/* Copyright (C) 2006, 2007, 2008, 2009 Free Software Foundation, Inc.
|
/* Copyright (C) 2006, 2007, 2008, 2009 Free Software Foundation, Inc.
|
|
|
This file is part of GCC.
|
This file is part of GCC.
|
|
|
GCC is free software; you can redistribute it and/or modify
|
GCC is free software; you can redistribute it and/or modify
|
it under the terms of the GNU General Public License as published by
|
it under the terms of the GNU General Public License as published by
|
the Free Software Foundation; either version 3, or (at your option)
|
the Free Software Foundation; either version 3, or (at your option)
|
any later version.
|
any later version.
|
|
|
GCC is distributed in the hope that it will be useful,
|
GCC is distributed in the hope that it will be useful,
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
GNU General Public License for more details.
|
GNU General Public License for more details.
|
|
|
Under Section 7 of GPL version 3, you are granted additional
|
Under Section 7 of GPL version 3, you are granted additional
|
permissions described in the GCC Runtime Library Exception, version
|
permissions described in the GCC Runtime Library Exception, version
|
3.1, as published by the Free Software Foundation.
|
3.1, as published by the Free Software Foundation.
|
|
|
You should have received a copy of the GNU General Public License and
|
You should have received a copy of the GNU General Public License and
|
a copy of the GCC Runtime Library Exception along with this program;
|
a copy of the GCC Runtime Library Exception along with this program;
|
see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
|
see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
|
<http://www.gnu.org/licenses/>. */
|
<http://www.gnu.org/licenses/>. */
|
|
|
/* Implemented from the specification included in the Intel C++ Compiler
|
/* Implemented from the specification included in the Intel C++ Compiler
|
User Guide and Reference, version 9.1. */
|
User Guide and Reference, version 9.1. */
|
|
|
#ifndef _TMMINTRIN_H_INCLUDED
|
#ifndef _TMMINTRIN_H_INCLUDED
|
#define _TMMINTRIN_H_INCLUDED
|
#define _TMMINTRIN_H_INCLUDED
|
|
|
#ifndef __SSSE3__
|
#ifndef __SSSE3__
|
# error "SSSE3 instruction set not enabled"
|
# error "SSSE3 instruction set not enabled"
|
#else
|
#else
|
|
|
/* We need definitions from the SSE3, SSE2 and SSE header files*/
|
/* We need definitions from the SSE3, SSE2 and SSE header files*/
|
#include <pmmintrin.h>
|
#include <pmmintrin.h>
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadd_epi16 (__m128i __X, __m128i __Y)
|
_mm_hadd_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phaddw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_phaddw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadd_epi32 (__m128i __X, __m128i __Y)
|
_mm_hadd_epi32 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phaddd128 ((__v4si)__X, (__v4si)__Y);
|
return (__m128i) __builtin_ia32_phaddd128 ((__v4si)__X, (__v4si)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadds_epi16 (__m128i __X, __m128i __Y)
|
_mm_hadds_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phaddsw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_phaddsw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadd_pi16 (__m64 __X, __m64 __Y)
|
_mm_hadd_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phaddw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_phaddw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadd_pi32 (__m64 __X, __m64 __Y)
|
_mm_hadd_pi32 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phaddd ((__v2si)__X, (__v2si)__Y);
|
return (__m64) __builtin_ia32_phaddd ((__v2si)__X, (__v2si)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hadds_pi16 (__m64 __X, __m64 __Y)
