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