.
This commit is contained in:
x
183
vp4dc.c
183
vp4dc.c
@ -1,183 +0,0 @@
|
||||
/**
|
||||
Copyright (C) powturbo 2013-2016
|
||||
GPL v2 License
|
||||
|
||||
This program is free software; you can redistribute it and/or modify
|
||||
it under the terms of the GNU General Public License as published by
|
||||
the Free Software Foundation; either version 2 of the License, or
|
||||
(at your option) any later version.
|
||||
|
||||
This program is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
GNU General Public License for more details.
|
||||
|
||||
You should have received a copy of the GNU General Public License along
|
||||
with this program; if not, write to the Free Software Foundation, Inc.,
|
||||
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
||||
|
||||
- homepage : https://sites.google.com/site/powturbo/
|
||||
- github : https://github.com/powturbo
|
||||
- twitter : https://twitter.com/powturbo
|
||||
- email : powturbo [_AT_] gmail [_DOT_] com
|
||||
**/
|
||||
// "Integer Compression" Turbo PforDelta
|
||||
#ifndef USIZE
|
||||
#include <stdint.h>
|
||||
|
||||
#include "conf.h"
|
||||
#include "bitpack.h"
|
||||
#include "vp4dc.h"
|
||||
|
||||
#define PAD8(__x) ( (((__x)+8-1)/8) )
|
||||
//------------------------------------------
|
||||
#define P4DSIZE 128 //64 //
|
||||
|
||||
#define P4D p4d
|
||||
#define P4DENC p4denc
|
||||
#define P4DNENC p4dnenc
|
||||
#define P4DE p4de
|
||||
|
||||
#define BITPACK bitpack
|
||||
#define USIZE 32
|
||||
#include __FILE__
|
||||
#undef USIZE
|
||||
|
||||
#define USIZE 16
|
||||
#include __FILE__
|
||||
#undef USIZE
|
||||
|
||||
#define USIZE 64
|
||||
#include __FILE__
|
||||
#undef USIZE
|
||||
|
||||
#undef BITPACK
|
||||
#undef P4DENC
|
||||
#undef P4DNENC
|
||||
#undef P4DE
|
||||
#undef P4D
|
||||
//------------------------------------------
|
||||
#define P4DENC p4dencv
|
||||
#define P4DNENC p4dnencv
|
||||
#define P4DE p4dev
|
||||
#define BITPACK bitpackv
|
||||
#define USIZE 32
|
||||
#include __FILE__
|
||||
#undef USIZE
|
||||
|
||||
#else
|
||||
#pragma clang diagnostic push
|
||||
#pragma clang diagnostic ignored "-Wparentheses"
|
||||
|
||||
#define uint_t TEMPLATE3(uint, USIZE, _t)
|
||||
#define P4DN (P4DSIZE/64)
|
||||
|
||||
#ifdef P4D
|
||||
//#define MINCALC(_i_) { l = PAD8(n*_i_) + PAD8(x*(bx-_i_)); x += cnt[_i_]; if(unlikely(l < ml)) b = _i_, ml = l; }
|
||||
#define MINCALC(_i_) { l = PAD8(n*_i_) + PAD8(x*(bx-_i_)); x += cnt[_i_]; unlikely(l < ml)?(ml=l,b=_i_):(ml=ml,b=b); }
|
||||
unsigned TEMPLATE2(P4D, USIZE)(uint_t *__restrict in, unsigned n, unsigned *pbx) {
|
||||
uint_t *ip,b=0; int i,ml,l; unsigned x, bx, cnt[USIZE+1] = {0};
|
||||
|
||||
for(ip = in; ip != in+(n&~3); ) {
|
||||
++cnt[TEMPLATE2(bsr, USIZE)(*ip)]; b |= *ip++;
|
||||
++cnt[TEMPLATE2(bsr, USIZE)(*ip)]; b |= *ip++;
|
||||
++cnt[TEMPLATE2(bsr, USIZE)(*ip)]; b |= *ip++;
|
||||
++cnt[TEMPLATE2(bsr, USIZE)(*ip)]; b |= *ip++;
|
||||
}
|
||||
while(ip != in+n) ++cnt[TEMPLATE2(bsr, USIZE)(*ip)], b |= *ip++;
|
||||
b = TEMPLATE2(bsr, USIZE)(b);
|
||||
|
||||
bx = b; ml = PAD8(n*b)+1-2-P4DN*8; x = cnt[b];
|
||||
#if USIZE > 32
|
||||
for(i = b-1; i >= 0; --i) MINCALC(i); // l = PAD8(n*i) + PAD8(x*(bx-i)); x += cnt[i]; if(unlikely(l < ml)) b = i, ml = l;
|
||||
#else
|
||||
switch(b-1) {
|
||||
case 31: MINCALC(31);
|
||||
case 30: MINCALC(30);
|
||||
case 29: MINCALC(29);
|
||||
case 28: MINCALC(28);
|
||||
case 27: MINCALC(27);
|
||||
case 26: MINCALC(26);
|
||||
case 25: MINCALC(25);
|
||||
case 24: MINCALC(24);
|
||||
case 23: MINCALC(23);
|
||||
case 22: MINCALC(22);
|
||||
case 21: MINCALC(21);
|
||||
case 20: MINCALC(20);
|
||||
|
||||
case 19: MINCALC(19);
|
||||
case 18: MINCALC(18);
|
||||
case 17: MINCALC(17);
|
||||
case 16: MINCALC(16);
|
||||
case 15: MINCALC(15);
|
||||
case 14: MINCALC(14);
|
||||
case 13: MINCALC(13);
|
||||
case 12: MINCALC(12);
|
||||
case 11: MINCALC(11);
|
||||
case 10: MINCALC(10);
|
||||
|
||||
case 9: MINCALC( 9);
|
||||
case 8: MINCALC( 8);
|
||||
case 7: MINCALC( 7);
|
||||
case 6: MINCALC( 6);
|
||||
case 5: MINCALC( 5);
|
||||
case 4: MINCALC( 4);
|
||||
case 3: MINCALC( 3);
|
||||
case 2: MINCALC( 2);
|
||||
case 1: MINCALC( 1);
|
||||
case 0: MINCALC( 0);
|
||||
}
|
||||
#endif
|
||||
*pbx = bx - b;
|
||||
return b;
|
||||
}
|
||||
#endif
|
||||
|
||||
unsigned char *TEMPLATE2(P4DE, USIZE)(uint_t *__restrict in, unsigned n, unsigned char *__restrict out, unsigned b, unsigned bx) { unsigned i, xn, c;
|
||||
if(!bx || b==USIZE) return TEMPLATE2(BITPACK, USIZE)(in, n, out, b);
|
||||
|
||||
uint_t msk = (1ull << b)-1,_in[P4DSIZE], inx[P4DSIZE*2];
|
||||
unsigned long long xmap[P4DN];
|
||||
unsigned miss[P4DSIZE];
|
||||
#if P4DN == 2
|
||||
xmap[0] = xmap[1] = 0;
|
||||
#else
|
||||
for(i = 0; i < P4DN; i++) xmap[i] = 0;
|
||||
#endif
|
||||
for(xn = i = 0; i != n&~3; ) {
|
||||
miss[xn] = i; xn += in[i] > msk; _in[i] = in[i] & msk; i++;
|
||||
miss[xn] = i; xn += in[i] > msk; _in[i] = in[i] & msk; i++;
