1*a28cd43dSSascha Wildner /*
2*a28cd43dSSascha Wildner * Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
3*a28cd43dSSascha Wildner * All rights reserved.
4*a28cd43dSSascha Wildner *
5*a28cd43dSSascha Wildner * This source code is licensed under both the BSD-style license (found in the
6*a28cd43dSSascha Wildner * LICENSE file in the root directory of this source tree) and the GPLv2 (found
7*a28cd43dSSascha Wildner * in the COPYING file in the root directory of this source tree).
8*a28cd43dSSascha Wildner * You may select, at your option, one of the above-listed licenses.
9*a28cd43dSSascha Wildner */
10*a28cd43dSSascha Wildner
11*a28cd43dSSascha Wildner /*-*************************************
12*a28cd43dSSascha Wildner * Dependencies
13*a28cd43dSSascha Wildner ***************************************/
14*a28cd43dSSascha Wildner #include "zstd_compress_literals.h"
15*a28cd43dSSascha Wildner
ZSTD_noCompressLiterals(void * dst,size_t dstCapacity,const void * src,size_t srcSize)16*a28cd43dSSascha Wildner size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
17*a28cd43dSSascha Wildner {
18*a28cd43dSSascha Wildner BYTE* const ostart = (BYTE* const)dst;
19*a28cd43dSSascha Wildner U32 const flSize = 1 + (srcSize>31) + (srcSize>4095);
20*a28cd43dSSascha Wildner
21*a28cd43dSSascha Wildner RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall, "");
22*a28cd43dSSascha Wildner
23*a28cd43dSSascha Wildner switch(flSize)
24*a28cd43dSSascha Wildner {
25*a28cd43dSSascha Wildner case 1: /* 2 - 1 - 5 */
26*a28cd43dSSascha Wildner ostart[0] = (BYTE)((U32)set_basic + (srcSize<<3));
27*a28cd43dSSascha Wildner break;
28*a28cd43dSSascha Wildner case 2: /* 2 - 2 - 12 */
29*a28cd43dSSascha Wildner MEM_writeLE16(ostart, (U16)((U32)set_basic + (1<<2) + (srcSize<<4)));
30*a28cd43dSSascha Wildner break;
31*a28cd43dSSascha Wildner case 3: /* 2 - 2 - 20 */
32*a28cd43dSSascha Wildner MEM_writeLE32(ostart, (U32)((U32)set_basic + (3<<2) + (srcSize<<4)));
33*a28cd43dSSascha Wildner break;
34*a28cd43dSSascha Wildner default: /* not necessary : flSize is {1,2,3} */
35*a28cd43dSSascha Wildner assert(0);
36*a28cd43dSSascha Wildner }
37*a28cd43dSSascha Wildner
38*a28cd43dSSascha Wildner ZSTD_memcpy(ostart + flSize, src, srcSize);
39*a28cd43dSSascha Wildner DEBUGLOG(5, "Raw literals: %u -> %u", (U32)srcSize, (U32)(srcSize + flSize));
40*a28cd43dSSascha Wildner return srcSize + flSize;
41*a28cd43dSSascha Wildner }
42*a28cd43dSSascha Wildner
ZSTD_compressRleLiteralsBlock(void * dst,size_t dstCapacity,const void * src,size_t srcSize)43*a28cd43dSSascha Wildner size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
44*a28cd43dSSascha Wildner {
45*a28cd43dSSascha Wildner BYTE* const ostart = (BYTE* const)dst;
46*a28cd43dSSascha Wildner U32 const flSize = 1 + (srcSize>31) + (srcSize>4095);
47*a28cd43dSSascha Wildner
48*a28cd43dSSascha Wildner (void)dstCapacity; /* dstCapacity already guaranteed to be >=4, hence large enough */
49*a28cd43dSSascha Wildner
50*a28cd43dSSascha Wildner switch(flSize)
51*a28cd43dSSascha Wildner {
52*a28cd43dSSascha Wildner case 1: /* 2 - 1 - 5 */
53*a28cd43dSSascha Wildner ostart[0] = (BYTE)((U32)set_rle + (srcSize<<3));
54*a28cd43dSSascha Wildner break;
55*a28cd43dSSascha Wildner case 2: /* 2 - 2 - 12 */
