X-Git-Url: https://scm.cri.ensmp.fr/git/Faustine.git/blobdiff_plain/1059e1cc0c2ecfa237406949aa26155b6a5b9154..66f23d4fabf89ad09adbd4dfc15ac6b5b2b7da83:/interpretor/preprocessor/faust-0.9.47mr3/architecture/intrinsic.hh diff --git a/interpretor/preprocessor/faust-0.9.47mr3/architecture/intrinsic.hh b/interpretor/preprocessor/faust-0.9.47mr3/architecture/intrinsic.hh deleted file mode 100644 index 25b594e..0000000 --- a/interpretor/preprocessor/faust-0.9.47mr3/architecture/intrinsic.hh +++ /dev/null @@ -1,1013 +0,0 @@ -/****************************************************/ -/* */ -/* intrinsic.hh: */ -/* */ -/* */ -/* Nicolas Scaringella */ -/* */ -/****************************************************/ - - - -// dans fichiers architectures desormais : -// inline void *aligned_calloc(size_t nmemb, size_t size) { return (void*)((unsigned)(calloc((nmemb*size)+15,sizeof(char)))+15 & 0xfffffff0); } - - - -#ifdef __SSE2__ - -/****************************************************/ -/* */ -/* SSE2 implementation */ -/* */ -/****************************************************/ - - -#include -#include -#include -//#include - - - -struct vec_int -{ - __m128i vec; - - vec_int() {} - - vec_int(int a) { vec = _mm_set_epi32(a,a,a,a); } - - vec_int(int a, int b, int c, int d) { vec = _mm_set_epi32(d,c,b,a); } - - vec_int(__m128i m) { vec = m; } - - operator __m128i() const { return vec; } - - const int& operator[](int i)const { int* ip = (int*)&vec; return *(ip+i); } - - int& operator[](int i) { int* ip = (int*)&vec; return *(ip+i); } - -}; - -struct vec_float -{ - __m128 vec; - - vec_float() {} - - vec_float(float a) { vec = _mm_set_ps1(a); } - - vec_float(float a, float b, float c, float d) { vec = _mm_set_ps(d,c,b,a); } - - vec_float(__m128 m) { vec = m; } - - //vec_float(vec_int vi) { vec = _mm_cvtepi32_ps(vi); } - - operator __m128() const { return vec; } - - const float& operator[](int i)const { float* fp = (float*)&vec; return *(fp+i); } - - float& operator[](int i) { float* fp = (float*)&vec; return *(fp+i); } - -}; - - -// Flush to zero mode: during underflow zero result is returned when the result is true -// Not compatible with the IEEE standard 754 ( which -// deliver denormalized result in case of underflow ) -#define NO_DENORMALIZE _mm_setcsr((_mm_getcsr() & ~_MM_FLUSH_ZERO_MASK) | (_MM_FLUSH_ZERO_ON)) -#define DENORMALIZE _mm_setcsr((_mm_getcsr() & ~_MM_FLUSH_ZERO_MASK) | (_MM_FLUSH_ZERO_OFF)) - - - - -// constants - -// 0 0 0 0 -#define VEC_INT_ZERO(a) _mm_xor_si128(a,a) - -// 0xffffffff 0xffffffff 0xffffffff 0xffffffff -#define VEC_INT_ONES(a) _mm_cmpeq_epi32(a,a) - -// Example: 2^10 - 1 = 1023 -> VEC_INT_PW2_MINUS_1(a,10) -#define VEC_INT_PW2_MINUS_1(a,pw) _mm_srli_epi32(_mm_cmpeq_epi32(a,a),32-pw) - -// 1 1 1 1: particular case -#define VEC_INT_ONE(a) _mm_srli_epi32(_mm_cmpeq_epi32(a,a),31) - -// Example: 2^10 = 1024 -> VEC_INT_PW2(a,10) -#define VEC_INT_PW2(a,pw) _mm_slli_epi32(_mm_srli_epi32(_mm_cmpeq_epi32(a,a),31),pw) - -// Example: -2^10 = -1024 -> VEC_INT_MINUS_PW2(a,10) -#define VEC_INT_MINUS_PW2(a,pw) _mm_slli_epi32(_mm_cmpeq_epi32(a,a),pw) - -// -1 -1 -1 -1: particular case -#define VEC_INT_MINUS_ONE(a) _mm_cmpeq_epi32(a,a) - -// 0.0 0.0 0.0 0.0 -#define VEC_FLOAT_ZERO(a) _mm_xor_ps(a,a) - -// 0xffffffff 0xffffffff 0xffffffff 0xffffffff -#define VEC_FLOAT_ONES(a) _mm_cmpeq_ps(a,a) - - - -// conversions entre vecteurs d'ints et de floats -inline vec_int float2int( vec_float a) { return _mm_cvtps_epi32(a); } -inline int float2int( float a ) { return int(a); } - -inline vec_float int2float( vec_int a) { return _mm_cvtepi32_ps(a); } -inline float int2float( int a ) { return float(a); } - - - - -// arithmetic -inline vec_float add_vec( vec_float a, vec_float b) { return _mm_add_ps(a,b); } -inline vec_float add_vec( vec_int a, vec_float b) { return _mm_add_ps(int2float(a),b); } -inline vec_float add_vec( vec_float a, vec_int b) { return _mm_add_ps(a,int2float(b)); } - -inline vec_int add_vec( vec_int a, vec_int b) { return _mm_add_epi32(a,b); } - - -inline vec_float add_scal( vec_float a, vec_float b) { return _mm_add_ss(a,b); } -inline vec_int add_scal( vec_int a, vec_int b) { return _mm_add_epi32(a,b); } // _mm_add_pi32 en MMX -//inline scal_int add_scal( scal_int a, scal_int b) { return _mm_add_pi32(a,b); } - - -inline vec_float sub_vec( vec_float a, vec_float b) { return _mm_sub_ps(a,b); } -inline vec_int sub_vec( vec_int a, vec_int b) { return _mm_sub_epi32(a,b); } -inline vec_float sub_scal( vec_float a, vec_float b) { return _mm_sub_ss(a,b); } -inline vec_int sub_scal( vec_int a, vec_int b) { return _mm_sub_epi32(a,b); } // _mm_sub_pi32 en MMX -//inline scal_int sub_scal( scal_int a, scal_int b) { return _mm_sub_pi32(a,b); } - - -inline vec_float mul_vec( vec_float a, vec_float b) { return _mm_mul_ps(a,b); } -inline vec_float mul_vec( vec_int a, vec_float b) { return _mm_mul_ps(int2float(a),b); } -inline vec_float mul_vec( vec_float a, vec_int b) { return _mm_mul_ps(a,int2float(b)); } - -inline vec_float mul_scal( vec_float a, vec_float b) { return _mm_mul_ss(a,b); } - -// INTEGER MULTIPLICATION -// low 32 bits of a 32 * 32 bit multiplication: each double-word X and Y is broken down into two words, A & B and C & D: -// X = ( A << 16 ) + B -// Y = ( C << 16 ) + D -// then: -// X * Y = (( A << 16 ) + B ) * (( C << 16 ) + D ) -// X * Y = ( A*C << 32 ) + ( A*D << 16 ) + ( B*C << 16 ) + B*D -// the partial result A*C does not appear in the low 32 bits result so does not need to be computed -// ( however, if it's different from zero, then there is an overflow ) - -inline vec_int mul_vec( vec_int a, vec_int b) { - - vec_int temp0 = _mm_shufflehi_epi16( _mm_shufflelo_epi16( b, 0xB1), 0xB1); - vec_int temp1 = _mm_and_si128( b, _mm_srli_epi32( _mm_cmpeq_epi32( b,b), 16)); - - vec_int temp2 = _mm_madd_epi16( a, temp0); - vec_int temp3 = _mm_madd_epi16( a, temp1); - - vec_int temp4 = _mm_slli_epi32( temp2, 16); - - return _mm_add_epi32( temp4, temp3); -} - -inline vec_int mul_scal( vec_int a, vec_int b) { - - vec_int temp0 = _mm_shufflelo_epi16( b, 0xB1); - vec_int temp1 = _mm_and_si128( b, _mm_cvtsi32_si128(0x00ff)); - - vec_int temp2 = _mm_madd_epi16( a, temp0); - vec_int temp3 = _mm_madd_epi16( a, temp1); - - vec_int temp4 = _mm_slli_epi32( temp2, 16); - - return _mm_add_epi32( temp4, temp3); -} - - -inline