|
||||
miss[xn] = i; xn += in[i] > msk; _in[i] = in[i] & msk; i++;
|
||||
miss[xn] = i; xn += in[i] > msk; _in[i] = in[i] & msk; i++;
|
||||
}
|
||||
while(i != n) { miss[xn] = i; xn += in[i] > msk; _in[i] = in[i] & msk; i++; }
|
||||
for(i = 0; i != xn; ++i) {
|
||||
c = miss[i];
|
||||
xmap[c>>6] |= (1ull << (c&0x3f));
|
||||
inx[i] = in[c] >> b;
|
||||
}
|
||||
#if P4DN == 2
|
||||
*(unsigned long long *)out = xmap[0]; out += 8;
|
||||
*(unsigned long long *)out = xmap[1]; out += 8;
|
||||
#else
|
||||
for(i=0;i < P4DN; i++) { *(unsigned long long *)out = xmap[i]; out += 8; }
|
||||
#endif
|
||||
out = TEMPLATE2(bitpack, USIZE)(inx, xn, out, bx);
|
||||
return TEMPLATE2(BITPACK, USIZE)(_in, n, out, b);
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(P4DENC, USIZE)(uint_t *__restrict in, unsigned n, unsigned char *__restrict out) {
|
||||
unsigned bx,b = TEMPLATE2(p4d, USIZE)(in, n, &bx);
|
||||
P4DSAVE(out, b, bx);
|
||||
return TEMPLATE2(P4DE, USIZE)(in, n, out, b, bx);
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(P4DNENC, USIZE)(uint_t *__restrict in, unsigned n, unsigned char *__restrict out) {
|
||||
uint_t *ip;
|
||||
for(ip = in; ip < in+n; ip += P4DSIZE) { unsigned l = (in+n) - ip; l = min(l, P4DSIZE);
|
||||
out = TEMPLATE2(P4DENC, USIZE)(ip, l, out);
|
||||
}
|
||||
return out;
|
||||
}
|
||||
#pragma clang diagnostic pop
|
||||
#endif
|
||||
57
vp4dc.h
57
vp4dc.h
@ -1,57 +0,0 @@
|
||||
/**
|
||||
Copyright (C) powturbo 2013-2015
|
||||
GPL v2 License
|
||||
|
||||
This program is free software; you can redistribute it and/or modify
|
||||
it under the terms of the GNU General Public License as published by
|
||||
the Free Software Foundation; either version 2 of the License, or
|
||||
(at your option) any later version.
|
||||
|
||||
This program is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
GNU General Public License for more details.
|
||||
|
||||
You should have received a copy of the GNU General Public License along
|
||||
with this program; if not, write to the Free Software Foundation, Inc.,
|
||||
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
||||
|
||||
- homepage : https://sites.google.com/site/powturbo/
|
||||
- github : https://github.com/powturbo
|
||||
- twitter : https://twitter.com/powturbo
|
||||
- email : powturbo [_AT_] gmail [_DOT_] com
|
||||
**/
|
||||
// vp4dc.h - "Integer Compression" TurboPfor (see vp4dd.h for decompression)
|
||||
|
||||
#define P4DEB(__b) (__b << 1)
|
||||
#define P4DEBX(__b, __bx) (__bx << 8 | __b << 1 | 1)
|
||||
#define P4DSAVE(__out, __b, __bx) do { if(!__bx) *__out++ = P4DEB(__b);else *(unsigned short *)__out = P4DEBX(__b, __bx), __out += 2; } while(0)
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
// compress integer array with n values to the buffer out. Return value = end of compressed buffer out
|
||||
unsigned char *p4denc32(unsigned *__restrict in, unsigned n, unsigned char *__restrict out);
|
||||
unsigned char *p4denc16(unsigned short *__restrict in, unsigned n, unsigned char *__restrict out);
|
||||
unsigned char *p4denc64(uint64_t *__restrict in, unsigned n, unsigned char *__restrict out);
|
||||
|
||||
// SIMD compress integer array with n values to the buffer out. Return value = end of compressed buffer out
|
||||
unsigned char *p4dencv32(unsigned *__restrict in, unsigned n, unsigned char *__restrict out);
|
||||
//unsigned char *p4denc16(unsigned short *__restrict in, unsigned n, unsigned char *__restrict out);
|
||||
|
||||
// calculate the best bit size b and bx.
|
||||
unsigned p4d32(unsigned *__restrict in, unsigned n, unsigned *pbx);
|
||||
unsigned p4d16(unsigned short *__restrict in, unsigned n, unsigned *pbx);
|
||||
unsigned p4d64(uint64_t *__restrict in, unsigned n, unsigned *pbx);
|
||||
|
||||
// same as p4denc, but with b and bx as parameters. Call after p4d32/p4d16
|
||||
unsigned char *p4de32(unsigned *__restrict in, unsigned n, unsigned char *__restrict out, unsigned b, unsigned bx);
|
||||
unsigned char *p4de16(unsigned short *__restrict in, unsigned n, unsigned char *__restrict out, unsigned b, unsigned bx);
|
||||
unsigned char *p4de64(uint64_t *__restrict in, unsigned n, unsigned char *__restrict out, unsigned b, unsigned bx);
|
||||
|
||||
unsigned char *p4dev32(unsigned *__restrict in, unsigned n, unsigned char *__restrict out, unsigned b, unsigned bx);
|
||||
|
||||
#ifdef __cplusplus
|
||||
}
|
||||
#endif
|
||||
590
vp4dd.c
590
vp4dd.c
@ -1,590 +0,0 @@
|
||||
/**
|
||||
Copyright (C) powturbo 2013-2016
|
||||
GPL v2 License
|
||||
|
||||
This program is free software; you can redistribute it and/or modify
|
||||
it under the terms of the GNU General Public License as published by
|
||||
the Free Software Foundation; either version 2 of the License, or
|
||||
(at your option) any later version.
|
||||
|
||||
This program is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
GNU General Public License for more details.
