portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-04-09 21:50:45 +01:00
Code Releases Activity

first attempt to introduce tables using fixed-size; still incomplete

Browse Source
This commit is contained in:
nmeyer-ur 2020-05-07 22:01:19 +02:00
parent acff9d6ed2
commit 2b81cbe2c2
1 changed files with 49 additions and 61 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

110
Grid/simd/Grid_a64fx-fixedsize.h
View File

@ -49,6 +49,8 @@ typedef svfloat16_t vech __attribute__((arm_sve_vector_bits(512)));
typedef svfloat32_t vecf __attribute__((arm_sve_vector_bits(512))); typedef svfloat32_t vecf __attribute__((arm_sve_vector_bits(512)));
typedef svfloat64_t vecd __attribute__((arm_sve_vector_bits(512))); typedef svfloat64_t vecd __attribute__((arm_sve_vector_bits(512)));
typedef svuint32_t veci __attribute__((arm_sve_vector_bits(512))); typedef svuint32_t veci __attribute__((arm_sve_vector_bits(512)));
typedef svuint32_t lutf __attribute__((arm_sve_vector_bits(512))); // LUTs for float
typedef svuint64_t lutd __attribute__((arm_sve_vector_bits(512))); // LUTs for double
#else #else
#pragma error("Oops. Wrong or undefined SVE vector size?") #pragma error("Oops. Wrong or undefined SVE vector size?")
#endif /* __ARM_FEATURE_SVE_BITS */ #endif /* __ARM_FEATURE_SVE_BITS */
@ -109,21 +111,24 @@ struct acle<double>{
typedef svuint64_t svuint; typedef svuint64_t svuint;
static inline pred pg1(){return svptrue_b64();} static inline pred pg1(){return svptrue_b64();}
static inline vec<uint64_t> tbl_swap(){ static inline lutd tbl_swap(){
const vec<uint64_t> t = {1, 0, 3, 2, 5, 4, 7, 6}; const uint64_t t[8] = {1, 0, 3, 2, 5, 4, 7, 6};
return t; pred pg1 = svptrue_b64();
return svld1(pg1, t);
} }
static inline vec<uint64_t> tbl0(){ static inline lutd tbl0(){
const vec<uint64_t> t = {4, 5, 6, 7, 0, 1, 2, 3}; const uint64_t t[8] = {4, 5, 6, 7, 0, 1, 2, 3};
return t; pred pg1 = svptrue_b64();
return svld1(pg1, t);
} }
static inline vec<uint64_t> tbl1(){ static inline lutd tbl1(){
const vec<uint64_t> t = {2, 3, 0, 1, 6, 7, 4, 5}; const uint64_t t[8] = {2, 3, 0, 1, 6, 7, 4, 5};
return t; pred pg1 = svptrue_b64();
return svld1(pg1, t);
} }
static inline pred pg_even(){return svzip1_b64(svptrue_b64(), svpfalse_b());} static inline pred pg_even(){return svzip1_b64(svptrue_b64(), svpfalse_b());}
static inline pred pg_odd() {return svzip1_b64(svpfalse_b(), svptrue_b64());} static inline pred pg_odd() {return svzip1_b64(svpfalse_b(), svptrue_b64());}
static inline svfloat64_t zero(){return svdup_f64(0.);} static inline vecd zero(){return svdup_f64(0.);}
}; };
template <> template <>
@ -136,25 +141,29 @@ struct acle<float>{
static inline pred pg1(){return svptrue_b32();} static inline pred pg1(){return svptrue_b32();}
// exchange neighboring elements // exchange neighboring elements
static inline vec<uint32_t> tbl_swap(){ static inline lutf tbl_swap(){
const vec<uint32_t> t = {1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14}; const uint32_t t[16] = {1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14};
return t; pred pg1 = svptrue_b32();
return svld1(pg1, t);
} }
static inline vec<uint32_t> tbl0(){ static inline lutf tbl0(){
const vec<uint32_t> t = {8, 9, 10, 11, 12, 13, 14, 15, 0, 1, 2, 3, 4, 5, 6, 7}; const uint32_t t[16] = {8, 9, 10, 11, 12, 13, 14, 15, 0, 1, 2, 3, 4, 5, 6, 7};
return t; pred pg1 = svptrue_b32();
return svld1(pg1, t);
} }
static inline vec<uint32_t> tbl1(){ static inline vec<uint32_t> tbl1(){
const vec<uint32_t> t = {4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11}; const lutf = {4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11};