vec_float div_vec( vec_float a, vec_float b) { return _mm_mul_ps(a,_mm_rcp_ps(b)); /*_mm_div_ps(a,b);*/ } -inline vec_int div_vec( vec_int a, vec_int b) { return _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(a),_mm_rcp_ps(_mm_cvtepi32_ps(b)))); } // A CHANGER !!!! -inline vec_float div_scal( vec_float a, vec_float b) { return _mm_mul_ss(a,_mm_rcp_ss(b)); /*_mm_div_ss(a,b);*/ } -inline vec_int div_scal( vec_int a, vec_int b) { return _mm_cvtps_epi32(_mm_mul_ss(_mm_cvtepi32_ps(a),_mm_rcp_ss(_mm_cvtepi32_ps(b)))); } // A CHANGER !!!!! -//inline scal_int div_scal( scal_int a, scal_int b) { return _mm_cvtsi32_si64((_mm_cvtsi64_si32(a))/(_mm_cvtsi64_si32(b))); } // A CHANGER !!!!! - - -inline vec_int mod_vec( vec_int a, vec_int N) { - - vec_int temp = _mm_sub_epi32(a,N); - vec_int zero = _mm_xor_si128(a,a); - - vec_int select = _mm_xor_si128( _mm_cmpgt_epi32(temp,zero), _mm_cmpeq_epi32(temp,zero)); // a - N >= 0 - - return _mm_or_si128(_mm_and_si128(select,temp),_mm_andnot_si128(select,a)); // if( a - N >=0 ) return a - N; else return a; -} - -inline vec_int mod_scal( vec_int a, vec_int N) { - - vec_int temp = _mm_sub_epi32(a,N); - vec_int zero = _mm_xor_si128(a,a); - - vec_int select = _mm_xor_si128( _mm_cmpgt_epi32(temp,zero), _mm_cmpeq_epi32(temp,zero)); // a - N >= 0 - - return _mm_or_si128(_mm_and_si128(select,temp),_mm_andnot_si128(select,a)); // if( a - N >=0 ) return a - N; else return a; -} - - - -// simulation of a*b + c -#define madd_vec(a,b,c) add_vec(mul_vec(a,b),c) -#define madd_scal(a,b,c) add_scal(mul_scal(a,b),c) - -//inline vec_float madd_scal( vec_float a, vec_float b, vec_float c) { return _mm_add_ss(_mm_mul_ss(a,b),c); } -//inline vec_int madd_vec( vec_int a, vec_int b, vec_int c) { return add_vec(mul_vec(a,b),c); } -//inline vec_int madd_scal( vec_int a, vec_int b, vec_int c) { return add_scal(mul_scal(a,b),c); } - - -// simulation of - ( a*b - c ) -//inline vec_float nmsub_vec( vec_float a, vec_float b, vec_float c) { } - -// simulation of a*(1/b) + c -inline vec_float divadd_vec( vec_float a, vec_float b, vec_float c) { return _mm_add_ps(_mm_mul_ps(a,_mm_rcp_ps(b)),c); } -inline vec_float divadd_scal( vec_float a, vec_float b, vec_float c) { return _mm_add_ss(_mm_mul_ss(a,_mm_rcp_ss(b)),c); } -// simulation of - ( a*(1/b) - c ) -//inline vec_float divsub_vec( vec_float a, vec_float b, vec_float c) { } - - - - -// shift ( and fill with 0's ) -inline vec_int shift_left_vec( vec_int a, vec_int num) { return _mm_sll_epi32(a,num); } -inline vec_int shift_left_vec( vec_int a, int num) { return _mm_slli_epi32(a,num); } -inline vec_int shift_left_scal( vec_int a, vec_int num) { return _mm_sll_epi32(a,num); } // _mm_sll_pi32(a,num) en MMX -//inline scal_int shift_left_scal( scal_int a, scal_int num) { return _mm_sll_pi32(a,num); } -inline vec_int shift_left_scal( vec_int a, int num) { return _mm_slli_epi32(a,num); } // _mm_slli_pi32(a,num) en MMX -//inline scal_int shift_left_scal( scal_int a, int num) { return _mm_slli_pi32(a,num); } - - -// shift ( and fill with the sign bit ) -inline vec_int shift_right_vec( vec_int a, vec_int num) { return _mm_sra_epi32(a,num); } -inline vec_int shift_right_vec( vec_int a, int num) { return _mm_srai_epi32(a,num); } -inline vec_int shift_right_scal( vec_int a, vec_int num) { return _mm_sra_epi32(a,num); } // _mm_sra_pi32(a,num) en MMX -//inline scal_int shift_right_scal( scal_int a, scal_int num) { return _mm_sra_pi32(a,num); } -inline vec_int shift_right_scal( vec_int a, int num) { return _mm_srai_epi32(a,num); } // _mm_srai_pi32(a,num) en MMX -//inline scal_int shift_right_scal( scal_int a, int num) { return _mm_srai_pi32(a,num); } - - -// shift ( and fill with 0's ) -inline vec_int shift_right_vec_logical( vec_int a, vec_int num) { return _mm_srl_epi32(a,num); } -inline vec_int shift_right_vec_logical( vec_int a, int num) { return _mm_srli_epi32(a,num); } -inline vec_int shift_right_scal_logical( vec_int a, vec_int num) { return _mm_srl_epi32(a,num); } // _mm_sra_pi32(a,num) en MMX -//inline scal_int shift_right_scal_logical( scal_int a, scal_int num) { return _mm_srl_pi32(a,num); } -inline vec_int shift_right_scal_logical( vec_int a, int num) { return _mm_srli_epi32(a,num); } // _mm_srai_pi32(a,num) en MMX -//inline scal_int shift_right_scal_logical( scal_int a, int num) { return _mm_srli_pi32(a,num); } - - -// Logic -// Ajouts YO;; supprime -//inline vec_float and_vec( vec_float a, vec_int b) { return _mm_and_ps(a,b); } -//inline vec_float and_vec( vec_int a, vec_float b) { return _mm_and_ps(a,b); } - -inline vec_float and_vec( vec_float a, vec_float b) { return _mm_and_ps(a,b); } -inline vec_int and_vec( vec_int a, vec_int b) { return _mm_and_si128(a,b); } -inline vec_float and_scal( vec_float a, vec_float b) { return _mm_and_ps(a,b); } -inline vec_int and_scal( vec_int a, vec_int b) { return _mm_and_si128(a,b); } // _mm_and_si64(a,b) en MMX -//inline scal_int and_scal( scal_int a, scal_int b) { return _mm_and_si64(a,b); } - - -inline vec_float or_vec( vec_float a, vec_float b) { return _mm_or_ps(a,b); } -inline vec_int or_vec( vec_int a, vec_int b) { return _mm_or_si128(a,b); } -inline vec_float or_scal( vec_float a, vec_float b) { return _mm_or_ps(a,b); } -inline vec_int or_scal( vec_int a, vec_int b) { return _mm_or_si128(a,b); } // _mm_or_si64(a,b) en MMX -//inline scal_int or_scal( scal_int a, scal_int b) { return _mm_or_si64(a,b); } - - -inline vec_float xor_vec( vec_float a, vec_float b) { return _mm_xor_ps(a,b); } -inline vec_int xor_vec( vec_int a, vec_int b) { return _mm_xor_si128(a,b); } -inline vec_float xor_scal( vec_float a, vec_float b) { return _mm_xor_ps(a,b); } -inline vec_int xor_scal( vec_int a, vec_int b) { return _mm_xor_si128(a,b); } // _mm_xor_si64(a,b) en MMX -//inline scal_int xor_scal( scal_int a, scal_int b) { return _mm_xor_si64(a,b); } - -//------------------------------------------------------------------------------------------------------------ -// YO : remplacement de inline vec_float par inline vec_int dans les operations de comparaison entre vec_float -// pour une meilleur compatibilité avec la compilation vectorielle -//------------------------------------------------------------------------------------------------------------ - -// cast (without conversion) -inline vec_float