|
||||
|
||||
You should have received a copy of the GNU General Public License along
|
||||
with this program; if not, write to the Free Software Foundation, Inc.,
|
||||
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
||||
|
||||
- homepage : https://sites.google.com/site/powturbo/
|
||||
- github : https://github.com/powturbo
|
||||
- twitter : https://twitter.com/powturbo
|
||||
- email : powturbo [_AT_] gmail [_DOT_] com
|
||||
**/
|
||||
// vp4dd.c - "Integer Compression" Turbo PforDelta
|
||||
#ifndef USIZE
|
||||
#include <stdint.h>
|
||||
|
||||
#include "conf.h"
|
||||
#include "bitunpack.h"
|
||||
#include "bitutil.h"
|
||||
#include "vp4dd.h"
|
||||
|
||||
#define PAD8(__x) ( (((__x)+8-1)/8) )
|
||||
|
||||
//#define __AVX2__ON // disabled (tested w/ notebook cpu i5-4200u and skylake i6700-3.7GHz, but slower than SSE3)
|
||||
|
||||
#if defined(__AVX2__) && defined(__AVX2__ON)
|
||||
#include <immintrin.h>
|
||||
static ALIGNED(unsigned char, shuffles[256][8], 32) = {
|
||||
{ 0,0,0,0,0,0,0,0 },
|
||||
{ 0,1,1,1,1,1,1,1 },
|
||||
{ 1,0,1,1,1,1,1,1 },
|
||||
{ 0,1,2,2,2,2,2,2 },
|
||||
{ 1,1,0,1,1,1,1,1 },
|
||||
{ 0,2,1,2,2,2,2,2 },
|
||||
{ 2,0,1,2,2,2,2,2 },
|
||||
{ 0,1,2,3,3,3,3,3 },
|
||||
{ 1,1,1,0,1,1,1,1 },
|
||||
{ 0,2,2,1,2,2,2,2 },
|
||||
{ 2,0,2,1,2,2,2,2 },
|
||||
{ 0,1,3,2,3,3,3,3 },
|
||||
{ 2,2,0,1,2,2,2,2 },
|
||||
{ 0,3,1,2,3,3,3,3 },
|
||||
{ 3,0,1,2,3,3,3,3 },
|
||||
{ 0,1,2,3,4,4,4,4 },
|
||||
{ 1,1,1,1,0,1,1,1 },
|
||||
{ 0,2,2,2,1,2,2,2 },
|
||||
{ 2,0,2,2,1,2,2,2 },
|
||||
{ 0,1,3,3,2,3,3,3 },
|
||||
{ 2,2,0,2,1,2,2,2 },
|
||||
{ 0,3,1,3,2,3,3,3 },
|
||||
{ 3,0,1,3,2,3,3,3 },
|
||||
{ 0,1,2,4,3,4,4,4 },
|
||||
{ 2,2,2,0,1,2,2,2 },
|
||||
{ 0,3,3,1,2,3,3,3 },
|
||||
{ 3,0,3,1,2,3,3,3 },
|
||||
{ 0,1,4,2,3,4,4,4 },
|
||||
{ 3,3,0,1,2,3,3,3 },
|
||||
{ 0,4,1,2,3,4,4,4 },
|
||||
{ 4,0,1,2,3,4,4,4 },
|
||||
{ 0,1,2,3,4,5,5,5 },
|
||||
{ 1,1,1,1,1,0,1,1 },
|
||||
{ 0,2,2,2,2,1,2,2 },
|
||||
{ 2,0,2,2,2,1,2,2 },
|
||||
{ 0,1,3,3,3,2,3,3 },
|
||||
{ 2,2,0,2,2,1,2,2 },
|
||||
{ 0,3,1,3,3,2,3,3 },
|
||||
{ 3,0,1,3,3,2,3,3 },
|
||||
{ 0,1,2,4,4,3,4,4 },
|
||||
{ 2,2,2,0,2,1,2,2 },
|
||||
{ 0,3,3,1,3,2,3,3 },
|
||||
{ 3,0,3,1,3,2,3,3 },
|
||||
{ 0,1,4,2,4,3,4,4 },
|
||||
{ 3,3,0,1,3,2,3,3 },
|
||||
{ 0,4,1,2,4,3,4,4 },
|
||||
{ 4,0,1,2,4,3,4,4 },
|
||||
{ 0,1,2,3,5,4,5,5 },
|
||||
{ 2,2,2,2,0,1,2,2 },
|
||||
{ 0,3,3,3,1,2,3,3 },
|
||||
{ 3,0,3,3,1,2,3,3 },
|
||||
{ 0,1,4,4,2,3,4,4 },
|
||||
{ 3,3,0,3,1,2,3,3 },
|
||||
{ 0,4,1,4,2,3,4,4 },
|
||||
{ 4,0,1,4,2,3,4,4 },
|
||||
{ 0,1,2,5,3,4,5,5 },
|
||||
{ 3,3,3,0,1,2,3,3 },
|
||||
{ 0,4,4,1,2,3,4,4 },
|
||||
{ 4,0,4,1,2,3,4,4 },
|
||||
{ 0,1,5,2,3,4,5,5 },
|
||||
{ 4,4,0,1,2,3,4,4 },
|
||||
{ 0,5,1,2,3,4,5,5 },
|
||||
{ 5,0,1,2,3,4,5,5 },
|
||||
{ 0,1,2,3,4,5,6,6 },
|
||||
{ 1,1,1,1,1,1,0,1 },
|
||||
{ 0,2,2,2,2,2,1,2 },
|
||||
{ 2,0,2,2,2,2,1,2 },
|
||||
{ 0,1,3,3,3,3,2,3 },
|
||||
{ 2,2,0,2,2,2,1,2 },
|
||||
{ 0,3,1,3,3,3,2,3 },
|
||||
{ 3,0,1,3,3,3,2,3 },
|
||||
{ 0,1,2,4,4,4,3,4 },
|
||||
{ 2,2,2,0,2,2,1,2 },
|
||||
{ 0,3,3,1,3,3,2,3 },
|
||||
{ 3,0,3,1,3,3,2,3 },
|
||||
{ 0,1,4,2,4,4,3,4 },
|
||||
{ 3,3,0,1,3,3,2,3 },
|
||||
{ 0,4,1,2,4,4,3,4 },
|
||||
{ 4,0,1,2,4,4,3,4 },
|
||||
{ 0,1,2,3,5,5,4,5 },
|
||||
{ 2,2,2,2,0,2,1,2 },
|
||||
{ 0,3,3,3,1,3,2,3 },
|
||||
{ 3,0,3,3,1,3,2,3 },
|
||||
{ 0,1,4,4,2,4,3,4 },
|
||||
{ 3,3,0,3,1,3,2,3 },
|
||||
{ 0,4,1,4,2,4,3,4 },
|
||||
{ 4,0,1,4,2,4,3,4 },
|
||||
{ 0,1,2,5,3,5,4,5 },
|
||||
{ 3,3,3,0,1,3,2,3 },
|
||||
{ 0,4,4,1,2,4,3,4 },
|
||||
{ 4,0,4,1,2,4,3,4 },
|
||||
{ 0,1,5,2,3,5,4,5 },
|
||||
{ 4,4,0,1,2,4,3,4 },
|
||||
{ 0,5,1,2,3,5,4,5 },
|
||||
{ 5,0,1,2,3,5,4,5 },
|
||||
{ 0,1,2,3,4,6,5,6 },
|
||||
{ 2,2,2,2,2,0,1,2 },
|
||||
{ 0,3,3,3,3,1,2,3 },
|
||||
{ 3,0,3,3,3,1,2,3 },
|
||||
{ 0,1,4,4,4,2,3,4 },
|
||||
{ 3,3,0,3,3,1,2,3 },
|
||||
{ 0,4,1,4,4,2,3,4 },
|
||||
{ 4,0,1,4,4,2,3,4 },
|
||||
{ 0,1,2,5,5,3,4,5 },
|
||||
{ 3,3,3,0,3,1,2,3 },
|
||||
{ 0,4,4,1,4,2,3,4 },
|
||||
{ 4,0,4,1,4,2,3,4 },
|
||||
{ 0,1,5,2,5,3,4,5 },
|
||||
{ 4,4,0,1,4,2,3,4 },
|
||||
{ 0,5,1,2,5,3,4,5 },
