 __A)
+{
+ __A = _mm512_mask_mov_ps (_mm512_set1_ps (__builtin_inff ()), __U, __A);
+ __MM512_REDUCE_OP (min_ps);
+}
-#define _mm512_maskz_getexp_pd(U, A) \
- ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
- (__v8df)_mm512_setzero_pd(), (__mmask8)(U), _MM_FROUND_CUR_DIRECTION))
-#endif
+extern __inline float
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_max_ps (__mmask16 __U, __m512 __A)
+{
+ __A = _mm512_mask_mov_ps (_mm512_set1_ps (-__builtin_inff ()), __U, __A);
+ __MM512_REDUCE_OP (max_ps);
+}
-#ifdef __OPTIMIZE__
-extern __inline __m512
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __v4di __T1 = (__v4di) _mm512_extracti64x4_epi64 (__A, 1); \
+ __v4di __T2 = (__v4di) _mm512_extracti64x4_epi64 (__A, 0); \
+ __m256i __T3 = (__m256i) (__T1 op __T2); \
+ __v2di __T4 = (__v2di) _mm256_extracti128_si256 (__T3, 1); \
+ __v2di __T5 = (__v2di) _mm256_extracti128_si256 (__T3, 0); \
+ __v2di __T6 = __T4 op __T5; \
+ return __T6[0] op __T6[1]
+
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_roundscale_ps (__m512 __A, const int __imm)
+_mm512_reduce_add_epi64 (__m512i __A)
{
- return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __A, __imm,
- (__v16sf)
- _mm512_undefined_ps (),
- -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (+);
}
-extern __inline __m512
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_roundscale_ps (__m512 __A, __mmask16 __B, __m512 __C,
- const int __imm)
+_mm512_reduce_mul_epi64 (__m512i __A)
{
- return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __C, __imm,
- (__v16sf) __A,
- (__mmask16) __B,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (*);
}
-extern __inline __m512
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_roundscale_ps (__mmask16 __A, __m512 __B, const int __imm)
+_mm512_reduce_and_epi64 (__m512i __A)
{
- return (__m512) __builtin_ia32_rndscaleps_mask ((__v16sf) __B,
- __imm,
- (__v16sf)
- _mm512_setzero_ps (),
- (__mmask16) __A,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (&);
}
-extern __inline __m512d
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_roundscale_pd (__m512d __A, const int __imm)
+_mm512_reduce_or_epi64 (__m512i __A)
{
- return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __A, __imm,
- (__v8df)
- _mm512_undefined_pd (),
- -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (|);
}
-extern __inline __m512d
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_roundscale_pd (__m512d __A, __mmask8 __B, __m512d __C,
- const int __imm)
+_mm512_mask_reduce_add_epi64 (__mmask8 __U, __m512i __A)
{
- return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __C, __imm,
- (__v8df) __A,
- (__mmask8) __B,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi64 (__U, __A);
+ __MM512_REDUCE_OP (+);
}
-extern __inline __m512d
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_maskz_roundscale_pd (__mmask8 __A, __m512d __B, const int __imm)
+_mm512_mask_reduce_mul_epi64 (__mmask8 __U, __m512i __A)
{
- return (__m512d) __builtin_ia32_rndscalepd_mask ((__v8df) __B,
- __imm,
- (__v8df)
- _mm512_setzero_pd (),
- (__mmask8) __A,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi64 (_mm512_set1_epi64 (1LL), __U, __A);
+ __MM512_REDUCE_OP (*);
}
-extern __inline __m128