|
_mm_hadds_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phaddsw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_phaddsw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsub_epi16 (__m128i __X, __m128i __Y)
|
_mm_hsub_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phsubw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_phsubw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsub_epi32 (__m128i __X, __m128i __Y)
|
_mm_hsub_epi32 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phsubd128 ((__v4si)__X, (__v4si)__Y);
|
return (__m128i) __builtin_ia32_phsubd128 ((__v4si)__X, (__v4si)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsubs_epi16 (__m128i __X, __m128i __Y)
|
_mm_hsubs_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_phsubsw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_phsubsw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsub_pi16 (__m64 __X, __m64 __Y)
|
_mm_hsub_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phsubw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_phsubw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsub_pi32 (__m64 __X, __m64 __Y)
|
_mm_hsub_pi32 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phsubd ((__v2si)__X, (__v2si)__Y);
|
return (__m64) __builtin_ia32_phsubd ((__v2si)__X, (__v2si)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_hsubs_pi16 (__m64 __X, __m64 __Y)
|
_mm_hsubs_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_phsubsw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_phsubsw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_maddubs_epi16 (__m128i __X, __m128i __Y)
|
_mm_maddubs_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_pmaddubsw128 ((__v16qi)__X, (__v16qi)__Y);
|
return (__m128i) __builtin_ia32_pmaddubsw128 ((__v16qi)__X, (__v16qi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_maddubs_pi16 (__m64 __X, __m64 __Y)
|
_mm_maddubs_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_pmaddubsw ((__v8qi)__X, (__v8qi)__Y);
|
return (__m64) __builtin_ia32_pmaddubsw ((__v8qi)__X, (__v8qi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_mulhrs_epi16 (__m128i __X, __m128i __Y)
|
_mm_mulhrs_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_pmulhrsw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_pmulhrsw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_mulhrs_pi16 (__m64 __X, __m64 __Y)
|
_mm_mulhrs_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_pmulhrsw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_pmulhrsw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_shuffle_epi8 (__m128i __X, __m128i __Y)
|
_mm_shuffle_epi8 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_pshufb128 ((__v16qi)__X, (__v16qi)__Y);
|
return (__m128i) __builtin_ia32_pshufb128 ((__v16qi)__X, (__v16qi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_shuffle_pi8 (__m64 __X, __m64 __Y)
|
_mm_shuffle_pi8 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_pshufb ((__v8qi)__X, (__v8qi)__Y);
|
return (__m64) __builtin_ia32_pshufb ((__v8qi)__X, (__v8qi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_epi8 (__m128i __X, __m128i __Y)
|
_mm_sign_epi8 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_psignb128 ((__v16qi)__X, (__v16qi)__Y);
|
return (__m128i) __builtin_ia32_psignb128 ((__v16qi)__X, (__v16qi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_epi16 (__m128i __X, __m128i __Y)
|
_mm_sign_epi16 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_psignw128 ((__v8hi)__X, (__v8hi)__Y);
|
return (__m128i) __builtin_ia32_psignw128 ((__v8hi)__X, (__v8hi)__Y);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_epi32 (__m128i __X, __m128i __Y)
|
_mm_sign_epi32 (__m128i __X, __m128i __Y)
