1===========================================================
2LZO stream format as understood by Linux's LZO decompressor
3===========================================================
4
5Introduction
6============
7
8  This is not a specification. No specification seems to be publicly available
9  for the LZO stream format. This document describes what input format the LZO
10  decompressor as implemented in the Linux kernel understands. The file subject
11  of this analysis is lib/lzo/lzo1x_decompress_safe.c. No analysis was made on
12  the compressor nor on any other implementations though it seems likely that
13  the format matches the standard one. The purpose of this document is to
14  better understand what the code does in order to propose more efficient fixes
15  for future bug reports.
16
17Description
18===========
19
20  The stream is composed of a series of instructions, operands, and data. The
21  instructions consist in a few bits representing an opcode, and bits forming
22  the operands for the instruction, whose size and position depend on the
23  opcode and on the number of literals copied by previous instruction. The
24  operands are used to indicate:
25
26    - a distance when copying data from the dictionary (past output buffer)
27    - a length (number of bytes to copy from dictionary)
28    - the number of literals to copy, which is retained in variable "state"
29      as a piece of information for next instructions.
30
31  Optionally depending on the opcode and operands, extra data may follow. These
32  extra data can be a complement for the operand (eg: a length or a distance
33  encoded on larger values), or a literal to be copied to the output buffer.
34
35  The first byte of the block follows a different encoding from other bytes, it
36  seems to be optimized for literal use only, since there is no dictionary yet
37  prior to that byte.
38
39  Lengths are always encoded on a variable size starting with a small number
40  of bits in the operand. If the number of bits isn't enough to represent the
41  length, up to 255 may be added in increments by consuming more bytes with a
42  rate of at most 255 per extra byte (thus the compression ratio cannot exceed
43  around 255:1). The variable length encoding using #bits is always the same::
44
45       length = byte & ((1 << #bits) - 1)
46       if (!length) {
47               length = ((1 << #bits) - 1)
48               length += 255*(number of zero bytes)
49               length += first-non-zero-byte
50       }
51       length += constant (generally 2 or 3)
52
53  For references to the dictionary, distances are relative to the output
54  pointer. Distances are encoded using very few bits belonging to certain
55  ranges, resulting in multiple copy instructions using different encodings.
56  Certain encodings involve one extra byte, others involve two extra bytes
57  forming a little-endian 16-bit quantity (marked LE16 below).
58
59  After any instruction except the large literal copy, 0, 1, 2 or 3 literals
60  are copied before starting the next instruction. The number of literals that
61  were copied may change the meaning and behaviour of the next instruction. In
62  practice, only one instruction needs to know whether 0, less than 4, or more
63  literals were copied. This is the information stored in the <state> variable
64  in this implementation. This number of immediate literals to be copied is
65  generally encoded in the last two bits of the instruction but may also be
66  taken from the last two bits of an extra operand (eg: distance).
67
68  End of stream is declared when a block copy of distance 0 is seen. Only one
69  instruction may encode this distance (0001HLLL), it takes one LE16 operand
70  for the distance, thus requiring 3 bytes.
71
72  .. important::
73
74     In the code some length checks are missing because certain instructions
75     are called under the assumption that a certain number of bytes follow
76     because it has already been guaranteed before parsing the instructions.
77     They just have to "refill" this credit if they consume extra bytes. This
78     is an implementation design choice independent on the algorithm or
79     encoding.
80
81Versions
82
830: Original version
841: LZO-RLE
85
86Version 1 of LZO implements an extension to encode runs of zeros using run
87length encoding. This improves speed for data with many zeros, which is a
88common case for zram. This modifies the bitstream in a backwards compatible way
89(v1 can correctly decompress v0 compressed data, but v0 cannot read v1 data).
90
91For maximum compatibility, both versions are available under different names
92(lzo and lzo-rle). Differences in the encoding are noted in this document with
93e.g.: version 1 only.
94
95Byte sequences
96==============
97
98  First byte encoding::
99
100      0..16   : follow regular instruction encoding, see below. It is worth
101                noting that code 16 will represent a block copy from the
102                dictionary which is empty, and that it will always be
103                invalid at this place.
104
105      17      : bitstream version. If the first byte is 17, and compressed
106                stream length is at least 5 bytes (length of shortest possible
107                versioned bitstream), the next byte gives the bitstream version
108                (version 1 only).