return t; pred pg1 = svptrue_b32();
return svld1(pg1, t);
} }
static inline vec<uint32_t> tbl2(){ static inline vec<uint32_t> tbl2(){
const vec<uint32_t> t = {2, 3, 0, 1, 6, 7, 4, 5, 10, 11, 8, 9, 14, 15, 12, 13}; const lutf = {2, 3, 0, 1, 6, 7, 4, 5, 10, 11, 8, 9, 14, 15, 12, 13};
return t; pred pg1 = svptrue_b32();
return svld1(pg1, t);
} }
static inline pred pg_even(){return svzip1_b32(svptrue_b32(), svpfalse_b());} static inline pred pg_even(){return svzip1_b32(svptrue_b32(), svpfalse_b());}
static inline pred pg_odd() {return svzip1_b32(svpfalse_b(), svptrue_b32());} static inline pred pg_odd() {return svzip1_b32(svpfalse_b(), svptrue_b32());}
static inline svfloat32_t zero(){return svdup_f32(0.);} static inline vecf zero(){return svdup_f32(0.);}
}; };
template <> template <>
@ -167,7 +176,7 @@ struct acle<uint16_t>{
static inline pred pg1(){return svptrue_b16();} static inline pred pg1(){return svptrue_b16();}
static inline pred pg_even(){return svzip1_b16(svptrue_b16(), svpfalse_b());} static inline pred pg_even(){return svzip1_b16(svptrue_b16(), svpfalse_b());}
static inline pred pg_odd() {return svzip1_b16(svpfalse_b(), svptrue_b16());} static inline pred pg_odd() {return svzip1_b16(svpfalse_b(), svptrue_b16());}
static inline svfloat16_t zero(){return svdup_f16(0.);} static inline vech zero(){return svdup_f16(0.);}
}; };
template <> template <>
@ -180,7 +189,6 @@ struct acle<Integer>{
//static inline svbool_t pg1(){return svptrue_b16();} //static inline svbool_t pg1(){return svptrue_b16();}
static inline pred pg1(){return svptrue_b32();} static inline pred pg1(){return svptrue_b32();}
static inline pred pg2(){return svptrue_pat_b32(SV_VL8);}
static inline pred pg_even(){return svzip1_b32(svptrue_b32(), svpfalse_b());} static inline pred pg_even(){return svzip1_b32(svptrue_b32(), svpfalse_b());}
static inline pred pg_odd() {return svzip1_b32(svpfalse_b(), svptrue_b32());} static inline pred pg_odd() {return svzip1_b32(svpfalse_b(), svptrue_b32());}
}; };
@ -416,24 +424,20 @@ struct Conj{
struct TimesMinusI{ struct TimesMinusI{
// Complex float // Complex float
inline vecf operator()(vecf a, vecf b){ inline vecf operator()(vecf a, vecf b){
const vec<typename acle<float>::uint> tbl_swap = acle<float>::tbl_swap(); lutf tbl_swap = acle<float>::tbl_swap();
pred pg1 = acle<float>::pg1(); pred pg1 = acle<float>::pg1();
pred pg_odd = acle<float>::pg_odd(); pred pg_odd = acle<float>::pg_odd();
typename acle<float>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v); vecf a_v = svtbl(a, tbl_swap);
vecf a_v = svld1(pg1, a.v);
a_v = svtbl(a_v, tbl_swap_v);
return svneg_x(pg_odd, a_v); return svneg_x(pg_odd, a_v);
} }
// Complex double // Complex double
inline vecd operator()(vecd a, vecd b){ inline vecd operator()(vecd a, vecd b){
const vec<typename acle<double>::uint> tbl_swap = acle<double>::tbl_swap(); lutd tbl_swap = acle<double>::tbl_swap();
pred pg1 = acle<double>::pg1(); pred pg1 = acle<double>::pg1();
pred pg_odd = acle<double>::pg_odd(); pred pg_odd = acle<double>::pg_odd();
typename acle<double>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v); vecd a_v = svtbl(a, tbl_swap);
vecd a_v = svld1(pg1, a.v);
a_v = svtbl(a_v, tbl_swap_v);
return svneg_x(pg_odd, a_v); return svneg_x(pg_odd, a_v);
} }
}; };
@ -441,24 +445,20 @@ struct TimesMinusI{
struct TimesI{ struct TimesI{
// Complex float // Complex float
inline vecf operator()(vecf a, vecf b){ inline vecf operator()(vecf a, vecf b){
const vec<typename acle<float>::uint> tbl_swap = acle<T>::tbl_swap(); lutf tbl_swap = acle<T>::tbl_swap();
