| /*===---- mmintrin.h - MMX intrinsics --------------------------------------=== |
| * |
| * Permission is hereby granted, free of charge, to any person obtaining a copy |
| * of this software and associated documentation files (the "Software"), to deal |
| * in the Software without restriction, including without limitation the rights |
| * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| * copies of the Software, and to permit persons to whom the Software is |
| * furnished to do so, subject to the following conditions: |
| * |
| * The above copyright notice and this permission notice shall be included in |
| * all copies or substantial portions of the Software. |
| * |
| * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN |
| * THE SOFTWARE. |
| * |
| *===-----------------------------------------------------------------------=== |
| */ |
| |
| #ifndef __MMINTRIN_H |
| #define __MMINTRIN_H |
| |
| #ifndef __MMX__ |
| #error "MMX instruction set not enabled" |
| #else |
| |
| typedef long long __m64 __attribute__((__vector_size__(8))); |
| |
| typedef int __v2si __attribute__((__vector_size__(8))); |
| typedef short __v4hi __attribute__((__vector_size__(8))); |
| typedef char __v8qi __attribute__((__vector_size__(8))); |
| |
| static inline void __attribute__((__always_inline__, __nodebug__)) |
| _mm_empty(void) |
| { |
| __builtin_ia32_emms(); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cvtsi32_si64(int __i) |
| { |
| return (__m64)(__v2si){__i, 0}; |
| } |
| |
| static inline int __attribute__((__always_inline__, __nodebug__)) |
| _mm_cvtsi64_si32(__m64 __m) |
| { |
| __v2si __mmx_var2 = (__v2si)__m; |
| return __mmx_var2[0]; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cvtsi64_m64(long long __i) |
| { |
| return (__m64)__i; |
| } |
| |
| static inline long long __attribute__((__always_inline__, __nodebug__)) |
| _mm_cvtm64_si64(__m64 __m) |
| { |
| return (long long)__m; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_packs_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_packs_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_packs_pu16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpackhi_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v8qi)__m1, (__v8qi)__m2, 4, 8+4, 5, |
| 8+5, 6, 8+6, 7, 8+7); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpackhi_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v4hi)__m1, (__v4hi)__m2, 2, 4+2, 3, |
| 4+3); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpackhi_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v2si)__m1, (__v2si)__m2, 1, 2+1); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpacklo_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v8qi)__m1, (__v8qi)__m2, 0, 8+0, 1, |
| 8+1, 2, 8+2, 3, 8+3); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpacklo_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v4hi)__m1, (__v4hi)__m2, 0, 4+0, 1, |
| 4+1); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_unpacklo_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_shufflevector((__v2si)__m1, (__v2si)__m2, 0, 2+0); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_add_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v8qi)__m1 + (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_add_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v4hi)__m1 + (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_add_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v2si)__m1 + (__v2si)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_adds_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_paddsb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_adds_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_paddsw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_adds_pu8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_paddusb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_adds_pu16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_paddusw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sub_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v8qi)__m1 - (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sub_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v4hi)__m1 - (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sub_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v2si)__m1 - (__v2si)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_subs_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_psubsb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_subs_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_psubsw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_subs_pu8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_psubusb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_subs_pu16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_psubusw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_madd_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pmaddwd((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_mulhi_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pmulhw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_mullo_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)((__v4hi)__m1 * (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sll_pi16(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psllw((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_slli_pi16(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_psllwi((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sll_pi32(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_pslld((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_slli_pi32(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_pslldi((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sll_si64(__m64 __m, __m64 __count) |
| { |
| return __builtin_ia32_psllq(__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_slli_si64(__m64 __m, int __count) |
| { |
| return __builtin_ia32_psllqi(__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sra_pi16(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psraw((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srai_pi16(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_psrawi((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_sra_pi32(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psrad((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srai_pi32(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_psradi((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srl_pi16(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psrlw((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srli_pi16(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_psrlwi((__v4hi)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srl_pi32(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psrld((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srli_pi32(__m64 __m, int __count) |
| { |
| return (__m64)__builtin_ia32_psrldi((__v2si)__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srl_si64(__m64 __m, __m64 __count) |
| { |
| return (__m64)__builtin_ia32_psrlq(__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_srli_si64(__m64 __m, int __count) |
| { |
| return __builtin_ia32_psrlqi(__m, __count); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_and_si64(__m64 __m1, __m64 __m2) |
| { |
| return __m1 & __m2; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_andnot_si64(__m64 __m1, __m64 __m2) |
| { |
| return ~__m1 & __m2; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_or_si64(__m64 __m1, __m64 __m2) |
| { |
| return __m1 | __m2; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_xor_si64(__m64 __m1, __m64 __m2) |
| { |
| return __m1 ^ __m2; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpeq_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpeqb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpeq_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpeqw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpeq_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpeqd((__v2si)__m1, (__v2si)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpgt_pi8(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpgtb((__v8qi)__m1, (__v8qi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpgt_pi16(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpgtw((__v4hi)__m1, (__v4hi)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_cmpgt_pi32(__m64 __m1, __m64 __m2) |
| { |
| return (__m64)__builtin_ia32_pcmpgtd((__v2si)__m1, (__v2si)__m2); |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_setzero_si64(void) |
| { |
| return (__m64){ 0LL }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set_pi32(int __i1, int __i0) |
| { |
| return (__m64)(__v2si){ __i0, __i1 }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set_pi16(short __s3, short __s2, short __s1, short __s0) |
| { |
| return (__m64)(__v4hi){ __s0, __s1, __s2, __s3 }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set_pi8(char __b7, char __b6, char __b5, char __b4, char __b3, char __b2, |
| char __b1, char __b0) |
| { |
| return (__m64)(__v8qi){ __b0, __b1, __b2, __b3, __b4, __b5, __b6, __b7 }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set1_pi32(int __i) |
| { |
| return (__m64)(__v2si){ __i, __i }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set1_pi16(short __s) |
| { |
| return (__m64)(__v4hi){ __s }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_set1_pi8(char __b) |
| { |
| return (__m64)(__v8qi){ __b }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_setr_pi32(int __i1, int __i0) |
| { |
| return (__m64)(__v2si){ __i1, __i0 }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_setr_pi16(short __s3, short __s2, short __s1, short __s0) |
| { |
| return (__m64)(__v4hi){ __s3, __s2, __s1, __s0 }; |
| } |
| |
| static inline __m64 __attribute__((__always_inline__, __nodebug__)) |
| _mm_setr_pi8(char __b7, char __b6, char __b5, char __b4, char __b3, char __b2, |
| char __b1, char __b0) |
| { |
| return (__m64)(__v8qi){ __b7, __b6, __b5, __b4, __b3, __b2, __b1, __b0 }; |
| } |
| |
| #endif /* __MMX__ */ |
| |
| #endif /* __MMINTRIN_H */ |
| |