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_roundscale_ss (__m128 __A, __m128 __B, const int __imm)
+_mm512_mask_reduce_and_epi64 (__mmask8 __U, __m512i __A)
{
- return (__m128) __builtin_ia32_rndscaless_round ((__v4sf) __A,
- (__v4sf) __B, __imm,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi64 (_mm512_set1_epi64 (~0LL), __U, __A);
+ __MM512_REDUCE_OP (&);
}
-extern __inline __m128d
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_roundscale_sd (__m128d __A, __m128d __B, const int __imm)
+_mm512_mask_reduce_or_epi64 (__mmask8 __U, __m512i __A)
{
- return (__m128d) __builtin_ia32_rndscalesd_round ((__v2df) __A,
- (__v2df) __B, __imm,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi64 (__U, __A);
+ __MM512_REDUCE_OP (|);
}
-#else
-#define _mm512_roundscale_ps(A, B) \
- ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(A), (int)(B),\
- (__v16sf)_mm512_undefined_ps(), (__mmask16)(-1), _MM_FROUND_CUR_DIRECTION))
-#define _mm512_mask_roundscale_ps(A, B, C, D) \
- ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(C), \
- (int)(D), \
- (__v16sf)(__m512)(A), \
- (__mmask16)(B), _MM_FROUND_CUR_DIRECTION))
-#define _mm512_maskz_roundscale_ps(A, B, C) \
- ((__m512) __builtin_ia32_rndscaleps_mask ((__v16sf)(__m512)(B), \
- (int)(C), \
- (__v16sf)_mm512_setzero_ps(),\
- (__mmask16)(A), _MM_FROUND_CUR_DIRECTION))
-#define _mm512_roundscale_pd(A, B) \
- ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(A), (int)(B),\
- (__v8df)_mm512_undefined_pd(), (__mmask8)(-1), _MM_FROUND_CUR_DIRECTION))
-#define _mm512_mask_roundscale_pd(A, B, C, D) \
- ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(C), \
- (int)(D), \
- (__v8df)(__m512d)(A), \
- (__mmask8)(B), _MM_FROUND_CUR_DIRECTION))
-#define _mm512_maskz_roundscale_pd(A, B, C) \
- ((__m512d) __builtin_ia32_rndscalepd_mask ((__v8df)(__m512d)(B), \
- (int)(C), \
- (__v8df)_mm512_setzero_pd(),\
- (__mmask8)(A), _MM_FROUND_CUR_DIRECTION))
-#define _mm_roundscale_ss(A, B, C) \
- ((__m128) __builtin_ia32_rndscaless_round ((__v4sf)(__m128)(A), \
- (__v4sf)(__m128)(B), (int)(C), _MM_FROUND_CUR_DIRECTION))
-#define _mm_roundscale_sd(A, B, C) \
- ((__m128d) __builtin_ia32_rndscalesd_round ((__v2df)(__m128d)(A), \
- (__v2df)(__m128d)(B), (int)(C), _MM_FROUND_CUR_DIRECTION))
-#endif
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m512i __T1 = _mm512_shuffle_i64x2 (__A, __A, 0x4e); \
+ __m512i __T2 = _mm512_##op (__A, __T1); \
+ __m512i __T3 \
+ = (__m512i) __builtin_shuffle ((__v8di) __T2, \
+ (__v8di) { 2, 3, 0, 1, 6, 7, 4, 5 });\
+ __m512i __T4 = _mm512_##op (__T2, __T3); \
+ __m512i __T5 \
+ = (__m512i) __builtin_shuffle ((__v8di) __T4, \
+ (__v8di) { 1, 0, 3, 2, 5, 4, 7, 6 });\
+ __v8di __T6 = (__v8di) _mm512_##op (__T4, __T5); \
+ return __T6[0]
-#ifdef __OPTIMIZE__
-extern __inline __mmask8
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cmp_pd_mask (__m512d __X, __m512d __Y, const int __P)
+_mm512_reduce_min_epi64 (__m512i __A)
{
- return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
- (__v8df) __Y, __P,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (min_epi64);
}
-extern __inline __mmask16
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_cmp_ps_mask (__m512 __X, __m512 __Y, const int __P)