|
||||
{ 5,0,1,2,5,3,4,5 },
|
||||
{ 0,1,2,3,6,4,5,6 },
|
||||
{ 3,3,3,3,0,1,2,3 },
|
||||
{ 0,4,4,4,1,2,3,4 },
|
||||
{ 4,0,4,4,1,2,3,4 },
|
||||
{ 0,1,5,5,2,3,4,5 },
|
||||
{ 4,4,0,4,1,2,3,4 },
|
||||
{ 0,5,1,5,2,3,4,5 },
|
||||
{ 5,0,1,5,2,3,4,5 },
|
||||
{ 0,1,2,6,3,4,5,6 },
|
||||
{ 4,4,4,0,1,2,3,4 },
|
||||
{ 0,5,5,1,2,3,4,5 },
|
||||
{ 5,0,5,1,2,3,4,5 },
|
||||
{ 0,1,6,2,3,4,5,6 },
|
||||
{ 5,5,0,1,2,3,4,5 },
|
||||
{ 0,6,1,2,3,4,5,6 },
|
||||
{ 6,0,1,2,3,4,5,6 },
|
||||
{ 0,1,2,3,4,5,6,7 },
|
||||
{ 1,1,1,1,1,1,1,0 },
|
||||
{ 0,2,2,2,2,2,2,1 },
|
||||
{ 2,0,2,2,2,2,2,1 },
|
||||
{ 0,1,3,3,3,3,3,2 },
|
||||
{ 2,2,0,2,2,2,2,1 },
|
||||
{ 0,3,1,3,3,3,3,2 },
|
||||
{ 3,0,1,3,3,3,3,2 },
|
||||
{ 0,1,2,4,4,4,4,3 },
|
||||
{ 2,2,2,0,2,2,2,1 },
|
||||
{ 0,3,3,1,3,3,3,2 },
|
||||
{ 3,0,3,1,3,3,3,2 },
|
||||
{ 0,1,4,2,4,4,4,3 },
|
||||
{ 3,3,0,1,3,3,3,2 },
|
||||
{ 0,4,1,2,4,4,4,3 },
|
||||
{ 4,0,1,2,4,4,4,3 },
|
||||
{ 0,1,2,3,5,5,5,4 },
|
||||
{ 2,2,2,2,0,2,2,1 },
|
||||
{ 0,3,3,3,1,3,3,2 },
|
||||
{ 3,0,3,3,1,3,3,2 },
|
||||
{ 0,1,4,4,2,4,4,3 },
|
||||
{ 3,3,0,3,1,3,3,2 },
|
||||
{ 0,4,1,4,2,4,4,3 },
|
||||
{ 4,0,1,4,2,4,4,3 },
|
||||
{ 0,1,2,5,3,5,5,4 },
|
||||
{ 3,3,3,0,1,3,3,2 },
|
||||
{ 0,4,4,1,2,4,4,3 },
|
||||
{ 4,0,4,1,2,4,4,3 },
|
||||
{ 0,1,5,2,3,5,5,4 },
|
||||
{ 4,4,0,1,2,4,4,3 },
|
||||
{ 0,5,1,2,3,5,5,4 },
|
||||
{ 5,0,1,2,3,5,5,4 },
|
||||
{ 0,1,2,3,4,6,6,5 },
|
||||
{ 2,2,2,2,2,0,2,1 },
|
||||
{ 0,3,3,3,3,1,3,2 },
|
||||
{ 3,0,3,3,3,1,3,2 },
|
||||
{ 0,1,4,4,4,2,4,3 },
|
||||
{ 3,3,0,3,3,1,3,2 },
|
||||
{ 0,4,1,4,4,2,4,3 },
|
||||
{ 4,0,1,4,4,2,4,3 },
|
||||
{ 0,1,2,5,5,3,5,4 },
|
||||
{ 3,3,3,0,3,1,3,2 },
|
||||
{ 0,4,4,1,4,2,4,3 },
|
||||
{ 4,0,4,1,4,2,4,3 },
|
||||
{ 0,1,5,2,5,3,5,4 },
|
||||
{ 4,4,0,1,4,2,4,3 },
|
||||
{ 0,5,1,2,5,3,5,4 },
|
||||
{ 5,0,1,2,5,3,5,4 },
|
||||
{ 0,1,2,3,6,4,6,5 },
|
||||
{ 3,3,3,3,0,1,3,2 },
|
||||
{ 0,4,4,4,1,2,4,3 },
|
||||
{ 4,0,4,4,1,2,4,3 },
|
||||
{ 0,1,5,5,2,3,5,4 },
|
||||
{ 4,4,0,4,1,2,4,3 },
|
||||
{ 0,5,1,5,2,3,5,4 },
|
||||
{ 5,0,1,5,2,3,5,4 },
|
||||
{ 0,1,2,6,3,4,6,5 },
|
||||
{ 4,4,4,0,1,2,4,3 },
|
||||
{ 0,5,5,1,2,3,5,4 },
|
||||
{ 5,0,5,1,2,3,5,4 },
|
||||
{ 0,1,6,2,3,4,6,5 },
|
||||
{ 5,5,0,1,2,3,5,4 },
|
||||
{ 0,6,1,2,3,4,6,5 },
|
||||
{ 6,0,1,2,3,4,6,5 },
|
||||
{ 0,1,2,3,4,5,7,6 },
|
||||
{ 2,2,2,2,2,2,0,1 },
|
||||
{ 0,3,3,3,3,3,1,2 },
|
||||
{ 3,0,3,3,3,3,1,2 },
|
||||
{ 0,1,4,4,4,4,2,3 },
|
||||
{ 3,3,0,3,3,3,1,2 },
|
||||
{ 0,4,1,4,4,4,2,3 },
|
||||
{ 4,0,1,4,4,4,2,3 },
|
||||
{ 0,1,2,5,5,5,3,4 },
|
||||
{ 3,3,3,0,3,3,1,2 },
|
||||
{ 0,4,4,1,4,4,2,3 },
|
||||
{ 4,0,4,1,4,4,2,3 },
|
||||
{ 0,1,5,2,5,5,3,4 },
|
||||
{ 4,4,0,1,4,4,2,3 },
|
||||
{ 0,5,1,2,5,5,3,4 },
|
||||
{ 5,0,1,2,5,5,3,4 },
|
||||
{ 0,1,2,3,6,6,4,5 },
|
||||
{ 3,3,3,3,0,3,1,2 },
|
||||
{ 0,4,4,4,1,4,2,3 },
|
||||
{ 4,0,4,4,1,4,2,3 },
|
||||
{ 0,1,5,5,2,5,3,4 },
|
||||
{ 4,4,0,4,1,4,2,3 },
|
||||
{ 0,5,1,5,2,5,3,4 },
|
||||
{ 5,0,1,5,2,5,3,4 },
|
||||
{ 0,1,2,6,3,6,4,5 },
|
||||
{ 4,4,4,0,1,4,2,3 },
|
||||
{ 0,5,5,1,2,5,3,4 },
|
||||
{ 5,0,5,1,2,5,3,4 },
|
||||
{ 0,1,6,2,3,6,4,5 },
|
||||
{ 5,5,0,1,2,5,3,4 },
|
||||
{ 0,6,1,2,3,6,4,5 },
|
||||
{ 6,0,1,2,3,6,4,5 },
|
||||
{ 0,1,2,3,4,7,5,6 },
|
||||
{ 3,3,3,3,3,0,1,2 },
|
||||
{ 0,4,4,4,4,1,2,3 },
|
||||
{ 4,0,4,4,4,1,2,3 },
|
||||
{ 0,1,5,5,5,2,3,4 },
|
||||
{ 4,4,0,4,4,1,2,3 },
|
||||
{ 0,5,1,5,5,2,3,4 },
|
||||
{ 5,0,1,5,5,2,3,4 },
|
||||
{ 0,1,2,6,6,3,4,5 },
|
||||
{ 4,4,4,0,4,1,2,3 },
|
||||
{ 0,5,5,1,5,2,3,4 },
|
||||
{ 5,0,5,1,5,2,3,4 },
|
||||
{ 0,1,6,2,6,3,4,5 },
|
||||
{ 5,5,0,1,5,2,3,4 },
|
||||
{ 0,6,1,2,6,3,4,5 },
|
||||
{ 6,0,1,2,6,3,4,5 },
|
||||
{ 0,1,2,3,7,4,5,6 },
|
||||
{ 4,4,4,4,0,1,2,3 },
|
||||
{ 0,5,5,5,1,2,3,4 },
|
||||
{ 5,0,5,5,1,2,3,4 },
|
||||
{ 0,1,6,6,2,3,4,5 },
|
||||
{ 5,5,0,5,1,2,3,4 },
|
||||
{ 0,6,1,6,2,3,4,5 },
|
||||
{ 6,0,1,6,2,3,4,5 },
|
||||
{ 0,1,2,7,3,4,5,6 },
|
||||
{ 5,5,5,0,1,2,3,4 },
|
||||
{ 0,6,6,1,2,3,4,5 },
|
||||
{ 6,0,6,1,2,3,4,5 },
|
||||
{ 0,1,7,2,3,4,5,6 },
|
||||
{ 6,6,0,1,2,3,4,5 },
|
||||
{ 0,7,1,2,3,4,5,6 },
|
||||
{ 7,0,1,2,3,4,5,6 },
|
||||