cast2vec_float(vec_int x) { return _mm_castsi128_ps(x); } -inline vec_int cast2vec_int(vec_float x) { return _mm_castps_si128(x); } - -// convertions -inline vec_float conv2vec_float(vec_int x) { return _mm_cvtepi32_ps(x); } -inline vec_int conv2vec_int(vec_float x) { return _mm_cvtps_epi32(x); } - - -// comparaison -//inline vec_float int2float( vec_int a) { return _mm_cvtepi32_ps(a); } - - - -inline vec_float gt_vec( vec_float a, vec_float b) { return _mm_cmpgt_ps(a,b); } -inline vec_float gt_vec( vec_int a, vec_float b) { return _mm_cmpgt_ps(_mm_cvtepi32_ps(a),b); } -inline vec_float gt_vec( vec_float a, vec_int b) { return _mm_cmpgt_ps(a,_mm_cvtepi32_ps(b)); } -inline vec_int gt_vec( vec_int a, vec_int b) { return _mm_cmpgt_epi32(a,b); } - -inline vec_float gt_scal( vec_float a, vec_float b) { return _mm_cmpgt_ps(a,b); } -inline vec_float gt_scal( vec_int a, vec_float b) { return _mm_cmpgt_ps(_mm_cvtepi32_ps(a),b); } -inline vec_float gt_scal( vec_float a, vec_int b) { return _mm_cmpgt_ps(a,_mm_cvtepi32_ps(b)); } -inline vec_int gt_scal( vec_int a, vec_int b) { return _mm_cmpgt_epi32(a,b); } - -// choose between two values choose(c,u,v) = c?u:v -// the type of the result depends of the types of u and v, not of the type of c - -inline vec_float choose(vec_float c, vec_float u, vec_float v) { return _mm_or_ps(_mm_and_ps(c,u), _mm_andnot_ps(c,v)); } -inline vec_float choose(vec_float c, vec_int u, vec_float v) { return _mm_or_ps(_mm_and_ps(c,_mm_cvtepi32_ps(u)), _mm_andnot_ps(c,v)); } -inline vec_float choose(vec_float c, vec_float u, vec_int v) { return _mm_or_ps(_mm_and_ps(c,u), _mm_andnot_ps(c,_mm_cvtepi32_ps(v))); } - -inline vec_float choose(vec_int c, vec_float u, vec_float v) { return choose(cast2vec_float(c), u, v); } -inline vec_float choose(vec_int c, vec_int u, vec_float v) { return choose(cast2vec_float(c), u, v); } -inline vec_float choose(vec_int c, vec_float u, vec_int v) { return choose(cast2vec_float(c), u, v); } - -inline vec_int choose(vec_int c, vec_int u, vec_int v) { return _mm_or_si128(_mm_and_si128(c,u), _mm_andnot_si128(c,v)); } -inline vec_int choose(vec_float c, vec_int u, vec_int v) { return choose(cast2vec_int(c), u, v); } - -// choose between two values choosezero(c,u) = c?u:0 -inline vec_float choosezero(vec_float c, vec_float u) { return _mm_and_ps(c,u); } -inline vec_float choosezero(vec_int c, vec_float u) { return choosezero(cast2vec_float(c), u); } - -inline vec_int choosezero(vec_int c, vec_int u) { return _mm_and_si128(c,u); } -inline vec_int choosezero(vec_float c, vec_int u) { return choosezero(cast2vec_int(c), u); } - - -//inline vec_int gt_vec( vec_float a, vec_float b) { return _mm_srli_epi32(_mm_cmpgt_ps(a,b),31); } -//inline vec_int gt_vec( vec_float a, vec_float b) { vec_univ v; v.f4 = _mm_cmpgt_ps(a,b); return _mm_srli_epi32(v.i4,31); } -//inline vec_int gt_vec( vec_int a, vec_int b) { return _mm_cmpgt_epi32(a,b); } -//inline vec_int gt_scal( vec_float a, vec_float b) { return _mm_srli_epi32(_mm_cmpgt_ss(a,b),31); } -//inline vec_int gt_scal( vec_float a, vec_float b) { vec_univ v; v.f4 = _mm_cmpgt_ss(a,b); return _mm_srli_epi32(v.i4,31); } -//inline vec_int gt_scal( vec_float a, vec_float b) { return _mm_srli_epi32(_mm_cmpgt_ss(a,b),31); } -//inline vec_int gt_scal( vec_int a, vec_float b) { return _mm_srli_epi32(_mm_cmpgt_ss(a,b),31); } -//inline vec_int gt_scal( vec_int a, vec_int b) { return _mm_cmpgt_epi32(a,b); } // _mm_cmpgt_pi32(a,b) en MMX - -//inline scal_int gt_scal( scal_int a, scal_int b) { return (__m128i) _mm_cmpgt_pi32(a,b); } - -#if 0 - -inline vec_int lt_vec( vec_float a, vec_float b) { return _mm_cmplt_ps(a,b); } -inline vec_int lt_vec( vec_int a, vec_int b) { return _mm_cmpgt_epi32(b,a); } -inline vec_int lt_scal( vec_float a, vec_float b) { return _mm_cmplt_ss(a,b); } -inline vec_int lt_scal( vec_int a, vec_int b) { return _mm_cmpgt_epi32(b,a); } // _mm_cmpgt_pi32(b,a) en MMX -//inline scal_int lt_scal( scal_int a, scal_int b) { return _mm_cmpgt_pi32(b,a); } - - - -inline vec_int ge_vec( vec_float a, vec_float b) { return _mm_cmpge_ps(a,b); } -inline vec_int ge_vec( vec_int a, vec_int b) { return _mm_xor_si128( _mm_cmpgt_epi32(a,b), _mm_cmpeq_epi32(a,b)); } -inline vec_int ge_scal( vec_float a, vec_float b) { return _mm_cmpge_ss(a,b); } -inline vec_int ge_scal( vec_int a, vec_int b) { return _mm_xor_si128( _mm_cmpgt_epi32(a,b), _mm_cmpeq_epi32(a,b)); } // _mm_xor_si64,_mm_cmpgt_pi32,_mm_cmpeq_pi32 MMX -//inline scal_int ge_scal( scal_int a, scal_int b) { return _mm_xor_si64( _mm_cmpgt_pi32(a,b),_mm_cmpeq_pi32(a,b)); } - - -inline vec_int le_vec( vec_float a, vec_float b) { return _mm_cmple_ps(a,b); } -inline vec_int le_vec( vec_int a, vec_int b) { return _mm_xor_si128( _mm_cmpgt_epi32(b,a), _mm_cmpeq_epi32(b,a)); } -inline vec_int le_scal( vec_float a, vec_float b) { return _mm_cmple_ss(a,b); } -inline vec_int le_scal( vec_int a, vec_int b) { return _mm_xor_si128( _mm_cmpgt_epi32(b,a), _mm_cmpeq_epi32(b,a)); } // _mm_xor_si64,_mm_cmpgt_pi32,_mm_cmpeq_pi32 MMX -//inline scal_int le_scal( scal_int a, scal_int b) { return _mm_xor_si64( _mm_cmpgt_pi32(b,a),_mm_cmpeq_pi32(b,a)); } - - -inline vec_int eq_vec( vec_float a, vec_float b) { return _mm_cmpeq_ps(a,b); } -inline vec_int eq_vec( vec_int a, vec_int b) { return _mm_cmpeq_epi32(a,b); } -inline vec_int eq_scal( vec_float a, vec_float b) { return _mm_cmpeq_ss(a,b); } -inline vec_int eq_scal( vec_int a, vec_int b) { return _mm_cmpeq_epi32(a,b); } // _mm_cmpeq_pi32(a,b) en MMX -//inline scal_int eq_scal( scal_int a, scal_int b) { return _mm_cmpeq_pi32(a,b); } - - -inline vec_int neq_vec( vec_float a, vec_float b) { return _mm_cmpneq_ps(a,b); } -inline vec_int neq_vec( vec_int a, vec_int b) { return _mm_andnot_si128(_mm_cmpeq_epi32(a,b), _mm_cmpeq_epi32(a,a)); } -inline vec_int neq_scal( vec_float a, vec_float b) { return _mm_cmpneq_ss(a,b); } -inline vec_int neq_scal( vec_int a, vec_int b) { return _mm_andnot_si128(_mm_cmpeq_epi32(a,b), _mm_cmpeq_epi32(a,a)); } // _mm_andnot_si64,_mm_cmpeq_pi32 MMX -//inline scal_int neq_scal( scal_int a, scal_int b) { return _mm_andnot_si64(_mm_cmpeq_pi32(a,b),SCAL_INT_ALL_ONE); } - -#endif - - -// memory - -#if 0 -inline vec_float set_vec( double a) { float val = float(a); vec_float temp = _mm_load_ss(&val); return _mm_shuffle_ps(temp,temp,0x00); } -inline