+_mm512_reduce_max_epi64 (__m512i __A)
{
- return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
- (__v16sf) __Y, __P,
- (__mmask16) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (max_epi64);
}
-extern __inline __mmask16
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cmp_ps_mask (__mmask16 __U, __m512 __X, __m512 __Y, const int __P)
+_mm512_mask_reduce_min_epi64 (__mmask8 __U, __m512i __A)
{
- return (__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf) __X,
- (__v16sf) __Y, __P,
- (__mmask16) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi64 (_mm512_set1_epi64 (__LONG_LONG_MAX__),
+ __U, __A);
+ __MM512_REDUCE_OP (min_epi64);
}
-extern __inline __mmask8
+extern __inline long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_mask_cmp_pd_mask (__mmask8 __U, __m512d __X, __m512d __Y, const int __P)
+_mm512_mask_reduce_max_epi64 (__mmask8 __U, __m512i __A)
{
- return (__mmask8) __builtin_ia32_cmppd512_mask ((__v8df) __X,
- (__v8df) __Y, __P,
- (__mmask8) __U,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi64 (_mm512_set1_epi64 (-__LONG_LONG_MAX__ - 1),
+ __U, __A);
+ __MM512_REDUCE_OP (max_epi64);
}
-extern __inline __mmask8
+extern __inline unsigned long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cmp_sd_mask (__m128d __X, __m128d __Y, const int __P)
+_mm512_reduce_min_epu64 (__m512i __A)
{
- return (__mmask8) __builtin_ia32_cmpsd_mask ((__v2df) __X,
- (__v2df) __Y, __P,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (min_epu64);
}
-extern __inline __mmask8
+extern __inline unsigned long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_mask_cmp_sd_mask (__mmask8 __M, __m128d __X, __m128d __Y, const int __P)
+_mm512_reduce_max_epu64 (__m512i __A)
{
- return (__mmask8) __builtin_ia32_cmpsd_mask ((__v2df) __X,
- (__v2df) __Y, __P,
- (__mmask8) __M,
- _MM_FROUND_CUR_DIRECTION);
+ __MM512_REDUCE_OP (max_epu64);
}
-extern __inline __mmask8
+extern __inline unsigned long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_cmp_ss_mask (__m128 __X, __m128 __Y, const int __P)
+_mm512_mask_reduce_min_epu64 (__mmask8 __U, __m512i __A)
{
- return (__mmask8) __builtin_ia32_cmpss_mask ((__v4sf) __X,
- (__v4sf) __Y, __P,
- (__mmask8) -1,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_mask_mov_epi64 (_mm512_set1_epi64 (~0LL), __U, __A);
+ __MM512_REDUCE_OP (min_epu64);
}
-extern __inline __mmask8
+extern __inline unsigned long long
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm_mask_cmp_ss_mask (__mmask8 __M, __m128 __X, __m128 __Y, const int __P)
+_mm512_mask_reduce_max_epu64 (__mmask8 __U, __m512i __A)
{
- return (__mmask8) __builtin_ia32_cmpss_mask ((__v4sf) __X,
- (__v4sf) __Y, __P,
- (__mmask8) __M,
- _MM_FROUND_CUR_DIRECTION);
+ __A = _mm512_maskz_mov_epi64 (__U, __A);
+ __MM512_REDUCE_OP (max_epu64);
}
-#else
-#define _mm512_cmp_pd_mask(X, Y, P) \
- ((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
- (__v8df)(__m512d)(Y), (int)(P),\
- (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m256d __T1 = (__m256d) _mm512_extractf64x4_pd (__A, 1); \
+ __m256d __T2 = (__m256d) _mm512_extractf64x4_pd (__A, 0); \
+ __m256d __T3 = __T1 op __T2; \
+ __m128d __T4 = _mm256_extractf128_pd (__T3, 1); \
+ __m128d __T5 = _mm256_extractf128_pd (__T3, 0); \