{ 0,1,2,3,4,5,6,7, }
|
||||
};
|
||||
#elif defined(__SSSE3__)
|
||||
#include <tmmintrin.h>
|
||||
static ALIGNED(char, shuffles[16][16], 16) = {
|
||||
#define _ 0x80
|
||||
{ _,_,_,_, _,_,_,_, _,_, _, _, _, _, _,_ },
|
||||
{ 0,1,2,3, _,_,_,_, _,_, _, _, _, _, _,_ },
|
||||
{ _,_,_,_, 0,1,2,3, _,_, _, _, _, _, _,_ },
|
||||
{ 0,1,2,3, 4,5,6,7, _,_, _, _, _, _, _,_ },
|
||||
{ _,_,_,_, _,_,_,_, 0,1, 2, 3, _, _, _,_ },
|
||||
{ 0,1,2,3, _,_,_,_, 4,5, 6, 7, _, _, _,_ },
|
||||
{ _,_,_,_, 0,1,2,3, 4,5, 6, 7, _, _, _,_ },
|
||||
{ 0,1,2,3, 4,5,6,7, 8,9,10,11, _, _, _,_ },
|
||||
{ _,_,_,_, _,_,_,_, _,_,_,_, 0, 1, 2, 3 },
|
||||
{ 0,1,2,3, _,_,_,_, _,_,_, _, 4, 5, 6, 7 },
|
||||
{ _,_,_,_, 0,1,2,3, _,_,_, _, 4, 5, 6, 7 },
|
||||
{ 0,1,2,3, 4,5,6,7, _,_, _, _, 8, 9,10,11 },
|
||||
{ _,_,_,_, _,_,_,_, 0,1, 2, 3, 4, 5, 6, 7 },
|
||||
{ 0,1,2,3, _,_,_,_, 4,5, 6, 7, 8, 9,10,11 },
|
||||
{ _,_,_,_, 0,1,2,3, 4,5, 6, 7, 8, 9,10,11 },
|
||||
{ 0,1,2,3, 4,5,6,7, 8,9,10,11, 12,13,14,15 },
|
||||
#undef _
|
||||
};
|
||||
#endif
|
||||
//---------------------- 32 bits ---------------------
|
||||
#define USIZE 32
|
||||
|
||||
//----- scalar -----------
|
||||
#define P4DD p4dd
|
||||
#define P4DDEC p4ddec
|
||||
#define BITUNPACK bitunpack
|
||||
|
||||
#define P4DDECX
|
||||
//delta 1
|
||||
#define P4DDD p4dd1d
|
||||
#define P4DDECD p4dd1dec
|
||||
#define BITUNPACKD bitd1unpack
|
||||
#define BITUNDD bitund1
|
||||
#include __FILE__
|
||||
|
||||
#undef P4DD
|
||||
#undef P4DDEC
|
||||
|
||||
#undef P4DDECX
|
||||
|
||||
#undef P4DDD
|
||||
#undef P4DDECD
|
||||
#undef BITUNPACKD
|
||||
#undef BITUNDD
|
||||
//delta 0
|
||||
#define P4DDD p4ddd
|
||||
#define P4DDECD p4dddec
|
||||
#define BITUNPACKD bitdunpack
|
||||
#define BITUNDD bitund
|
||||
#include __FILE__
|
||||
|
||||
#undef P4DDD
|
||||
#undef P4DDECD
|
||||
#undef BITUNPACKD
|
||||
#undef BITUNDD
|
||||
#undef BITUNPACK
|
||||
|
||||
//------ SIMD -------------
|
||||
#define P4SIMD
|
||||
#define P4DD p4ddv
|
||||
#define P4DDEC p4ddecv
|
||||
#define BITUNPACK bitunpackv
|
||||
|
||||
//#define P4DDECX
|
||||
|
||||
#define P4DDD p4dd1dv
|
||||
#define P4DDECD p4dd1decv
|
||||
#define BITUNPACKD bitd1unpackv
|
||||
#define BITUNPACKD_ _bitd1unpackv
|
||||
#define BITUNDD bitund1
|
||||
#include __FILE__
|
||||
|
||||
#undef P4DDEC
|
||||
//#undef BITUNPACK
|
||||
//#undef P4DDECX
|
||||
|
||||
#undef P4DDD
|
||||
#undef P4DDECD
|
||||
#undef BITUNPACKD
|
||||
#undef BITUNPACKD_
|
||||
#undef BITUNDD
|
||||
|
||||
#define P4DDD p4dddv
|
||||
#define P4DDECD p4dddecv
|
||||
#define BITUNPACKD bitdunpackv
|
||||
#define BITUNPACKD_ _bitdunpackv
|
||||
#define BITUNDD bitund
|
||||
#include __FILE__
|
||||
|
||||
#undef P4DDD
|
||||
#undef P4DDECD
|
||||
#undef BITUNPACKD
|
||||
#undef BITUNDD
|
||||
|
||||
#undef P4DD
|
||||
#undef BITUNPACK
|
||||
#undef USIZE
|
||||
#undef P4SIMD
|
||||
//----------------------------------
|
||||
#define USIZE 64
|
||||
#define P4DD p4dd
|
||||
#define P4DDEC p4ddec
|
||||
#define BITUNPACK bitunpack
|
||||
#include __FILE__
|
||||
#undef P4DD
|
||||
#undef P4DDEC
|
||||
#undef BITUNPACK
|
||||
|
||||
#else
|
||||
#define uint_t TEMPLATE3(uint, USIZE, _t)
|
||||
|
||||
#ifdef P4DDEC
|
||||
unsigned char *_bitunpackv32( const unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned b, unsigned *__restrict pex, unsigned char *bb);
|
||||
|
||||
unsigned char *TEMPLATE2(P4DD, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out, unsigned b, unsigned bx) {
|
||||
uint_t ex[0x100+8];
|
||||
|
||||
if(!(b & 1)) return TEMPLATE2(BITUNPACK, USIZE)(in, n, out, b>>1);
|
||||
b >>= 1;
|
||||
#if defined(P4SIMD) && defined(__SSE3__) && USIZE == 32 && P4DN == 2
|
||||
unsigned char *pb = in;
|
||||
in = TEMPLATE2(bitunpack, USIZE)(in+16, popcnt64(*(unsigned long long *)in) + popcnt64(*(unsigned long long *)(in+8)), ex, bx);
|
||||
return _bitunpackv32(in, n, out, b, ex, pb);
|
||||
#else
|
||||
#if P4DN == 2
|
||||
unsigned long long bb[P4DN]; unsigned num=0,i;
|
||||
bb[0] = *(unsigned long long *)in; in += 8;
|
||||