109                Otherwise, the bitstream version is 0.
110
111      18..21  : copy 0..3 literals
112                state = (byte - 17) = 0..3  [ copy <state> literals ]
113                skip byte
114
115      22..255 : copy literal string
116                length = (byte - 17) = 4..238
117                state = 4 [ don't copy extra literals ]
118                skip byte
119
120  Instruction encoding::
121
122      0 0 0 0 X X X X  (0..15)
123        Depends on the number of literals copied by the last instruction.
124        If last instruction did not copy any literal (state == 0), this
125        encoding will be a copy of 4 or more literal, and must be interpreted
126        like this :
127
128           0 0 0 0 L L L L  (0..15)  : copy long literal string
129           length = 3 + (L ?: 15 + (zero_bytes * 255) + non_zero_byte)
130           state = 4  (no extra literals are copied)
131
132        If last instruction used to copy between 1 to 3 literals (encoded in
133        the instruction's opcode or distance), the instruction is a copy of a
134        2-byte block from the dictionary within a 1kB distance. It is worth
135        noting that this instruction provides little savings since it uses 2
136        bytes to encode a copy of 2 other bytes but it encodes the number of
137        following literals for free. It must be interpreted like this :
138
139           0 0 0 0 D D S S  (0..15)  : copy 2 bytes from <= 1kB distance
140           length = 2
141           state = S (copy S literals after this block)
142         Always followed by exactly one byte : H H H H H H H H
143           distance = (H << 2) + D + 1
144
145        If last instruction used to copy 4 or more literals (as detected by
146        state == 4), the instruction becomes a copy of a 3-byte block from the
147        dictionary from a 2..3kB distance, and must be interpreted like this :
148
149           0 0 0 0 D D S S  (0..15)  : copy 3 bytes from 2..3 kB distance
150           length = 3
151           state = S (copy S literals after this block)
152         Always followed by exactly one byte : H H H H H H H H
153           distance = (H << 2) + D + 2049
154
155      0 0 0 1 H L L L  (16..31)
156           Copy of a block within 16..48kB distance (preferably less than 10B)
157           length = 2 + (L ?: 7 + (zero_bytes * 255) + non_zero_byte)
158        Always followed by exactly one LE16 :  D D D D D D D D : D D D D D D S S
159           distance = 16384 + (H << 14) + D
160           state = S (copy S literals after this block)
161           End of stream is reached if distance == 16384
162           In version 1 only, to prevent ambiguity with the RLE case when
163           ((distance & 0x803f) == 0x803f) && (261 <= length <= 264), the
164           compressor must not emit block copies where distance and length
165           meet these conditions.
166
167        In version 1 only, this instruction is also used to encode a run of
168           zeros if distance = 0xbfff, i.e. H = 1 and the D bits are all 1.
169           In this case, it is followed by a fourth byte, X.
170           run length = ((X << 3) | (0 0 0 0 0 L L L)) + 4
171
172      0 0 1 L L L L L  (32..63)
173           Copy of small block within 16kB distance (preferably less than 34B)
174           length = 2 + (L ?: 31 + (zero_bytes * 255) + non_zero_byte)
175        Always followed by exactly one LE16 :  D D D D D D D D : D D D D D D S S
176           distance = D + 1
177           state = S (copy S literals after this block)
178
179      0 1 L D D D S S  (64..127)
180           Copy 3-4 bytes from block within 2kB distance
181           state = S (copy S literals after this block)
182           length = 3 + L
183         Always followed by exactly one byte : H H H H H H H H
184           distance = (H << 3) + D + 1
185
186      1 L L D D D S S  (128..255)
187           Copy 5-8 bytes from block within 2kB distance
188           state = S (copy S literals after this block)
189           length = 5 + L
190         Always followed by exactly one byte : H H H H H H H H
191           distance = (H << 3) + D + 1
192
193Authors
194=======
195
196  This document was written by Willy Tarreau <w@1wt.eu> on 2014/07/19 during an
197  analysis of the decompression code available in Linux 3.16-rc5, and updated
198  by Dave Rodgman <dave.rodgman@arm.com> on 2018/10/30 to introduce run-length
199  encoding. The code is tricky, it is possible that this document contains
200  mistakes or that a few corner cases were overlooked. In any case, please
201  report any doubt, fix, or proposed updates to the author(s) so that the
202  document can be updated.
203