/** Copyright (c) 2016-2019, Powturbo All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - homepage : https://sites.google.com/site/powturbo/ - github : https://github.com/powturbo - twitter : https://twitter.com/powturbo - email : powturbo [_AT_] gmail [_DOT_] com **/ // TubeBase64: ssse3 + arm neon functions (see also turbob64avx2) #include #if defined(__AVX__) #include #define FUNPREF tb64avx #elif defined(__SSE4_1__) #include #define FUNPREF tb64sse #elif defined(__SSSE3__) #ifdef __powerpc64__ #define __SSE__ 1 #define __SSE2__ 1 #define __SSE3__ 1 #define NO_WARN_X86_INTRINSICS 1 #endif #define FUNPREF tb64sse #include #elif defined(__SSE2__) #include #elif defined(__ARM_NEON) #include #endif #define UA_MEMCPY #include "conf.h" #include "turbob64.h" #include "turbob64_.h" #ifdef __ARM_NEON //----------------------------------- arm neon -------------------------------- #define _ 0xff // invald entry static const unsigned char lut[] = { _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _, _,62, _, _, _,63, 52,53,54,55,56,57,58,59,60,61, _, _, _, _, _, _, _, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14, 15,16,17,18,19,20,21,22,23,24,25, _, _, _, _, _, _,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40, 41,42,43,44,45,46,47,48,49,50,51, _, _, _, _, _, }; #undef _ #ifndef vld1q_u8_x4 static inline uint8x16x4_t vld1q_u8_x4(const uint8_t *lut) { uint8x16x4_t v; v.val[0] = vld1q_u8(lut); v.val[1] = vld1q_u8(lut+16); v.val[2] = vld1q_u8(lut+32); v.val[3] = vld1q_u8(lut+48); return v; } #endif #define B64D(iv, ov) {\ iv.val[0] = vqtbx4q_u8(vqtbl4q_u8(vlut1, veorq_u8(iv.val[0], cv40)), vlut0, iv.val[0]);\ iv.val[1] = vqtbx4q_u8(vqtbl4q_u8(vlut1, veorq_u8(iv.val[1], cv40)), vlut0, iv.val[1]);\ iv.val[2] = vqtbx4q_u8(vqtbl4q_u8(vlut1, veorq_u8(iv.val[2], cv40)), vlut0, iv.val[2]);\ iv.val[3] = vqtbx4q_u8(vqtbl4q_u8(vlut1, veorq_u8(iv.val[3], cv40)), vlut0, iv.val[3]);\ \ ov.val[0] = vorrq_u8(vshlq_n_u8(iv.val[0], 2), vshrq_n_u8(iv.val[1], 4));\ ov.val[1] = vorrq_u8(vshlq_n_u8(iv.val[1], 4), vshrq_n_u8(iv.val[2], 2));\ ov.val[2] = vorrq_u8(vshlq_n_u8(iv.val[2], 6), iv.val[3] );\ } #define _MM_B64CHK(iv, xv) xv = vorrq_u8(xv, vorrq_u8(vorrq_u8(iv.val[0], iv.val[1]), vorrq_u8(iv.val[2], iv.val[3]))) size_t tb64ssedec(const unsigned char *in, size_t inlen, unsigned char *out) { const unsigned char *ip; unsigned char *op; const uint8x16x4_t vlut0 = vld1q_u8_x4( lut), vlut1 = vld1q_u8_x4(&lut[64]); const uint8x16_t cv40 = vdupq_n_u8(0x40); uint8x16_t xv = vdupq_n_u8(0); #define ND 256 for(ip = in, op = out; ip != in+(inlen&~(ND-1)); ip += ND, op += (ND/4)*3) { PREFETCH(ip,256,0); uint8x16x4_t iv0 = vld4q_u8(ip), iv1 = vld4q_u8(ip+64); uint8x16x3_t ov0,ov1; B64D(iv0, ov0); #if ND > 128 CHECK1(_MM_B64CHK(iv0,xv)); #else CHECK0(_MM_B64CHK(iv0,xv)); #endif B64D(iv1, ov1); CHECK1(_MM_B64CHK(iv1,xv)); #if ND > 128 iv0 = vld4q_u8(ip+128); iv1 = vld4q_u8(ip+192); #endif vst3q_u8(op, ov0); vst3q_u8(op+48, ov1); #if ND > 128 B64D(iv0,ov0); CHECK1(_MM_B64CHK(iv0,xv)); B64D(iv1,ov1); vst3q_u8(op+ 96, ov0); vst3q_u8(op+144, ov1); CHECK0(_MM_B64CHK(iv1,xv)); #endif } for( ; ip != in+(inlen&~(64-1)); ip += 64, op += (64/4)*3) { uint8x16x4_t iv = vld4q_u8(ip); uint8x16x3_t ov; B64D(iv,ov); vst3q_u8(op, ov); CHECK0(xv = vorrq_u8(xv, vorrq_u8(vorrq_u8(iv.val[0], iv.val[1]), vorrq_u8(iv.val[2], iv.val[3])))); } size_t rc; if(!