pred pg1 = acle<float>::pg1(); pred pg1 = acle<float>::pg1();
pred pg_even = acle<float>::pg_even(); pred pg_even = acle<float>::pg_even();
typename acle<float>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v); vecf a_v = svtbl(a, tbl_swap);
vecf a_v = svld1(pg1, a.v);
a_v = svtbl(a_v, tbl_swap_v);
return svneg_x(pg_even, a_v); return svneg_x(pg_even, a_v);
} }
// Complex double // Complex double
inline vecd operator()(vecd a, vecd b){ inline vecd operator()(vecd a, vecd b){
const vec<typename acle<double>::uint> tbl_swap = acle<double>::tbl_swap(); lutd tbl_swap = acle<double>::tbl_swap();
pred pg1 = acle<double>::pg1(); pred pg1 = acle<double>::pg1();
pred pg_even = acle<double>::pg_even(); pred pg_even = acle<double>::pg_even();
typename acle<double>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v); vecd a_v = svtbl(a, tbl_swap);
vecd a_v = svld1(pg1, a.v);
a_v = svtbl(a_v, tbl_swap_v);
return svneg_x(pg_even, a_v); return svneg_x(pg_even, a_v);
} }
}; };
@ -642,22 +642,16 @@ struct Permute{
return svext(in, in, (uint64_t)(16u / 2u)); return svext(in, in, (uint64_t)(16u / 2u));
} }
static inline vecf Permute1(vecf in) { static inline vecf Permute1(vecf in) {
const vec<typename acle<float>::uint> tbl_swap = acle<float>::tbl1(); lutf tbl_swap = acle<float>::tbl1();
pred pg1 = acle<float>::pg1(); return svtbl(in, tbl_swap);
typename acle<float>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v);
return svtbl(in, tbl_swap_v);
} }
static inline vecf Permute2(vecf in) { static inline vecf Permute2(vecf in) {
const vec<typename acle<float>::uint> tbl_swap = acle<float>::tbl2(); lutf tbl_swap = acle<float>::tbl2();
pred pg1 = acle<float>::pg1(); return svtbl(in, tbl_swap);
typename acle<float>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v);
return svtbl(in, tbl_swap_v);
} }
static inline vecf Permute3(vecf in) { static inline vecf Permute3(vecf in) {
const vec<typename acle<float>::uint> tbl_swap = acle<float>::tbl_swap(); lutf tbl_swap = acle<float>::tbl_swap();
pred pg1 = acle<float>::pg1(); return svtbl(in, tbl_swap);
typename acle<float>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v);
return svtbl(in, tbl_swap_v);
} }
// double // double
@ -665,17 +659,12 @@ struct Permute{
return svext(in, in, (uint64_t)(8u / 2u)); return svext(in, in, (uint64_t)(8u / 2u));
} }
static inline vecd Permute1(vecd in) { static inline vecd Permute1(vecd in) {
lutd tbl_swap = acle<double>::tbl1();
const vec<typename acle<double>::uint> tbl_swap = acle<double>::tbl1(); return svtbl(in, tbl_swap);
pred pg1 = acle<double>::pg1();
typename acle<double>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v);
return svtbl(in, tbl_swap_v);
} }
static inline vecd Permute2(vecd in) { static inline vecd Permute2(vecd in) {
const vec<typename acle<double>::uint> tbl_swap = acle<double>::tbl_swap(); lutd tbl_swap = acle<double>::tbl_swap();
pred pg1 = acle<double>::pg1(); return svtbl(in, tbl_swap);
typename acle<double>::svuint tbl_swap_v = svld1(pg1, tbl_swap.v);
return svtbl(in, tbl_swap_v);
} }
static inline vecd Permute3(vecd in) { static inline vecd Permute3(vecd in) {
return in; return in;
@ -776,7 +765,6 @@ inline Grid::RealD Reduce<Grid::RealD, vecd>::operator()(vecd in){
pred pg1 = acle<double>::pg1(); pred pg1 = acle<double>::pg1();
return svred(pg1, in); return svred(pg1, in);
} }
//Integer Reduce //Integer Reduce
template <> template <>
inline Integer Reduce<Integer, veci>::operator()(veci in){ inline Integer Reduce<Integer, veci>::operator()(veci in){