bb[1] = *(unsigned long long *)in; in += 8;
|
||||
in = TEMPLATE2(bitunpack, USIZE)(in, popcnt64(bb[0]) + popcnt64(bb[1]), ex, bx);
|
||||
#else
|
||||
unsigned long long bb[P4DN]; unsigned num=0;
|
||||
for(i = 0; i < P4DN; i++) { bb[i] = *(unsigned long long *)in; in += 8; num += popcnt64(bb[i]); }
|
||||
in = TEMPLATE2(bitunpack, USIZE)(in, num, ex, bx);
|
||||
#endif
|
||||
|
||||
in = TEMPLATE2(BITUNPACK, USIZE)(in, n, out, b);
|
||||
#ifdef __AVX2__ON
|
||||
uint_t *op,*pex = ex;
|
||||
#if P4DN == 2
|
||||
for(op = out; b0; b0 >>= 8,op += 8) { unsigned m = (unsigned char)b0, mc = popcnt32(m), s = pex[mc]; pex[mc]=0;
|
||||
_mm256_storeu_si256((__m256i *)op, _mm256_add_epi32(_mm256_loadu_si256((const __m256i*)op), _mm256_permutevar8x32_epi32(_mm256_slli_epi32(_mm256_load_si256((const __m256i*)pex), b), _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i *)shuffles[m])) )) ); pex += mc; *pex=s;
|
||||
}
|
||||
for(op = out+64; b1; b1 >>= 8,op += 8) { unsigned m = (unsigned char)b1, mc=popcnt32(m), s = pex[mc]; pex[mc]=0;
|
||||
_mm256_storeu_si256((__m256i *)op, _mm256_add_epi32(_mm256_loadu_si256((const __m256i*)op), _mm256_permutevar8x32_epi32(_mm256_slli_epi32(_mm256_load_si256((const __m256i*)pex), b), _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i *)shuffles[m])) )) ); pex += mc; *pex=s;
|
||||
}
|
||||
#else
|
||||
for(i = 0; i < P4DN; i++) {
|
||||
for(op = out; bb[i]; bb[i] >>= 8,op += 8) { unsigned m = (unsigned char)bb[i], mc=popcnt32(m), s = pex[mc]; pex[mc]=0;
|
||||
_mm256_storeu_si256((__m256i *)op, _mm256_add_epi32(_mm256_loadu_si256((const __m256i*)op), _mm256_permutevar8x32_epi32(_mm256_slli_epi32(_mm256_load_si256((const __m256i*)pex), b), _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i *)shuffles[m])) )) ); pex += mc; *pex=s;
|
||||
} out += 64;
|
||||
}
|
||||
#endif
|
||||
#elif defined(__SSSE3__) && USIZE == 32
|
||||
uint_t *op,*pex = ex;
|
||||
#if P4DN == 2
|
||||
for(op = out; bb[0]; bb[0] >>= 4,op+=4) { const unsigned m = bb[0]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
}
|
||||
for(op=out+64; bb[1]; bb[1] >>= 4,op+=4) { const unsigned m = bb[1]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
}
|
||||
#else
|
||||
for(i = 0; i < P4DN; i++) { // Loop unrolling
|
||||
for(op = out; bb[i]; bb[i] >>= 4,op+=4) { const unsigned m = bb[i]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
} out+=64;
|
||||
}
|
||||
#endif
|
||||
#else
|
||||
unsigned k = 0;
|
||||
#if P4DN == 2
|
||||
while(bb[0]) { unsigned x = ctz64(bb[0]); out[x] += ex[k++]<<b; bb[0] ^= (1ull<<x); } out += 64;
|
||||
while(bb[1]) { unsigned x = ctz64(bb[1]); out[x] += ex[k++]<<b; bb[1] ^= (1ull<<x); }
|
||||
#else
|
||||
for(i = 0; i < P4DN; i++) {
|
||||
while(bb[i]) { unsigned x = ctz64(b0); out[x] += ex[k++]<<b; bb[i] ^= (1ull<<x); } out += 64;
|
||||
}
|
||||
#endif
|
||||
#endif
|
||||
return in;
|
||||
#endif
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(P4DDEC, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out) {
|
||||
unsigned b = *in++,bx;
|
||||
if(b&1) bx = *in++;
|
||||
return TEMPLATE2(P4DD, USIZE)(in, n, out, b, bx);
|
||||
}
|
||||
#endif
|
||||
|
||||
#ifdef P4DDECD
|
||||
unsigned char *_bitd1unpackv32( const unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned *__restrict pex, unsigned char *bb);
|
||||
unsigned char *_bitdunpackv32( const unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned *__restrict pex, unsigned char *bb);
|
||||
|
||||
unsigned char *TEMPLATE2(P4DDD, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out, unsigned start, unsigned b, unsigned bx) {
|
||||
uint_t ex[0x100+8], *_op = out;
|
||||
if(!(b & 1)) return TEMPLATE2(BITUNPACKD, USIZE)(in, n, out, start, b>>1);
|
||||
b >>= 1;
|
||||
|
||||
#if defined(P4SIMD) && defined(__SSE3__) && USIZE == 32 && P4DN == 2
|
||||
unsigned char *pb = in;
|
||||
in = TEMPLATE2(bitunpack, USIZE)(in+16, popcnt64(*(unsigned long long *)in) + popcnt64(*(unsigned long long *)(in+8)), ex, bx);
|
||||
return TEMPLATE2(BITUNPACKD_, USIZE)(in, n, out, start, b, ex, pb);