vec_float set_vec( float a) { float val = a; vec_float temp = _mm_load_ss(&val); return _mm_shuffle_ps(temp,temp,0x00); } -inline vec_int set_vec( long int a) { vec_int temp = _mm_cvtsi32_si128(int(a)); temp = _mm_unpacklo_epi32(temp,temp); return _mm_unpacklo_epi32(temp,temp); } -inline vec_int set_vec( int a) { vec_int temp = _mm_cvtsi32_si128(a); temp = _mm_unpacklo_epi32(temp,temp); return _mm_unpacklo_epi32(temp,temp);} -inline vec_int set_vec( short a) { vec_int temp = _mm_cvtsi32_si128(int(a)); temp = _mm_unpacklo_epi32(temp,temp); return _mm_unpacklo_epi32(temp,temp); } -//inline scal_int set_vec( long int a) { _mm_cvtsi32_si64(int(a)); } -//inline scal_int set_vec( int a) { _mm_cvtsi32_si64(a); } -//inline scal_int set_vec( short a) { _mm_cvtsi32_si64(int(a)); } -#endif - -#if 0 - -inline vec_float set_vec( double a, double b, double c, double d) { __vec_float temp; temp.s[0]=float(a); temp.s[1]=float(b); temp.s[2]=float(c); temp.s[3]=float(d); return temp.v; } -inline vec_float set_vec( float a, float b, float c, float d) { __vec_float temp; temp.s[0]=a; temp.s[1]=b; temp.s[2]=c; temp.s[3]=d; return temp.v; } -inline vec_int set_vec( int a, int b, int c, int d) { __vec_int temp; temp.s[0]=a; temp.s[1]=b; temp.s[2]=c; temp.s[3]=d; return temp.v; } -inline vec_int set_vec( long int a, long int b, long int c, long int d) { __vec_int temp; temp.s[0]=int(a); temp.s[1]=int(b); temp.s[2]=int(c); temp.s[3]=int(d); return temp.v; } -inline vec_int set_vec( short a, short b, short c, short d) { __vec_int temp; temp.s[0]=short(a); temp.s[1]=short(b); temp.s[2]=short(c); temp.s[3]=short(d); return temp.v; } - -#endif - -inline vec_float set_vec( float a, float b, float c, float d) { return vec_float(a,b,c,d); } -inline vec_int set_vec( int a, int b, int c, int d) { return vec_int(a,b,c,d); } - -inline vec_float set_vec( float a) { return vec_float(a); } -inline vec_int set_vec( int a) { return vec_int(a); } - -inline vec_float load_a_vec( float* a) { return _mm_load_ps(a); } -inline vec_int load_a_vec( int* a) { return _mm_load_si128((__m128i*)a); } - -inline vec_float load_u_vec( float* a) { return _mm_loadu_ps(a); } -inline vec_int load_u_vec( int* a) { return _mm_loadu_si128((__m128i*)a); } - -// nouvelles fonctions d'écriture sans polluer le cache -inline void store_stream( float* a, vec_float b) { return _mm_stream_ps(a,b); } -inline void store_stream( int* a, vec_int b) { return _mm_stream_si128((__m128i*)a,b); } - -inline void store_a_vec( float* a, vec_float b) { return _mm_store_ps(a,b); } -inline void store_a_vec( int* a, vec_int b) { return _mm_store_si128((__m128i*)a,b); } - -inline void store_u_vec( float* a, vec_float b) { return _mm_storeu_ps(a,b); } -inline void store_u_vec( int* a, vec_int b) { return _mm_storeu_si128((__m128i*)a,b); } - - -inline vec_float load_scal(float* a) { return _mm_load_ss(a); } -inline vec_int load_scal(int* a) { return _mm_cvtsi32_si128(*a); } -//inline scal_int load_scal(int* a) { return _mm_cvtsi32_si64(*a); } - -inline void store_scal(float* a, vec_float content) { return _mm_store_ss(a,content); } -inline void store_scal(int* a, vec_int content) { *a = _mm_cvtsi128_si32(content); return; } -//inline void store_scal(int* a, scal_int content) { *a = _mm_cvtsi64_si32(content); return; } - - - -inline vec_float REC0(vec_float a) { return _mm_unpacklo_ps(a,a); } -inline vec_float REC1(vec_float a) { return _mm_unpacklo_ps(a,a); } -inline vec_float REC2(vec_float a) { return _mm_shuffle_ps(a,a,0x90); } -inline vec_float REC3(vec_float a) { return _mm_shuffle_ps(a,a,0x1B); } - -inline vec_int REC0(vec_int a) { return _mm_unpacklo_epi32(a,a); } -inline vec_int REC1(vec_int a) { return _mm_unpacklo_epi32(a,a); } -inline vec_int REC2(vec_int a) { return _mm_shuffle_epi32(a,0x90); } -inline vec_int REC3(vec_int a) { return _mm_shuffle_epi32(a,0x1B); } - - - - - -// scalar to vector: takes 4 vector which lower elements stands for a scalar value and rebuild a vector from these 4 scalar -//inline vec_float SCAL2VEC(vec_float a0,vec_float a1,vec_float a2,vec_float a3) { return _mm_shuffle_ps(_mm_shuffle_ps(a0,a1,0x00),_mm_shuffle_ps(a2,a3,0x00),0x88); } -inline vec_float SCAL2VEC(vec_float a0,vec_float a1,vec_float a2,vec_float a3) { return _mm_unpacklo_ps(_mm_unpacklo_ps(a0,a2),_mm_unpacklo_ps(a1,a3)); } -inline vec_int SCAL2VEC(vec_int a0,vec_int a1,vec_int a2,vec_int a3) { return _mm_unpacklo_epi32(_mm_unpacklo_epi32(a0,a2),_mm_unpacklo_epi32(a1,a3)); } -//inline vec_int SCAL2VEC(scal_int a0,scal_int a1,scal_int a2,scal_int a3) { return _mm_unpacklo_epi32( _mm_movpi64_epi64(_mm_unpacklo_pi32(a0,a2)), _mm_movpi64_epi64(_mm_unpacklo_pi32(a1,a3)) ); } // ou _mm_set_epi64( _mm_unpacklo_pi32(a0,a1), _mm_unpacklo_pi32(a2,a3)) - -inline vec_float SCAL2VEC(double a0, double a1, double a2, double a3) { return _mm_set_ps(float(a3),float(a2),float(a1),float(a0)); } -inline vec_float SCAL2VEC(float a0, float a1, float a2, float a3) { return _mm_set_ps(a3,a2,a1,a0); } -inline vec_int SCAL2VEC(long a0, long a1, long a2, long a3) { return _mm_set_epi32(int(a3),int(a2),int(a1),int(a0)); } -inline vec_int SCAL2VEC(int a0, int a1, int a2, int a3) { return _mm_set_epi32(a3,a2,a1,a0); } -inline vec_int SCAL2VEC(short a0, short a1, short a2, short a3) { return _mm_set_epi32(int(a3),int(a2),int(a1),int(a0)); } - - -// vector to scalar: build a scalar vector from one element of the initial vector -inline vec_float VEC2SCALVEC0(vec_float a) { return a; } // return x,x,x,a0 // _mm_shuffle_ps(a,a,Ox00) would return a0,a0,a0,a0 -inline vec_float VEC2SCALVEC1(vec_float a) { return _mm_shuffle_ps(a,a,0x55); } // return a1,a1,a1,a1 -inline vec_float VEC2SCALVEC2(vec_float a) { return _mm_shuffle_ps(a,a,0xAA); } // return a2,a2,a2,a2 -inline vec_float VEC2SCALVEC3(vec_float a) { return _mm_shuffle_ps(a,a,0xFF); } // return a3,a3,a3,a3 - -inline vec_int VEC2SCALVEC0(vec_int a) { return a; } // return x,x,x,a0 // _mm_shuffle_epi32(a,Ox00) would return a0,a0,a0,a0 -inline vec_int VEC2SCALVEC1(vec_int a) { return _mm_shuffle_epi32(a,0x55); } // return a1,a1,a1,a1 -inline vec_int VEC2SCALVEC2(vec_int a) { return _mm_shuffle_epi32(a,0xAA); } // return a2,a2,a2,a2 -inline vec_int VEC2SCALVEC3(vec_int a) { return _mm_shuffle_epi32(a,0xFF); } // return a3,a3,a3,a3 - -//inline scal_int