56*a28cd43dSSascha Wildner MEM_writeLE16(ostart, (U16)((U32)set_rle + (1<<2) + (srcSize<<4)));
57*a28cd43dSSascha Wildner break;
58*a28cd43dSSascha Wildner case 3: /* 2 - 2 - 20 */
59*a28cd43dSSascha Wildner MEM_writeLE32(ostart, (U32)((U32)set_rle + (3<<2) + (srcSize<<4)));
60*a28cd43dSSascha Wildner break;
61*a28cd43dSSascha Wildner default: /* not necessary : flSize is {1,2,3} */
62*a28cd43dSSascha Wildner assert(0);
63*a28cd43dSSascha Wildner }
64*a28cd43dSSascha Wildner
65*a28cd43dSSascha Wildner ostart[flSize] = *(const BYTE*)src;
66*a28cd43dSSascha Wildner DEBUGLOG(5, "RLE literals: %u -> %u", (U32)srcSize, (U32)flSize + 1);
67*a28cd43dSSascha Wildner return flSize+1;
68*a28cd43dSSascha Wildner }
69*a28cd43dSSascha Wildner
ZSTD_compressLiterals(ZSTD_hufCTables_t const * prevHuf,ZSTD_hufCTables_t * nextHuf,ZSTD_strategy strategy,int disableLiteralCompression,void * dst,size_t dstCapacity,const void * src,size_t srcSize,void * entropyWorkspace,size_t entropyWorkspaceSize,const int bmi2)70*a28cd43dSSascha Wildner size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
71*a28cd43dSSascha Wildner ZSTD_hufCTables_t* nextHuf,
72*a28cd43dSSascha Wildner ZSTD_strategy strategy, int disableLiteralCompression,
73*a28cd43dSSascha Wildner void* dst, size_t dstCapacity,
74*a28cd43dSSascha Wildner const void* src, size_t srcSize,
75*a28cd43dSSascha Wildner void* entropyWorkspace, size_t entropyWorkspaceSize,
76*a28cd43dSSascha Wildner const int bmi2)
77*a28cd43dSSascha Wildner {
78*a28cd43dSSascha Wildner size_t const minGain = ZSTD_minGain(srcSize, strategy);
79*a28cd43dSSascha Wildner size_t const lhSize = 3 + (srcSize >= 1 KB) + (srcSize >= 16 KB);
80*a28cd43dSSascha Wildner BYTE* const ostart = (BYTE*)dst;
81*a28cd43dSSascha Wildner U32 singleStream = srcSize < 256;
82*a28cd43dSSascha Wildner symbolEncodingType_e hType = set_compressed;
83*a28cd43dSSascha Wildner size_t cLitSize;
84*a28cd43dSSascha Wildner
85*a28cd43dSSascha Wildner DEBUGLOG(5,"ZSTD_compressLiterals (disableLiteralCompression=%i srcSize=%u)",
86*a28cd43dSSascha Wildner disableLiteralCompression, (U32)srcSize);
87*a28cd43dSSascha Wildner
88*a28cd43dSSascha Wildner /* Prepare nextEntropy assuming reusing the existing table */
89*a28cd43dSSascha Wildner ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
90*a28cd43dSSascha Wildner
91*a28cd43dSSascha Wildner if (disableLiteralCompression)
92*a28cd43dSSascha Wildner return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
93*a28cd43dSSascha Wildner
94*a28cd43dSSascha Wildner /* small ? don't even attempt compression (speed opt) */
95*a28cd43dSSascha Wildner # define COMPRESS_LITERALS_SIZE_MIN 63
96*a28cd43dSSascha Wildner { size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
97*a28cd43dSSascha Wildner if (srcSize <= minLitSize) return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
98*a28cd43dSSascha Wildner }
99*a28cd43dSSascha Wildner
100*a28cd43dSSascha Wildner RETURN_ERROR_IF(dstCapacity < lhSize+1, dstSize_tooSmall, "not enough space for compression");
101*a28cd43dSSascha Wildner { HUF_repeat repeat = prevHuf->repeatMode;
102*a28cd43dSSascha Wildner int const preferRepeat = strategy < ZSTD_lazy ? srcSize <= 1024 : 0;
103*a28cd43dSSascha Wildner if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1;
104*a28cd43dSSascha Wildner cLitSize = singleStream ?