(rc=tb64xdec(ip, inlen&(64-1), op)) || vaddvq_u8(vshrq_n_u8(xv,7))) return 0; //decode all return (op-out)+rc; } //-------------------------------------------------------------------------------------------------- #define B64E(iv, ov) {\ ov.val[0] = vshrq_n_u8(iv.val[0], 2);\ ov.val[1] = vandq_u8(vorrq_u8(vshlq_n_u8(iv.val[0], 4), vshrq_n_u8(iv.val[1], 4)), cv3f);\ ov.val[2] = vandq_u8(vorrq_u8(vshlq_n_u8(iv.val[1], 2), vshrq_n_u8(iv.val[2], 6)), cv3f);\ ov.val[3] = vandq_u8( iv.val[2], cv3f);\ \ ov.val[0] = vqtbl4q_u8(vlut, ov.val[0]);\ ov.val[1] = vqtbl4q_u8(vlut, ov.val[1]);\ ov.val[2] = vqtbl4q_u8(vlut, ov.val[2]);\ ov.val[3] = vqtbl4q_u8(vlut, ov.val[3]);\ } size_t tb64sseenc(const unsigned char* in, size_t inlen, unsigned char *out) { static unsigned char lut[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; const unsigned char *ip; unsigned char *op; const size_t outlen = TB64ENCLEN(inlen); const uint8x16x4_t vlut = vld1q_u8_x4(lut); const uint8x16_t cv3f = vdupq_n_u8(0x3f); #define NE 128 // 256// for(ip = in, op = out; op != out+(outlen&~(NE-1)); op += NE, ip += (NE/4)*3) { uint8x16x3_t iv0 = vld3q_u8(ip), iv1 = vld3q_u8(ip+48); uint8x16x4_t ov0,ov1; B64E(iv0, ov0); B64E(iv1, ov1); #if NE > 128 iv0 = vld3q_u8(ip+ 96); iv1 = vld3q_u8(ip+144); #endif vst4q_u8(op, ov0); vst4q_u8(op+64, ov1); //PREFETCH(ip,256,0); #if NE > 128 B64E(iv0, ov0); B64E(iv1, ov1); vst4q_u8(op+128, ov0); vst4q_u8(op+192, ov1); #endif } for( ; op != out+(outlen&~(64-1)); op += 64, ip += (64/4)*3) { const uint8x16x3_t iv = vld3q_u8(ip); uint8x16x4_t ov; B64E(iv, ov); vst4q_u8(op,ov); } EXTAIL(); return outlen; } #elif defined(__SSSE3__) //----------------- SSSE3 / SSE4.1 / AVX (derived from the AVX2 functions ) ----------------------------------------------------------------- #define OVD 4 size_t TEMPLATE2(FUNPREF, dec)(const unsigned char *in, size_t inlen, unsigned char *out) { if(inlen >= 16+OVD) { const unsigned char *ip; unsigned char *op; #ifdef __AVX__ #define ND 64 #else #define ND 32 #endif __m128i vx = _mm_setzero_si128(); const __m128i delta_asso = _mm_setr_epi8(0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0f, 0x00, 0x0f); const __m128i delta_values = _mm_setr_epi8(0x00, 0x00, 0x00, 0x13, 0x04, 0xbf, 0xbf, 0xb9, 0xb9, 0x00, 0x10, 0xc3, 0xbf, 0xbf, 0xb9, 0xb9); #ifndef NB64CHECK const __m128i check_asso = _mm_setr_epi8(0x0d, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x03, 0x07, 0x0b, 0x0b, 0x0b, 0x0f); const __m128i check_values = _mm_setr_epi8(0x80, 0x80, 0x80, 0x80, 0xcf, 0xbf, 0xd5, 0xa6, 0xb5, 0x86, 0xd1, 0x80, 0xb1, 0x80, 0x91, 0x80); #endif const __m128i cpv = _mm_set_epi8( -1, -1, -1, -1, 