VEC2SCALVEC0(vec_int a) { return _mm_movepi64_pi64(a); } // ATTENTION !!!! : -//inline scal_int VEC2SCALVEC1(vec_int a) { __m64 temp = _mm_movepi64_pi64(a); return _mm_unpackhi_pi32(temp,temp); } // VEC2SCALVEC0 et 1 peuvent être réunis en une instruction plus efficace -//inline scal_int VEC2SCALVEC2(vec_int a) { return _mm_movepi64_pi64(_mm_shuffle_epi32(a,0xAA)); } -//inline scal_int VEC2SCALVEC3(vec_int a) { return _mm_movepi64_pi64(_mm_shuffle_epi32(a,0xFF)); } - - - -// vector to scalar: build a single scalar from a vector -inline float VEC2SCAL0(vec_float a) { float temp; _mm_store_ss(&temp,a); return temp; } -inline float VEC2SCAL1(vec_float a) { float temp; _mm_store_ss(&temp,_mm_shuffle_ps(a,a,0x55)); return temp; } -inline float VEC2SCAL2(vec_float a) { float temp; _mm_store_ss(&temp,_mm_shuffle_ps(a,a,0xAA)); return temp; } -inline float VEC2SCAL3(vec_float a) { float temp; _mm_store_ss(&temp,_mm_shuffle_ps(a,a,0xFF)); return temp; } - -inline int VEC2SCAL0(vec_int a) { return _mm_cvtsi128_si32(a); } -inline int VEC2SCAL1(vec_int a) { return _mm_cvtsi128_si32(_mm_shuffle_epi32(a,0x55)); } -inline int VEC2SCAL2(vec_int a) { return _mm_cvtsi128_si32(_mm_shuffle_epi32(a,0xAA)); } -inline int VEC2SCAL3(vec_int a) { return _mm_cvtsi128_si32(_mm_shuffle_epi32(a,0xFF)); } -//inline int VEC2SCAL0(scal_int a) { return _mm_cvtsi64_si32(a); } - - - -// select: if( select == 0 ) then a ; else b ; -inline vec_float select_vec( vec_float select, vec_float a, vec_float b) { return _mm_or_ps( _mm_andnot_ps(select,a), _mm_and_ps(select,b));} -inline vec_float select_scal( vec_float select, vec_float a, vec_float b){ return _mm_or_ps( _mm_andnot_ps(select,a), _mm_and_ps(select,b));} -inline vec_float select_vec( vec_int select, vec_float a, vec_float b) { __m128 temp = _mm_cvtepi32_ps(select); return _mm_or_ps( _mm_andnot_ps(temp,a), _mm_and_ps(temp,b));} -inline vec_float select_scal( vec_int select, vec_float a, vec_float b){ __m128 temp = _mm_cvtepi32_ps(select); return _mm_or_ps( _mm_andnot_ps(temp,a), _mm_and_ps(temp,b));} -inline vec_int select_vec( vec_int select, vec_int a, vec_int b) { return _mm_or_si128( _mm_andnot_si128(select,a), _mm_and_si128(select,b)); } -inline vec_int select_scal( vec_int select, vec_int a, vec_int b) { return _mm_or_si128( _mm_andnot_si128(select,a), _mm_and_si128(select,b)); } // ou MMX - - - -// vectorial version of the "mem" Faust key-word -// return a[2] a[1] a[0] b[3] -inline vec_float mem1_vec( vec_float a, vec_float b) { return _mm_shuffle_ps(_mm_shuffle_ps(b,a,0x4E),a,0x99); } -inline vec_int mem1_vec( vec_int a, vec_int b) { - return _mm_unpacklo_epi32( _mm_shuffle_epi32( _mm_unpacklo_epi32( _mm_shuffle_epi32( b, 0xFF), a), 0xEE), _mm_shuffle_epi32( a, 0x88) ); -} - -// return a[1] a[0] b[3] b[2] -inline vec_float mem2_vec( vec_float a, vec_float b) { return _mm_shuffle_ps(b,a,0x4E); } -inline vec_int mem2_vec( vec_int a, vec_int b) { return _mm_shuffle_epi32( _mm_unpackhi_epi32( b, _mm_shuffle_epi32( a, 0x44)), 0xD8 ); - -} - -// return a[0] b[3] b[2] b[1] -inline vec_float mem3_vec( vec_float a, vec_float b) { return _mm_shuffle_ps(b,_mm_shuffle_ps(b,a,0x4E),0x99); } -inline vec_int mem3_vec( vec_int a, vec_int b) { - return _mm_unpacklo_epi32( _mm_shuffle_epi32( b, 0x99), _mm_shuffle_epi32( _mm_unpackhi_epi32( b, _mm_shuffle_epi32( a, 0x00)), 0xEE) ); -} - - - - - -// conversion -inline vec_float bool2float( vec_float a ) { return _mm_and_ps(a,set_vec(1.0f)); } -inline vec_float bool2float( vec_int a ) { return _mm_cvtepi32_ps(_mm_and_si128(a,_mm_sub_epi32(_mm_xor_si128(a,a),_mm_cmpeq_epi32(a,a)))); } - -inline vec_int bool2int( vec_int a) { return _mm_and_si128(a,_mm_sub_epi32(_mm_xor_si128(a,a),_mm_cmpeq_epi32(a,a))); } -inline vec_int bool2int( vec_float a ) { return _mm_cvtps_epi32(_mm_and_ps(a,set_vec(1.0f))); } - -inline vec_int boolfloat2boolint( vec_float a ) { vec_int temp; asm volatile("" : "=xmm" (temp) : "0" (a)); return temp; } -inline vec_float boolint2boolfloat( vec_int a ) { vec_float temp; asm volatile("" : "=xmm" (temp) : "0" (a)); return temp; } - - - - - - - -#elif defined(__ALTIVEC__) - - -/****************************************************/ -/* */ -/* ALTIVEC implementation */ -/* */ -/****************************************************/ - - -//#define vec_float vector float -//#define vec_int vector signed int -//#define vec_bool vector bool int - -struct vec_int -{ - vector signed int vec; - vec_int() {} - vec_int(vector signed int m) { vec = m; } -// operator __m128i() const { return vec; } - -}; - -struct vec_float -{ -// union { __m128 vec; __m128i i4; }; - vector float vec; - vec_float() {} - vec_float(vector float m) { vec = m; } - //vec_float(vec_int a) { vec = _mm_cvtepi32_ps(a); } - - //operator __m128() const { return vec; } -}; - -typedef union{ - float s[4]; - vec_float v; -} __vec_float; - - -typedef union{ - int s[4]; - vec_int v; -} __vec_int; - - -// Non-Java mode: during underflow zero result is returned -// Not compatible with the Java-IEEE-C9X standard ( which -// deliver denormalized result in case of underflow ) -#define NO_DENORMALIZE vec_mtvscr(vec_or(vec_mfvscr(),(vector unsigned short)(0x8000))) -#define DENORMALIZE vec_mtvscr(vec_or(vec_mfvscr(),(vector unsigned short)(0x0000))) - - - - -// constants - -// 0 0 0 0 -#define VEC_INT_ZERO(a) vec_xor(a,a) - -// 0xffffffff 0xffffffff 0xffffffff 0xffffffff -#define VEC_INT_ONES(a) (vector signed int)vec_cmpeq(a,a) - -// 1 1 1 1 -#define VEC_INT_ONE(a) vec_splat_s32(1) - -// -1 -1 -1 -1 -#define VEC_INT_MINUS_ONE(a) vec_splat_s32(-1) - -// a must belong to [-16,15] -// no efficient equivalent with SSE2 -#define VEC_INT_MINUS_16_TO_15(a) vec_splat_s32(a) - -// This not exactly equivalent to the SSE2 version -// the power must belong to [17,31] -// ( that is 2^17 - 1 = 32767 is the minimum -// and 2^31 - 1 = 2147483647 is the maximum; -// if you need 2^32 - 1, use VEC_INT_ONES ) -// Example: 2^19 - 1 = 524287 -> VEC_INT_PW2_MINUS_1(a,19) -#define VEC_INT_PW2_MINUS_1(a,pw) vec_sr((vector signed int)vec_cmpeq(a,a), vec_splat_u32(32-pw)) - - -// This not exactly equivalent to the SSE2 version -// the power must belong to [4,18] -// ( that is 2^18 = 262144 is the maximum -// and 2^4 = 16 is the minimum; -// if you need 2^0 = 1, use VEC_INT_ONE, -// if you need 2^1 = 