|
||||
#else
|
||||
unsigned long long bb[P4DN]; unsigned num=0,i;
|
||||
for(i = 0; i < P4DN; i++) { bb[i] = *(unsigned long long *)in; in += 8; num += popcnt64(bb[i]); }
|
||||
in = TEMPLATE2(bitunpack, USIZE)(in, num, ex, bx);
|
||||
|
||||
in = TEMPLATE2(BITUNPACK, USIZE)(in, n, out, b);
|
||||
#ifdef __AVX2__ON
|
||||
uint_t *op,*pex = ex;
|
||||
for(i = 0; i < P4DN; i++) {
|
||||
for(op = out; bb[i]; bb[i] >>= 8,op += 8) { unsigned m = (unsigned char)bb[i], mc=popcnt32(m), s = pex[mc]; pex[mc]=0;
|
||||
_mm256_storeu_si256((__m256i *)op, _mm256_add_epi32(_mm256_loadu_si256((const __m256i*)op), _mm256_permutevar8x32_epi32(_mm256_slli_epi32(_mm256_load_si256((const __m256i*)pex), b), _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i *)shuffles[m])) )) ); pex += mc; *pex=s;
|
||||
} out += 64;
|
||||
}
|
||||
#elif defined(__SSSE3__)
|
||||
uint_t *op,*pex = ex;
|
||||
#if P4DN == 2
|
||||
for(op = out; bb[0]; bb[0] >>= 4,op+=4) { const unsigned m = bb[0]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
}
|
||||
for(op=out+64; bb[1]; bb[1] >>= 4,op+=4) { const unsigned m = bb[1]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
}
|
||||
#else
|
||||
for(i = 0; i < P4DN; i++) { // Loop unrolling
|
||||
for(op = out; bb[i]; bb[i] >>= 4,op+=4) { const unsigned m = bb[i]&0xf;
|
||||
_mm_storeu_si128((__m128i *)op, _mm_add_epi32(_mm_loadu_si128((__m128i*)op), _mm_shuffle_epi8(_mm_slli_epi32(_mm_loadu_si128((__m128i*)pex), b), _mm_load_si128((__m128i*)shuffles[m]) ) )); pex += popcnt32(m);
|
||||
} out+=64;
|
||||
}
|
||||
#endif
|
||||
#else
|
||||
unsigned k = 0;
|
||||
for(i = 0; i < P4DN; i++) {
|
||||
while(bb[i]) { unsigned x = ctz64(bb[i]); out[x] += ex[k++]<<b; bb[i] ^= (1ull<<x); } out += 64;
|
||||
}
|
||||
#endif
|
||||
TEMPLATE2(BITUNDD, USIZE)(_op, n, start);
|
||||
return in;
|
||||
#endif
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(P4DDECD, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out, unsigned start) {
|
||||
unsigned b = *in++,bx;
|
||||
if(b&1) bx = *in++;
|
||||
return TEMPLATE2(P4DDD, USIZE)(in, n, out, start, b, bx);
|
||||
}
|
||||
#endif
|
||||
|
||||
#ifdef P4DDECX
|
||||
unsigned char *TEMPLATE2(p4ddecx, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out) {
|
||||
unsigned b,i;
|
||||
struct p4d p4d;
|
||||
p4dini(&p4d, &in, n, &b);
|
||||
|
||||
if(unlikely(p4d.i & 1)) {
|
||||
for(i = 0; i < n; i++)
|
||||
out[i] = TEMPLATE2(vp4dget, USIZE)(&p4d, in, b, i);
|
||||
} else for(i = 0; i < n; i++) out[i] = TEMPLATE2(_bitgetx, USIZE)(in, b, i*b);
|
||||
return in + PAD8(n*b);
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(p4dfdecx, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out, unsigned start) {
|
||||
unsigned b,i;
|
||||
struct p4d p4d;
|
||||
p4dini(&p4d, &in, n, &b);
|
||||
|
||||
if(unlikely(p4d.i & 1)) {
|
||||
for(i = 0; i < n; i++)
|
||||
out[i] = TEMPLATE2(vp4dget, USIZE)(&p4d, in, b, i)+start+i+1;
|
||||
} else for(i = 0; i < n; i++) out[i] = TEMPLATE2(_bitgetx, USIZE)(in, b, i*b)+start+i+1;
|
||||
return in + PAD8(n*b);
|
||||
}
|
||||
|
||||
unsigned char *TEMPLATE2(p4df0decx, USIZE)(unsigned char *__restrict in, unsigned n, uint_t *__restrict out, unsigned start) {
|
||||
unsigned b,i;
|
||||
struct p4d p4d;
|
||||
p4dini(&p4d, &in, n, &b);
|
||||
|
||||
if(unlikely(p4d.i & 1)) {
|
||||
for(i = 0; i < n; i++)
|
||||
out[i] = TEMPLATE2(vp4dget, USIZE)(&p4d, in, b, i)+start;// return p4d.ex + PAD8((p4d.cum[P4DN-1] + popcnt64(p4d.xmap[P4DN-1]))*p4d.bx);
|
||||
} else for(i = 0; i < n; i++) out[i] = TEMPLATE2(_bitgetx, USIZE)(in, b, i*b)+start;
|
||||
return in + PAD8(n*b);
|
||||
}
|
||||
#endif
|
||||
|
||||
#endif
|
||||
|
||||
115
vp4dd.h
115
vp4dd.h
@ -1,115 +0,0 @@
|
||||
/**
|
||||
Copyright (C) powturbo 2013-2015
|
||||
GPL v2 License
|
||||
|
||||
This program is free software; you can redistribute it and/or modify
|
||||
it under the terms of the GNU General Public License as published by
|
||||
the Free Software Foundation; either version 2 of the License, or
|
||||
(at your option) any later version.