|
{
|
{
|
return (__m128i) __builtin_ia32_psignd128 ((__v4si)__X, (__v4si)__Y);
|
return (__m128i) __builtin_ia32_psignd128 ((__v4si)__X, (__v4si)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_pi8 (__m64 __X, __m64 __Y)
|
_mm_sign_pi8 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_psignb ((__v8qi)__X, (__v8qi)__Y);
|
return (__m64) __builtin_ia32_psignb ((__v8qi)__X, (__v8qi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_pi16 (__m64 __X, __m64 __Y)
|
_mm_sign_pi16 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_psignw ((__v4hi)__X, (__v4hi)__Y);
|
return (__m64) __builtin_ia32_psignw ((__v4hi)__X, (__v4hi)__Y);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_sign_pi32 (__m64 __X, __m64 __Y)
|
_mm_sign_pi32 (__m64 __X, __m64 __Y)
|
{
|
{
|
return (__m64) __builtin_ia32_psignd ((__v2si)__X, (__v2si)__Y);
|
return (__m64) __builtin_ia32_psignd ((__v2si)__X, (__v2si)__Y);
|
}
|
}
|
|
|
#ifdef __OPTIMIZE__
|
#ifdef __OPTIMIZE__
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_alignr_epi8(__m128i __X, __m128i __Y, const int __N)
|
_mm_alignr_epi8(__m128i __X, __m128i __Y, const int __N)
|
{
|
{
|
return (__m128i) __builtin_ia32_palignr128 ((__v2di)__X,
|
return (__m128i) __builtin_ia32_palignr128 ((__v2di)__X,
|
(__v2di)__Y, __N * 8);
|
(__v2di)__Y, __N * 8);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_alignr_pi8(__m64 __X, __m64 __Y, const int __N)
|
_mm_alignr_pi8(__m64 __X, __m64 __Y, const int __N)
|
{
|
{
|
return (__m64) __builtin_ia32_palignr ((__v1di)__X,
|
return (__m64) __builtin_ia32_palignr ((__v1di)__X,
|
(__v1di)__Y, __N * 8);
|
(__v1di)__Y, __N * 8);
|
}
|
}
|
#else
|
#else
|
#define _mm_alignr_epi8(X, Y, N) \
|
#define _mm_alignr_epi8(X, Y, N) \
|
((__m128i) __builtin_ia32_palignr128 ((__v2di)(__m128i)(X), \
|
((__m128i) __builtin_ia32_palignr128 ((__v2di)(__m128i)(X), \
|
(__v2di)(__m128i)(Y), \
|
(__v2di)(__m128i)(Y), \
|
(int)(N) * 8))
|
(int)(N) * 8))
|
#define _mm_alignr_pi8(X, Y, N) \
|
#define _mm_alignr_pi8(X, Y, N) \
|
((__m64) __builtin_ia32_palignr ((__v1di)(__m64)(X), \
|
((__m64) __builtin_ia32_palignr ((__v1di)(__m64)(X), \
|
(__v1di)(__m64)(Y), \
|
(__v1di)(__m64)(Y), \
|
(int)(N) * 8))
|
(int)(N) * 8))
|
#endif
|
#endif
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_epi8 (__m128i __X)
|
_mm_abs_epi8 (__m128i __X)
|
{
|
{
|
return (__m128i) __builtin_ia32_pabsb128 ((__v16qi)__X);
|
return (__m128i) __builtin_ia32_pabsb128 ((__v16qi)__X);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_epi16 (__m128i __X)
|
_mm_abs_epi16 (__m128i __X)
|
{
|
{
|
return (__m128i) __builtin_ia32_pabsw128 ((__v8hi)__X);
|
return (__m128i) __builtin_ia32_pabsw128 ((__v8hi)__X);
|
}
|
}
|
|
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_epi32 (__m128i __X)
|
_mm_abs_epi32 (__m128i __X)
|
{
|
{
|
return (__m128i) __builtin_ia32_pabsd128 ((__v4si)__X);
|
return (__m128i) __builtin_ia32_pabsd128 ((__v4si)__X);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_pi8 (__m64 __X)
|
_mm_abs_pi8 (__m64 __X)
|
{
|
{
|
return (__m64) __builtin_ia32_pabsb ((__v8qi)__X);
|
return (__m64) __builtin_ia32_pabsb ((__v8qi)__X);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_pi16 (__m64 __X)
|
_mm_abs_pi16 (__m64 __X)
|
{
|
{
|
return (__m64) __builtin_ia32_pabsw ((__v4hi)__X);
|
return (__m64) __builtin_ia32_pabsw ((__v4hi)__X);
|
}
|
}
|
|
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
|
_mm_abs_pi32 (__m64 __X)
|
_mm_abs_pi32 (__m64 __X)
|
{
|
{
|
return (__m64) __builtin_ia32_pabsd ((__v2si)__X);
|
return (__m64) __builtin_ia32_pabsd ((__v2si)__X);
|
}
|
}
|
|
|
#endif /* __SSSE3__ */
|
#endif /* __SSSE3__ */
|
|
|
#endif /* _TMMINTRIN_H_INCLUDED */
|
#endif /* _TMMINTRIN_H_INCLUDED */
|
|
|