2, use VEC_INT_MINUS_16_TO_15(2) -// if you need 2^2 = 4, use VEC_INT_MINUS_16_TO_15(4) -// if you need 2^3 = 8, use VEC_INT_MINUS_16_TO_15(8) ) -// Example: 2^10 = 1024 -> VEC_INT_PW2(a,10) -#define VEC_INT_PW2(a,pw) vec_sl(vec_splat_s32(8), vec_splat_u32(pw-3)) - - -//vec_sr(a,(vector unsigned int)num); - -// 0.0 0.0 0.0 0.0 -#define VEC_FLOAT_ZERO(a) vec_xor(a,a) - -// 0xffffffff 0xffffffff 0xffffffff 0xffffffff -#define VEC_FLOAT_ONES(a) (vector float)vec_cmpeq(a,a) - - - - - - - - - - -// arithmetic -inline vec_float add_vec( vec_float a, vec_float b) { return vec_add(a,b); } -inline vec_float add_scal(vec_float a, vec_float b) { return vec_add(a,b); } -inline vec_int add_vec( vec_int a, vec_int b) { return vec_add(a,b); } -inline vec_int add_scal(vec_int a, vec_int b) { return vec_add(a,b); } - - -inline vec_float sub_vec( vec_float a, vec_float b) { return vec_sub(a,b); } -inline vec_float sub_scal( vec_float a, vec_float b) { return vec_sub(a,b); } -inline vec_int sub_vec( vec_int a, vec_int b) { return vec_sub(a,b); } -inline vec_int sub_scal( vec_int a, vec_int b) { return vec_sub(a,b); } - - -inline vec_float mul_vec( vec_float a, vec_float b) { return vec_madd(a,b,(vec_float)(vec_splat_s32(int(0x00000000)))); } -inline vec_float mul_scal( vec_float a, vec_float b) { return vec_madd(a,b,(vec_float)(vec_splat_s32(int(0x00000000)))); } - - -// low 32 bits of a 32 * 32 bit multiplication: each double-word X and Y is broken down into two words, A & B and C & D: -// X = ( A << 16 ) + B -// Y = ( C << 16 ) + D -// then: -// X * Y = (( A << 16 ) + B ) * (( C << 16 ) + D ) -// X * Y = ( A*C << 32 ) + ( A*D << 16 ) + ( B*C << 16 ) + B*D -// the partial result A*C does not appear in the low 32 bits result so does not need to be computed ( however, if it's different -// from zero, then there is an overflow ) -// In this implementation A*D + B*C is computed in a single "vec_msum" - -inline vec_int mul_vec( vec_int a, vec_int b) { - const vector unsigned int VEC_SIXTEEN_UINT32 = vec_splat_u32(-16); - return (vector signed int)vec_add( vec_sl( vec_msum( (vector unsigned short)a, (vector unsigned short)(vec_rl( b, VEC_SIXTEEN_UINT32 ) ), vec_splat_u32(0) ), VEC_SIXTEEN_UINT32 ), vec_mulo( (vector unsigned short)a, (vector unsigned short)b ) ); - } - -inline vec_int mul_scal( vec_int a, vec_int b) { - const vector unsigned int VEC_SIXTEEN_UINT32 = vec_splat_u32(-16); - return (vector signed int)vec_add( vec_sl( vec_msum( (vector unsigned short)a, (vector unsigned short)(vec_rl( b, VEC_SIXTEEN_UINT32 ) ), vec_splat_u32(0) ), VEC_SIXTEEN_UINT32 ), vec_mulo( (vector unsigned short)a, (vector unsigned short)b ) ); - } - - -//inline vec_int mul_vec( vec_int a, vec_int b) { return (vec_int)vec_round(vec_madd((vec_float)(a),(vec_float)(b),(vec_float)(vec_splat_s32(int(0x00000000))))); } -//inline vec_int mul_scal( vec_int a, vec_int b) { return (vec_int)vec_round(vec_madd((vec_float)(a),(vec_float)(b),(vec_float)(vec_splat_s32(int(0x00000000))))); } - - - -inline vec_float div_vec( vec_float a, vec_float b) { return vec_madd(a,vec_re(b),(vec_float)(vec_splat_s32(int(0x00000000)))); } -inline vec_float div_scal( vec_float a, vec_float b) { return vec_madd(a,vec_re(b),(vec_float)(vec_splat_s32(int(0x00000000)))); } -inline vec_int div_vec( vec_int a, vec_int b) { return (vec_int)vec_round(vec_madd((vec_float)(a),vec_re((vec_float)(b)),(vec_float)(vec_splat_s32(int(0x00000000))))); } -inline vec_int div_scal( vec_int a, vec_int b) { return (vec_int)vec_round(vec_madd((vec_float)(a),vec_re((vec_float)(b)),(vec_float)(vec_splat_s32(int(0x00000000))))); } - - -inline vec_int mod_vec( vec_int a, vec_int N) { - - vec_int temp = vec_sub(a,N); - vec_int zero = vec_splat_s32(int(0x00000000)); - - vector bool int select = (vector bool int )(vec_xor(vec_cmpgt(temp,zero),vec_cmpeq(temp,zero))); // a - N >= 0 - - return vec_sel(a,temp,select); // if( a - N >=0 ) return a - N; else return a; -} - -inline vec_int mod_scal( vec_int a, vec_int N) { - - vec_int temp = vec_sub(a,N); - vec_int zero = vec_splat_s32(int(0x00000000)); - - vector bool int select = (vector bool int )(vec_xor(vec_cmpgt(temp,zero),vec_cmpeq(temp,zero))); // a - N >= 0 - - return vec_sel(a,temp,select); // if( a - N >=0 ) return a - N; else return a; -} - - - - -// return a*b + c -inline vec_float madd_vec( vec_float a, vec_float b, vec_float c) { return vec_madd(a,b,c); } -inline vec_float madd_scal( vec_float a, vec_float b, vec_float c) { return vec_madd(a,b,c); } - -// return - ( a*b - c ) -inline vec_float nmsub_vec( vec_float a, vec_float b, vec_float c) { return vec_nmsub(a,b,c); } -inline vec_float nmsub_scal( vec_float a, vec_float b, vec_float c) { return vec_nmsub(a,b,c); } - - -// return a*(1/b) + c -inline vec_float divadd_vec( vec_float a, vec_float b, vec_float c) { return vec_madd(a,vec_re(b),c); } -inline vec_float divadd_scal( vec_float a, vec_float b, vec_float c) { return vec_madd(a,vec_re(b),c); } - - -// return - ( a*(1/b) - c ) -inline vec_float divsub_vec( vec_float a, vec_float b, vec_float c) { return vec_nmsub(a,vec_re(b),c); } -inline vec_float divsub_scal( vec_float a, vec_float b, vec_float c) { return vec_nmsub(a,vec_re(b),c); } - - - - -// logic -inline vec_float and_vec( vec_float a, vec_float b) { return vec_and(a,b); } -inline vec_float and_scal( vec_float a, vec_float b) { return vec_and(a,b); } -inline vec_int and_vec( vec_int a, vec_int b) { return vec_and(a,b); } -inline vec_int and_scal( vec_int a, vec_int b) { return vec_and(a,b); } - -inline vec_float or_vec( vec_float a, vec_float b) { return vec_or(a,b); } -inline vec_float or_scal( vec_float a, vec_float b) { return vec_or(a,b); } -inline vec_int or_vec( vec_int a, vec_int b) { return vec_or(a,b); } -inline vec_int or_scal( vec_int a, vec_int b) { return vec_or(a,b); } - -inline vec_float xor_vec( vec_float a, vec_float b) { return vec_xor(a,b); } -inline vec_float xor_scal( vec_float a, vec_float b) { return vec_xor(a,b); } -inline vec_int xor_vec( vec_int a, vec_int b) { return vec_xor(a,b); } -inline vec_int xor_scal( vec_int a, vec_int b) { return vec_xor(a,b); } - - - -// shift left -inline vec_int shift_left_vec( vec_int a, vec_int num) { return vec_sl(a,(vector unsigned int)num); } -inline vec_int shift_left_scal( vec_int a, vec_int num) { return