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