|
||||
|
||||
This program is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
GNU General Public License for more details.
|
||||
|
||||
You should have received a copy of the GNU General Public License along
|
||||
with this program; if not, write to the Free Software Foundation, Inc.,
|
||||
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
||||
|
||||
- homepage : https://sites.google.com/site/powturbo/
|
||||
- github : https://github.com/powturbo
|
||||
- twitter : https://twitter.com/powturbo
|
||||
- email : powturbo [_AT_] gmail [_DOT_] com
|
||||
**/
|
||||
// vp4dd.h - "Integer Compression" Turbo PforDelta
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
#include <stdint.h>
|
||||
#define P4DSIZE 128 //64 //
|
||||
#define P4DN (P4DSIZE/64)
|
||||
|
||||
//---------------- Bulk decompress of TurboPFor compressed integer array -------------------------------------------------------
|
||||
// decompress a previously (with p4denc32) 32 bits packed array. Return value = end of packed buffer in
|
||||
unsigned char *p4ddec32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out);
|
||||
unsigned char *p4ddec64( unsigned char *__restrict in, unsigned n, uint64_t *__restrict out);
|
||||
|
||||
unsigned char *p4dd32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned b, unsigned bx);
|
||||
unsigned char *p4dd64( unsigned char *__restrict in, unsigned n, uint64_t *__restrict out, unsigned b, unsigned bx);
|
||||
|
||||
unsigned char *p4ddecv32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out); // SIMD
|
||||
|
||||
//-- delta min = 0
|
||||
unsigned char *p4dddec32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
unsigned char *p4dddecv32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
//-- delta min = 1
|
||||
unsigned char *p4dd1dec32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
unsigned char *p4dd1decv32(unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
|
||||
// same as abose, b and bx not stored within the compressed stream header (see idxcr.c/idxqry.c for an example)
|
||||
unsigned char *p4ddv32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned b, unsigned bx);
|
||||
|
||||
unsigned char *p4ddd32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned bx);
|
||||
unsigned char *p4dddv32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned bx);
|
||||
|
||||
unsigned char *p4dd1d32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned bx);
|
||||
unsigned char *p4dd1dv32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start, unsigned b, unsigned bx);
|
||||
|
||||
//---------------- Direct Access functions to compressed TurboPFor array -------------------------------------------------------
|
||||
#define P4D_PAD8(__x) ( (((__x)+8-1)/8) )
|
||||
#define P4D_B(__x) (((__x) >> 1) & 0x3f)
|
||||
#define P4D_XB(__x) (((__x) & 1)?((__x) >> 8):0)
|
||||
#define P4D_ININC(__in, __x) __in += 1+(__x & 1)
|
||||
|
||||
static inline unsigned vp4dbits(unsigned char *__restrict in, int *bx) { unsigned i = *(unsigned short *)in; *bx = P4D_XB(i); return P4D_B(i); }
|
||||
|
||||
struct p4d {
|
||||
unsigned long long *xmap;
|
||||
unsigned char *ex;
|
||||
unsigned i,bx,cum[P4DN+1];
|
||||
int oval,idx;
|
||||
};
|
||||
|
||||
// prepare direct access usage
|
||||
static inline void p4dini(struct p4d *p4d, unsigned char *__restrict *pin, unsigned n, unsigned *b) { unsigned char *in = *pin;
|
||||
static unsigned long long xmap[P4DN+1] = { 0 };
|
||||
|
||||
unsigned i = *(unsigned short *)in;
|
||||
p4d->i = i;
|
||||
*b = P4D_B(i);
|
||||
p4d->bx = P4D_XB(i);
|
||||
P4D_ININC(in, i);
|
||||
|
||||
p4d->xmap = (i&1)?(unsigned long long *)in:xmap;
|
||||
p4d->ex = in + ((i&1)?8*P4DN:0);
|
||||
for(p4d->cum[0] = 0, i=1; i < P4DN; i++) p4d->cum[i] = p4d->cum[i-1] + popcnt64(p4d->xmap[i-1]);
|
||||
*pin = p4d->ex + P4D_PAD8((p4d->cum[P4DN-1] + popcnt64(p4d->xmap[P4DN-1]))*p4d->bx);
|
||||
p4d->oval = p4d->idx = -1;
|
||||
}
|
||||
|
||||
// Get a single value with index "idx" from a p4denc32 packed array
|
||||
static ALWAYS_INLINE unsigned vp4dget32(struct p4d *p4d, unsigned char *__restrict in, unsigned b, unsigned idx) { unsigned bi, cl, u = _bitgetx32(in, b, idx*b);
|
||||
if(unlikely(p4d->xmap[bi = idx>>6] & (1ull<<(cl = (idx & 0x3f))))) u |= _bitgetx32(p4d->ex, p4d->bx, (p4d->cum[bi] + popcnt64(p4d->xmap[bi] & ~((~0ull)<<cl)))*p4d->bx ) << b;
|
||||
return u;
|
||||
}
|
||||
|
||||
// like vp4dget32 but for 16 bits packed array (with p4denc16)
|
||||
static ALWAYS_INLINE unsigned vp4dget16(struct p4d *p4d, unsigned char *__restrict in, unsigned b, unsigned idx) { unsigned bi, cl, u = _bitgetx16(in, b, idx*b);
|
||||
if(unlikely(p4d->xmap[bi = idx>>6] & (1ull<<(cl = (idx & 0x3f))))) u |= _bitgetx32(p4d->ex, p4d->bx, (p4d->cum[bi] + popcnt64(p4d->xmap[bi] & ~((~0ull)<<cl)))*p4d->bx ) << b;
|
||||
return u;
|
||||
}
|
||||
|
||||
// Get the next single value greater of equal to val
|
||||
static ALWAYS_INLINE int vp4dgeq(struct p4d *p4d, unsigned char *__restrict in, unsigned b, int val) { do p4d->oval += vp4dget32(p4d, in, b, ++p4d->idx)+1; while(p4d->oval < val); return p4d->oval; }
|
||||
|
||||
/* like p4ddec32 but using direct access. This is only a demo showing direct access usage. Use p4ddec32 for instead for decompressing entire blocks */
|
||||
unsigned char *p4ddecx32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out);
|
||||
unsigned char *p4dfdecx32( unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
unsigned char *p4df0decx32(unsigned char *__restrict in, unsigned n, unsigned *__restrict out, unsigned start);
|
||||
|
||||
#define P4DSIZE 128 //64 //
|
||||
#define P4DN (P4DSIZE/64)
|
||||
#ifdef __cplusplus
|
||||
}
|
||||
#endif
|
||||
Reference in New Issue
Block a user