vec_sl(a,(vector unsigned int)num); } - -// shift ( and fill with the sign bit ) -inline vec_int shift_right_vec( vec_int a, vec_int num) { return vec_sra(a,(vector unsigned int)num); } -inline vec_int shift_right_scal( vec_int a, vec_int num) { return vec_sra(a,(vector unsigned int)num); } - -// shift ( and fill with 0's ) -//inline vec_int shift_right_vec_logical( vec_int a, int num) { return vec_sr(a, set_vec(num) ); } -//inline vec_int shift_right_scal_logical( vec_int a, int num) { return vec_sr(a, set_vec(num) ); ); } -inline vec_int shift_right_vec_logical( vec_int a, vec_int num) { return vec_sr(a,(vector unsigned int)num); } -inline vec_int shift_right_scal_logical( vec_int a, vec_int num) { return vec_sr(a,(vector unsigned int)num); } - - - - -// comparaison -inline vec_float gt_vec( vec_float a, vec_float b) { return (vector float)vec_cmpgt(a,b); } -inline vec_float gt_scal( vec_float a, vec_float b) { return (vector float)vec_cmpgt(a,b); } -inline vec_int gt_vec( vec_int a, vec_int b) { return (vector signed int)vec_cmpgt(a,b); } -inline vec_int gt_scal( vec_int a, vec_int b) { return (vector signed int)vec_cmpgt(a,b); } - - -inline vec_float lt_vec( vec_float a, vec_float b) { return (vector float)vec_cmplt(a,b); } -inline vec_float lt_scal( vec_float a, vec_float b) { return (vector float)vec_cmplt(a,b); } -inline vec_int lt_vec( vec_int a, vec_int b) { return (vector signed int)vec_cmplt(a,b); } -inline vec_int lt_scal( vec_int a, vec_int b) { return (vector signed int)vec_cmplt(a,b); } - - -inline vec_float ge_vec( vec_float a, vec_float b) { return (vector float)vec_cmpge(a,b); } -inline vec_float ge_scal( vec_float a, vec_float b) { return (vector float)vec_cmpge(a,b); } -inline vec_int ge_vec( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmpgt(a,b),vec_cmpeq(a,b)); } -inline vec_int ge_scal( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmpgt(a,b),vec_cmpeq(a,b)); } - - -inline vec_float le_vec( vec_float a, vec_float b) { return (vector float)vec_cmple(a,b); } -inline vec_float le_scal( vec_float a, vec_float b) { return (vector float)vec_cmple(a,b); } -inline vec_int le_vec( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmplt(a,b),vec_cmpeq(a,b)); } -inline vec_int le_scal( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmplt(a,b),vec_cmpeq(a,b)); } - - -inline vec_float eq_vec( vec_float a, vec_float b) { return (vector float)vec_cmpeq(a,b); } -inline vec_float eq_scal( vec_float a, vec_float b) { return (vector float)vec_cmpeq(a,b); } -inline vec_int eq_vec( vec_int a, vec_int b) { return (vector signed int)vec_cmpeq(a,b); } -inline vec_int eq_scal( vec_int a, vec_int b) { return (vector signed int)vec_cmpeq(a,b); } - - -inline vec_float neq_vec( vec_float a, vec_float b) { return (vector float)vec_xor(vec_cmpeq(a,b),vec_cmpeq(a,a)); } -inline vec_float neq_scal( vec_float a, vec_float b) { return (vector float)vec_xor(vec_cmpeq(a,b),vec_cmpeq(a,a)); } -inline vec_int neq_vec( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmpeq(a,b),vec_cmpeq(a,a)); } -inline vec_int neq_scal( vec_int a, vec_int b) { return (vector signed int)vec_xor(vec_cmpeq(a,b),vec_cmpeq(a,a)); } - - - -// memory -inline vec_float set_vec( vec_float a) { return a; } -inline vec_float set_vec( __vec_float a) { return a.v; } -inline vec_int set_vec( vec_int a) { return a; } -inline vec_int set_vec( __vec_int a) { return a.v; } - - - -inline vec_float set_vec( double a) { - float af; - af = (float)a; - vector float temp; temp = vec_lde(0,&af); - temp = vec_perm(temp,temp,vec_lvsl(0,&af)); - return vec_splat(temp,0); - //__vec_float temp; float af = float(a); temp.s[0]=af; temp.s[1]=af; temp.s[2]=af; temp.s[3]=af; return temp.v; -} -inline vec_float set_vec( float a) { - float af; - af = a; - vector float temp; temp = vec_lde(0,&af); - temp = vec_perm(temp,temp,vec_lvsl(0,&af)); - return vec_splat(temp,0); - //__vec_float temp; temp.s[0]=a; temp.s[1]=a; temp.s[2]=a; temp.s[3]=a; return temp.v; -} -inline vec_int set_vec( long int a) { - int ai; - ai = (int)a; - vector signed int temp; temp = vec_lde(0,&ai); - temp = vec_perm(temp,temp,vec_lvsl(0,&ai)); - return vec_splat(temp,0); - //__vec_int temp; int al = int(a); temp.s[0]=al; temp.s[1]=al; temp.s[2]=al; temp.s[3]=al; return temp.v; -} -inline vec_int set_vec( int a) { - int ai; - ai = a; - vector signed int temp; temp = vec_lde(0,&ai); - temp = vec_perm(temp,temp,vec_lvsl(0,&ai)); - return vec_splat(temp,0); - //__vec_int temp; temp.s[0]=a; temp.s[1]=a; temp.s[2]=a; temp.s[3]=a; return temp.v; -} -inline vec_int set_vec( short a) { - int ai; - ai = (int)a; - vector signed int temp; temp = vec_lde(0,&ai); - temp = vec_perm(temp,temp,vec_lvsl(0,&ai)); - return vec_splat(temp,0); - //__vec_int temp; int as = int(a); temp.s[0]=as; temp.s[1]=as; temp.s[2]=as; temp.s[3]=as; return temp.v; -} - -inline vec_float set_vec( double a, double b, double c, double d) { __vec_float temp; temp.s[0]=float(a); temp.s[1]=float(b); temp.s[2]=float(c); temp.s[3]=float(d); return temp.v; } -inline vec_float set_vec( float a, float b, float c, float d) { __vec_float temp; temp.s[0]=a; temp.s[1]=b; temp.s[2]=c; temp.s[3]=d; return temp.v; } -inline vec_int set_vec( int a, int b, int c, int d) { __vec_int temp; temp.s[0]=a; temp.s[1]=b; temp.s[2]=c; temp.s[3]=d; return temp.v; } -inline vec_int set_vec( long int a, long int b, long int c, long int d) { __vec_int temp; temp.s[0]=int(a); temp.s[1]=int(b); temp.s[2]=int(c); temp.s[3]=int(d); return temp.v; } -inline vec_int set_vec( short a, short b, short c, short d) { __vec_int temp; temp.s[0]=short(a); temp.s[1]=short(b); temp.s[2]=short(c); temp.s[3]=short(d); return temp.v; } - - -inline vec_float load_a_vec( float* a) { return vec_ld(0,a); } -//inline vec_float load_u_vec( float* a) { } -inline vec_int load_a_vec( int* a) { return vec_ld(0,a); } -//inline vec_int load_u_vec( int* a) { } - -inline void store_a_vec( float* a, vec_float b) { return vec_st(b,0,a); } -//inline void store_u_vec( float* a, vec_float b) { } -inline void store_a_vec( int* a, vec_int b) { return vec_st(b,0,a); } -//inline void store_u_vec( int* a, vec_int b) { } - - -inline vec_float load_scal(float* a) { vector float temp; temp = vec_lde(0,a); return vec_perm(temp,temp,vec_lvsl(0,a)); } -inline vec_int load_scal(int* a) { vector