12, 13, 14, 8, 9, 10, 4, 5, 6, 0, 1, 2); for(ip = in, op = out; ip < (in+inlen)-(ND+OVD); ip += ND, op += (ND/4)*3) { __m128i iv0 = _mm_loadu_si128((__m128i *) ip), iv1 = _mm_loadu_si128((__m128i *)(ip+16)); __m128i ov0,shifted0; MM_MAP8TO6(iv0, shifted0,delta_asso, delta_values, ov0); MM_PACK8TO6(ov0, cpv); __m128i ov1,shifted1; MM_MAP8TO6(iv1, shifted1,delta_asso, delta_values, ov1); MM_PACK8TO6(ov1, cpv); _mm_storeu_si128((__m128i*) op, ov0); _mm_storeu_si128((__m128i*)(op+12), ov1); PREFETCH(ip,1024,0); #if ND > 32 __m128i iv2 = _mm_loadu_si128((__m128i *)(ip+32)), iv3 = _mm_loadu_si128((__m128i *)(ip+48)); #endif CHECK0(MM_B64CHK(iv0, shifted0, check_asso, check_values, vx)); CHECK1(MM_B64CHK(iv1, shifted1, check_asso, check_values, vx)); #if ND > 32 __m128i ov2,shifted2; MM_MAP8TO6(iv2, shifted2,delta_asso, delta_values, ov2); MM_PACK8TO6(ov2, cpv); __m128i ov3,shifted3; MM_MAP8TO6(iv3, shifted3,delta_asso, delta_values, ov3); MM_PACK8TO6(ov3, cpv); _mm_storeu_si128((__m128i*)(op+24), ov2); _mm_storeu_si128((__m128i*)(op+36), ov3); CHECK1(MM_B64CHK(iv2, shifted2, check_asso, check_values, vx)); CHECK1(MM_B64CHK(iv3, shifted3, check_asso, check_values, vx)); #endif } #ifdef __AVX__ for(; ip < (in+inlen)-(16+OVD); ip += 16, op += (16/4)*3) { #else if(ip < (in+inlen)-(16+OVD)) { #endif __m128i iv0 = _mm_loadu_si128((__m128i *) ip); __m128i ov0, shifted0; MM_MAP8TO6(iv0, shifted0,delta_asso, delta_values, ov0); MM_PACK8TO6(ov0, cpv); _mm_storeu_si128((__m128i*) op, ov0); CHECK1(MM_B64CHK(iv0, shifted0, check_asso, check_values, vx)); #ifndef __AVX__ ip += 16; op += (16/4)*3; #endif } size_t rc; if(!(rc = _tb64xdec(ip, inlen-(ip-in), op)) || _mm_movemask_epi8(vx)) return 0; return (op-out)+rc; } return _tb64xdec(in, inlen, out); } #define OVE 8 size_t TEMPLATE2(FUNPREF, enc)(const unsigned char* in, size_t inlen, unsigned char *out) { const unsigned char *ip = in; unsigned char *op = out; size_t outlen = TB64ENCLEN(inlen); #ifdef __AVX__ #define NE 64 #else #define NE 32 #endif if(outlen >= 16+OVE) { const __m128i shuf = _mm_set_epi8(10,11, 9, 10, 7, 8, 6, 7, 4, 5, 3, 4, 1, 2, 0, 1); for(; op <= (out+outlen)-(NE+OVE); op += NE, ip += (NE/4)*3) { PREFETCH(ip,1024,0); __m128i v0 = _mm_loadu_si128((__m128i*)ip), v1 = _mm_loadu_si128((__m128i*)(ip+12)); #if NE > 32 __m128i v2 = _mm_loadu_si128((__m128i*)(ip+24)), v3 = _mm_loadu_si128((__m128i*)(ip+36)); #endif v0 = _mm_shuffle_epi8(v0, shuf); v1 = _mm_shuffle_epi8(v1, shuf); v0 = mm_unpack6to8(v0); v1 = mm_unpack6to8(v1); v0 = mm_map6to8(v0); v1 = mm_map6to8(v1); _mm_storeu_si128((__m128i*) op, v0); _mm_storeu_si128((__m128i*)(op+16), v1); #if NE > 32 v2 = _mm_shuffle_epi8(v2, shuf); v3 = _mm_shuffle_epi8(v3, shuf); v2 = mm_unpack6to8(v2); v3 = mm_unpack6to8(v3); v2 = mm_map6to8(v2); v3 = mm_map6to8(v3); _mm_storeu_si128((__m128i*)(op+32), v2); _mm_storeu_si128((__m128i*)(op+48), v3); #endif } for(; op <= (out+outlen)-(16+OVE); op += 16, ip += (16/4)*3) { __m128i v0 = _mm_loadu_si128((__m128i*)ip); v0 = _mm_shuffle_epi8(v0, shuf); v0 = mm_unpack6to8(v0); v0 = mm_map6to8(v0); _mm_storeu_si128((__m128i*) op, v0); } } EXTAIL(); return outlen; } #endif //------------------------------------------------------------------------------------------------------------------- #if !defined(__AVX__) //include only 1 time size_t tb64memcpy(const unsigned char* in, size_t inlen, unsigned char *out) { memcpy(out, in, inlen); return inlen; } static unsigned _cpuisa; //--------------------- CPU detection ------------------------------------------- #if defined(__i386__) || defined(__x86_64__) #if _MSC_VER >=1300 #include #elif defined (__INTEL_COMPILER) #include #endif static inline void cpuid(int reg[4], int id) { #if defined (_MSC_VER) //|| defined (__INTEL_COMPILER) __cpuidex(reg, id, 0); #elif defined(__i386__) || defined(__x86_64__) __asm("cpuid" : "=a"(reg[0]),"=b"(reg[1]),"=c"(reg[2]),"=d"(reg[3]) : "a"(id),"c"(0) : ); #endif } static inline uint64_t xgetbv (int ctr) { #if(defined _MSC_VER && (_MSC_FULL_VER >= 160040219) || defined __INTEL_COMPILER) return _xgetbv(ctr); #elif defined(__i386__) || defined(__x86_64__) unsigned a, d; __asm("xgetbv" : "=a"(a),"=d"(d) : "c"(ctr) : ); return (uint64_t)d << 32 | a; #else unsigned a=0, d=0; return (uint64_t)d << 32 | a; #endif } #endif #define AVX512F 0x001 #define AVX512DQ 0x002 #define AVX512IFMA 0x004 #define AVX512PF 0x008 #define AVX512ER 0x010 #define AVX512CD 0x020 #define AVX512BW 0x040 #define AVX512VL 0x080 #define AVX512VNNI 0x100 #define AVX512VBMI 0x200 #define AVX512VBMI2 0x400 #define IS_SSE 0x10 #define IS_SSE2 0x20 #define IS_SSE3 0x30 #define IS_SSSE3 0x32 #define IS_POWER9 0x34 // powerpc #define IS_NEON 0x38 // arm neon #define IS_SSE41 0x40 #define IS_SSE41x 0x41 //+popcount #define IS_SSE42 0x42 #define IS_AVX 0x50 #define IS_AVX2 0x60 #define IS_AVX512 0x800 unsigned cpuisa(void) { int c[4] = {0}; if(_cpuisa) return _cpuisa; _cpuisa++; #if defined(__i386__) || defined(__x86_64__) cpuid(c, 0); if(c[0]) { cpuid(c, 1); //family = ((c >> 8) & 0xf) + ((c >> 20) & 0xff) //model = ((c >> 4) & 0xf) + ((c >> 12) & 0xf0) if( c[3] & (1 << 25)) { _cpuisa = IS_SSE; if( c[3] & (1 << 26)) { _cpuisa = IS_SSE2; if( c[2] & (1 << 0)) { _cpuisa = IS_SSE3; // _cpuisa = IS_SSE3SLOW; // Atom SSSE3 slow if( c[2] & (1 << 9)) { _cpuisa = IS_SSSE3; if( c[2] & (1 << 19)) { _cpuisa = IS_SSE41; if( c[2] & (1 << 23)) { _cpuisa = IS_SSE41x; // +popcount if( c[2] & (1 << 20)) { _cpuisa = IS_SSE42; // SSE4.2 if((c[2] & (1 << 28)) && (c[2] & (1 << 27)) && // OSXSAVE (c[2] & (1 << 26)) && // XSAVE (xgetbv(0) & 6)==6) { _cpuisa = IS_AVX; // AVX if(c[2]& (1 << 3)) _cpuisa |= 1; // +FMA3 if(c[2]& (1 << 16)) _cpuisa |= 2; // +FMA4 if(c[2]& (1 << 25)) _cpuisa |= 4; // +AES cpuid(c, 7); if(c[1] & (1 << 5)) { _cpuisa = IS_AVX2; if(c[1] & (1 << 16)) { cpuid(c, 0xd); if((c[0] & 0x60)==0x60) { _cpuisa = IS_AVX512; cpuid(c, 7); if(c[1] & (1<<16)) _cpuisa |= AVX512F; if(c[1] & (1<<17)) _cpuisa |= AVX512DQ; if(c[1] & (1<<21)) _cpuisa |= AVX512IFMA; if(c[1] & (1<<26)) _cpuisa |= AVX512PF; if(c[1] & (1<<27)) _cpuisa |= AVX512ER; if(c[1] & (1<<28)) _cpuisa |= AVX512CD; if(c[1] & (1<<30)) _cpuisa |= AVX512BW; if(c[1] & (1u<<31)) _cpuisa |= AVX512VL; if(c[2] & (1<< 1)) _cpuisa |= AVX512VBMI; if(c[2] & (1<<11)) _cpuisa |= AVX512VNNI; if(c[2] & (1<< 6)) _cpuisa |= AVX512VBMI2; }}} }}}}}}}}} #elif defined(__powerpc64__) _cpuisa = IS_POWER9; // power9 #elif defined(__ARM_NEON) _cpuisa = IS_NEON; // ARM_NEON #endif return _cpuisa; } unsigned cpuini(unsigned cpuisa) { if(cpuisa) _cpuisa = cpuisa; return _cpuisa; } char *cpustr(unsigned cpuisa) { if(!cpuisa) cpuisa = _cpuisa; #if defined(__i386__) || defined(__x86_64__) if(cpuisa >= IS_AVX512) { if(cpuisa & AVX512VBMI2) return "avx512vbmi2"; if(cpuisa & AVX512VBMI) return "avx512vbmi"; if(cpuisa & AVX512VNNI) return "avx512vnni"; if(cpuisa & AVX512VL) return "avx512vl"; if(cpuisa & AVX512BW) return "avx512bw"; if(cpuisa & AVX512CD) return "avx512cd"; if(cpuisa & AVX512ER) return "avx512er"; if(cpuisa & AVX512PF) return "avx512pf"; if(cpuisa & AVX512IFMA) return "avx512ifma"; if(cpuisa & AVX512DQ) return "avx512dq"; if(cpuisa & AVX512F) return "avx512f"; return "avx512"; } else if(cpuisa >= IS_AVX2) return "avx2"; else if(cpuisa >= IS_AVX) switch(cpuisa&0xf) { case 1: return "avx+fma3"; case 2: return "avx+fma4"; case 4: return "avx+aes"; case 5: return "avx+fma3+aes"; default:return "avx"; } else if(cpuisa >= IS_SSE42) return "sse4.2"; else if(cpuisa >= IS_SSE41x) return "sse4.1+popcnt"; else if(cpuisa >= IS_SSE41) return "sse4.1"; else if(cpuisa >= IS_SSSE3) return "ssse3"; else if(cpuisa >= IS_SSE3) return "sse3"; else if(cpuisa >= IS_SSE2) return "sse2"; else if(cpuisa >= IS_SSE) return "sse"; #elif defined(__powerpc64__) if(cpuisa >= IS_POWER9) return "power9"; #elif defined(__ARM_NEON) if(cpuisa >= IS_NEON) return "arm_neon"; #endif return "none"; } //--------------------------------------------------------------------------------- TB64FUNC _tb64e = tb64xenc; TB64FUNC _tb64d = tb64xdec; static int tb64set; void tb64ini(unsigned id, unsigned isshort) { int i; if(tb64set) return; tb64set++; i = id?id:cpuisa(); #if defined(__i386__) || defined(__x86_64__) #ifdef USE_AVX512 if(i >= IS_AVX512) { _tb64e = i >= (IS_AVX512|AVX512VL)?tb64avx2enc:tb64avx2enc; _tb64d = tb64avx512dec; } else #endif #ifndef NO_AVX2 if(i >= IS_AVX2) { _tb64e = isshort?_tb64avx2enc:tb64avx2enc; _tb64d = isshort?_tb64avx2dec:tb64avx2dec; } else #endif #ifndef NO_AVX if(i >= IS_AVX) { _tb64e = tb64avxenc; _tb64d = tb64avxdec; } else #endif #endif #if defined(__i386__) || defined(__x86_64__) || defined(__ARM_NEON) || defined(__powerpc64__) #ifndef NO_SSE if(i >= IS_SSSE3) { _tb64e = tb64sseenc; _tb64d = tb64ssedec; } #endif #endif } size_t tb64enc(const unsigned char *in, size_t inlen, unsigned char *out) { if(!tb64set) tb64ini(0,0); return _tb64e(in,inlen,out); } size_t tb64dec(const unsigned char *in, size_t inlen, unsigned char *out) { if(!tb64set) tb64ini(0,0); return _tb64d(in,inlen,out); } #endif