105*a28cd43dSSascha Wildner HUF_compress1X_repeat(
106*a28cd43dSSascha Wildner ostart+lhSize, dstCapacity-lhSize, src, srcSize,
107*a28cd43dSSascha Wildner HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
108*a28cd43dSSascha Wildner (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2) :
109*a28cd43dSSascha Wildner HUF_compress4X_repeat(
110*a28cd43dSSascha Wildner ostart+lhSize, dstCapacity-lhSize, src, srcSize,
111*a28cd43dSSascha Wildner HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
112*a28cd43dSSascha Wildner (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2);
113*a28cd43dSSascha Wildner if (repeat != HUF_repeat_none) {
114*a28cd43dSSascha Wildner /* reused the existing table */
115*a28cd43dSSascha Wildner DEBUGLOG(5, "Reusing previous huffman table");
116*a28cd43dSSascha Wildner hType = set_repeat;
117*a28cd43dSSascha Wildner }
118*a28cd43dSSascha Wildner }
119*a28cd43dSSascha Wildner
120*a28cd43dSSascha Wildner if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) {
121*a28cd43dSSascha Wildner ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
122*a28cd43dSSascha Wildner return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
123*a28cd43dSSascha Wildner }
124*a28cd43dSSascha Wildner if (cLitSize==1) {
125*a28cd43dSSascha Wildner ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
126*a28cd43dSSascha Wildner return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
127*a28cd43dSSascha Wildner }
128*a28cd43dSSascha Wildner
129*a28cd43dSSascha Wildner if (hType == set_compressed) {
130*a28cd43dSSascha Wildner /* using a newly constructed table */
131*a28cd43dSSascha Wildner nextHuf->repeatMode = HUF_repeat_check;
132*a28cd43dSSascha Wildner }
133*a28cd43dSSascha Wildner
134*a28cd43dSSascha Wildner /* Build header */
135*a28cd43dSSascha Wildner switch(lhSize)
136*a28cd43dSSascha Wildner {
137*a28cd43dSSascha Wildner case 3: /* 2 - 2 - 10 - 10 */
138*a28cd43dSSascha Wildner { U32 const lhc = hType + ((!singleStream) << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<14);
139*a28cd43dSSascha Wildner MEM_writeLE24(ostart, lhc);
140*a28cd43dSSascha Wildner break;
141*a28cd43dSSascha Wildner }
142*a28cd43dSSascha Wildner case 4: /* 2 - 2 - 14 - 14 */
143*a28cd43dSSascha Wildner { U32 const lhc = hType + (2 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<18);
144*a28cd43dSSascha Wildner MEM_writeLE32(ostart, lhc);
145*a28cd43dSSascha Wildner break;
146*a28cd43dSSascha Wildner }
147*a28cd43dSSascha Wildner case 5: /* 2 - 2 - 18 - 18 */
148*a28cd43dSSascha Wildner { U32 const lhc = hType + (3 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<22);
149*a28cd43dSSascha Wildner MEM_writeLE32(ostart, lhc);
150*a28cd43dSSascha Wildner ostart[4] = (BYTE)(cLitSize >> 10);
151*a28cd43dSSascha Wildner break;
152*a28cd43dSSascha Wildner }
153*a28cd43dSSascha Wildner default: /* not possible : lhSize is {3,4,5} */
154*a28cd43dSSascha Wildner assert(0);
155*a28cd43dSSascha Wildner }
156*a28cd43dSSascha Wildner DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)srcSize, (U32)(lhSize+cLitSize));
157*a28cd43dSSascha Wildner return lhSize+cLitSize;
158*a28cd43dSSascha Wildner }
159