+ __m128d __T6 = __T4 op __T5; \
+ return __T6[0] op __T6[1]
-#define _mm512_cmp_ps_mask(X, Y, P) \
- ((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
- (__v16sf)(__m512)(Y), (int)(P),\
- (__mmask16)-1,_MM_FROUND_CUR_DIRECTION))
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_add_pd (__m512d __A)
+{
+ __MM512_REDUCE_OP (+);
+}
-#define _mm512_mask_cmp_pd_mask(M, X, Y, P) \
- ((__mmask8) __builtin_ia32_cmppd512_mask ((__v8df)(__m512d)(X), \
- (__v8df)(__m512d)(Y), (int)(P),\
- (__mmask8)M, _MM_FROUND_CUR_DIRECTION))
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_mul_pd (__m512d __A)
+{
+ __MM512_REDUCE_OP (*);
+}
-#define _mm512_mask_cmp_ps_mask(M, X, Y, P) \
- ((__mmask16) __builtin_ia32_cmpps512_mask ((__v16sf)(__m512)(X), \
- (__v16sf)(__m512)(Y), (int)(P),\
- (__mmask16)M,_MM_FROUND_CUR_DIRECTION))
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_add_pd (__mmask8 __U, __m512d __A)
+{
+ __A = _mm512_maskz_mov_pd (__U, __A);
+ __MM512_REDUCE_OP (+);
+}
-#define _mm_cmp_sd_mask(X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), (int)(P),\
- (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_mul_pd (__mmask8 __U, __m512d __A)
+{
+ __A = _mm512_mask_mov_pd (_mm512_set1_pd (1.0), __U, __A);
+ __MM512_REDUCE_OP (*);
+}
-#define _mm_mask_cmp_sd_mask(M, X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpsd_mask ((__v2df)(__m128d)(X), \
- (__v2df)(__m128d)(Y), (int)(P),\
- M,_MM_FROUND_CUR_DIRECTION))
+#undef __MM512_REDUCE_OP
+#define __MM512_REDUCE_OP(op) \
+ __m256d __T1 = (__m256d) _mm512_extractf64x4_pd (__A, 1); \
+ __m256d __T2 = (__m256d) _mm512_extractf64x4_pd (__A, 0); \
+ __m256d __T3 = _mm256_##op (__T1, __T2); \
+ __m128d __T4 = _mm256_extractf128_pd (__T3, 1); \
+ __m128d __T5 = _mm256_extractf128_pd (__T3, 0); \
+ __m128d __T6 = _mm_##op (__T4, __T5); \
+ __m128d __T7 = (__m128d) __builtin_shuffle (__T6, (__v2di) { 1, 0 }); \
+ __m128d __T8 = _mm_##op (__T6, __T7); \
+ return __T8[0]
-#define _mm_cmp_ss_mask(X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), (int)(P), \
- (__mmask8)-1,_MM_FROUND_CUR_DIRECTION))
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_min_pd (__m512d __A)
+{
+ __MM512_REDUCE_OP (min_pd);
+}
-#define _mm_mask_cmp_ss_mask(M, X, Y, P) \
- ((__mmask8) __builtin_ia32_cmpss_mask ((__v4sf)(__m128)(X), \
- (__v4sf)(__m128)(Y), (int)(P), \
- M,_MM_FROUND_CUR_DIRECTION))
-#endif
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_reduce_max_pd (__m512d __A)
+{
+ __MM512_REDUCE_OP (max_pd);
+}
-extern __inline __mmask16
+extern __inline double
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm512_kmov (__mmask16 __A)
+_mm512_mask_reduce_min_pd (__mmask8 __U, __m512d __A)
+{
+ __A = _mm512_mask_mov_pd (_mm512_set1_pd (__builtin_inf ()), __U, __A);
+ __MM512_REDUCE_OP (min_pd);
+}
+
+extern __inline double
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm512_mask_reduce_max_pd (__mmask8 __U, __m512d __A)
{
- return __builtin_ia32_kmov16 (__A);
+ __A = _mm512_mask_mov_pd (_mm512_set1_pd (-__builtin_inf ()), __U, __A);
+ __MM512_REDUCE_OP (max_pd);
}
+#undef __MM512_REDUCE_OP
+
#ifdef __DISABLE_AVX512F__
#undef __DISABLE_AVX512F__
#pragma GCC pop_options
projects / thirdparty / gcc.git / blobdiff