signed int temp; temp = vec_lde(0,a); return vec_perm(temp,temp,vec_lvsl(0,a)); } -inline void store_scal(float* a, vec_float content) { vec_float temp = vec_splat(content,0); return vec_ste(temp,0,a); } -inline void store_scal(int* a, vec_int content) { vec_int temp = vec_splat(content,0); return vec_ste(temp,0,a); } - - - - -inline vec_float REC0(vec_float a) { return vec_mergeh(a,a); } -inline vec_float REC1(vec_float a) { return vec_mergeh(a,a); } -inline vec_float REC2(vec_float a) { return vec_sld(vec_splat(a,0),a,12); } -inline vec_float REC3(vec_float a) { - vector float temp1 = vec_mergel(a,a); - vector float temp2 = vec_mergeh(a,a); - return vec_mergel(vec_mergel(temp1,temp2),vec_mergeh(temp1,temp2)); -} - -inline vec_int REC0(vec_int a) { return vec_mergeh(a,a); } -inline vec_int REC1(vec_int a) { return vec_mergeh(a,a); } -inline vec_int REC2(vec_int a) { return vec_sld(vec_splat(a,0),a,12); } -inline vec_int REC3(vec_int a) { - vector signed int temp1 = vec_mergel(a,a); - vector signed int temp2 = vec_mergeh(a,a); - return vec_mergel(vec_mergel(temp1,temp2),vec_mergeh(temp1,temp2)); -} - - -// scalar to vector: takes 4 vector which lower elements stands for a scalar value and rebuild a vector from these 4 scalar -inline vec_float SCAL2VEC(vec_float a0,vec_float a1,vec_float a2,vec_float a3) { - return vec_perm(vec_mergeh(a0,a1),vec_mergeh(a2,a3),vec_lvsl(8,(float*)(0))); - //return vec_perm(vec_perm(a0,a1,VEC_PERM0),vec_perm(a2,a3,VEC_PERM0),VEC_PERM1); -} -inline vec_int SCAL2VEC(vec_int a0,vec_int a1,vec_int a2,vec_int a3) { - return vec_perm(vec_mergeh(a0,a1),vec_mergeh(a2,a3),vec_lvsl(8,(int*)(0))); - //return vec_perm(vec_perm(a0,a1,VEC_PERM0),vec_perm(a2,a3,VEC_PERM0),VEC_PERM1); -} - -inline vec_float SCAL2VEC(float a0, float a1, float a2, float a3) { __vec_float temp; temp.s[0]=a0; temp.s[1]=a1; temp.s[2]=a2; temp.s[3]=a3; return temp.v; } -inline vec_int SCAL2VEC(int a0, int a1, int a2, int a3) { __vec_int temp; temp.s[0]=a0; temp.s[1]=a1; temp.s[2]=a2; temp.s[3]=a3; return temp.v; } - - -// vector to scalar: build a scalar vector from one element of the initial vector -inline vec_float VEC2SCALVEC0(vec_float a) { return a; } // return x,x,x,a0 // vec_splat(a,0) would return a0,a0,a0,a0 -inline vec_float VEC2SCALVEC1(vec_float a) { return vec_splat(a,1); } // return a1,a1,a1,a1 -inline vec_float VEC2SCALVEC2(vec_float a) { return vec_splat(a,2); } // return a2,a2,a2,a2 -inline vec_float VEC2SCALVEC3(vec_float a) { return vec_splat(a,3); } // return a3,a3,a3,a3 - -inline vec_int VEC2SCALVEC0(vec_int a) { return a; } // return x,x,x,a0 // vec_splat(a,0) would return a0,a0,a0,a0 -inline vec_int VEC2SCALVEC1(vec_int a) { return vec_splat(a,1); } // return a1,a1,a1,a1 -inline vec_int VEC2SCALVEC2(vec_int a) { return vec_splat(a,2); } // return a2,a2,a2,a2 -inline vec_int VEC2SCALVEC3(vec_int a) { return vec_splat(a,3); } // return a3,a3,a3,a3 - - -// vector to scalar: build a single scalar from a vector -inline float VEC2SCAL0(vec_float a) { float temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,0),0,&temp); return temp; } -inline float VEC2SCAL1(vec_float a) { float temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,1),0,&temp); return temp; } -inline float VEC2SCAL2(vec_float a) { float temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,2),0,&temp); return temp; } -inline float VEC2SCAL3(vec_float a) { float temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,3),0,&temp); return temp; } - -inline int VEC2SCAL0(vec_int a) { int temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,0),0,&temp); return temp; } -inline int VEC2SCAL1(vec_int a) { int temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,1),0,&temp); return temp; } -inline int VEC2SCAL2(vec_int a) { int temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,2),0,&temp); return temp; } -inline int VEC2SCAL3(vec_int a) { int temp __attribute__ ((aligned(16))); vec_ste(vec_splat(a,3),0,&temp); return temp; } - - - -// select: if( select == 0 ) then a ; else b ; -inline vec_float select_vec( vec_float select, vec_float a, vec_float b) { return vec_sel(a,b,(vector bool int)select ); } -inline vec_float select_scal( vec_float select, vec_float a, vec_float b) {return vec_sel(a,b,(vector bool int)select ); } -inline vec_float select_vec( vec_int select, vec_float a, vec_float b) { return vec_sel(a,b,(vector bool int)select ); } -inline vec_float select_scal( vec_int select, vec_float a, vec_float b) {return vec_sel(a,b,(vector bool int)select ); } -inline vec_int select_vec( vec_int select, vec_int a, vec_int b) { return vec_sel(a,b,(vector bool int)select ); } -inline vec_int select_scal( vec_int select, vec_int a, vec_int b) { return vec_sel(a,b,(vector bool int)select ); } - - - - - -// vectorial version of the "mem" Faust key-word -// result = { a[2] a[1] a[0] b[3] } -inline vec_float mem1_vec( vec_float a, vec_float b) { return vec_sld(b,a,12); } -inline vec_int mem1_vec( vec_int a, vec_int b) { return vec_sld(b,a,12); } - -// result = { a[1] a[0] b[3] b[2] } -inline vec_float mem2_vec( vec_float a, vec_float b) { return vec_sld(b,a,8); } -inline vec_int mem2_vec( vec_int a, vec_int b) { return vec_sld(b,a,8); } - -// result = { a[0] b[3] b[2] b[1] } -inline vec_float mem3_vec( vec_float a, vec_float b) { return vec_sld(b,a,4); } -inline vec_int mem3_vec( vec_int a, vec_int b) { return vec_sld(b,a,4); } - - - - - -// conversion -inline vec_float bool2float( vec_float a ) { return vec_and(a,(vec_float)(vec_splat_s32(int(0x00000001)))); } -inline vec_float bool2float( vec_int a ) { return (vec_float)(vec_and(a,vec_splat_s32(int(0x00000001)))); } - -inline vec_int bool2int( vec_int a) { return (vec_int)vec_and(a,vec_splat_s32(int(0x00000001))); } -inline vec_int bool2int( vec_float a ) { return (vec_int)vec_round(vec_and(a,(vec_float)(vec_splat_s32(int(0x00000001))))); } - -inline vec_int boolfloat2boolint( vec_float a ) { return (vector signed int)a; } -inline vec_float boolint2boolfloat( vec_int a ) { return (vector float)a; } - -inline vec_int float2int( vec_float a) { return (vec_int)vec_round(a); } -inline int float2int( float a ) { return int(a); } - -inline vec_float int2float( vec_int a) { return vec_ctf(a,0); } -inline float int2float( int a ) { return float(a); } - - -#endif