Documentation/lzo.txt

   1 ===========================================================
   2 LZO stream format as understood by Linux's LZO decompressor
   3 ===========================================================
   4
   5 Introduction
   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
  17 Description
  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
  81 Versions
  82
  83 0: Original version
  84 1: LZO-RLE
  85
  86 Version 1 of LZO implements an extension to encode runs of zeros using run
  87 length encoding. This improves speed for data with many zeros, which is a
  88 common 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
  91 For maximum compatibility, both versions are available under different names
  92 (lzo and lzo-rle). Differences in the encoding are noted in this document with
  93 e.g.: version 1 only.
  94
  95 Byte 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, the next byte
 106                 gives the bitstream version (version 1 only). If the first byte
 107                 is not 17, the bitstream version is 0.
 108
 109       18..21  : copy 0..3 literals
 110                 state = (byte - 17) = 0..3  [ copy <state> literals ]
 111                 skip byte
 112
 113       22..255 : copy literal string
 114                 length = (byte - 17) = 4..238
 115                 state = 4 [ don't copy extra literals ]
 116                 skip byte
 117
 118   Instruction encoding::
 119
 120       0 0 0 0 X X X X  (0..15)
 121         Depends on the number of literals copied by the last instruction.
 122         If last instruction did not copy any literal (state == 0), this
 123         encoding will be a copy of 4 or more literal, and must be interpreted
 124         like this :
 125
 126            0 0 0 0 L L L L  (0..15)  : copy long literal string
 127            length = 3 + (L ?: 15 + (zero_bytes * 255) + non_zero_byte)
 128            state = 4  (no extra literals are copied)
 129
 130         If last instruction used to copy between 1 to 3 literals (encoded in
 131         the instruction's opcode or distance), the instruction is a copy of a
 132         2-byte block from the dictionary within a 1kB distance. It is worth
 133         noting that this instruction provides little savings since it uses 2
 134         bytes to encode a copy of 2 other bytes but it encodes the number of
 135         following literals for free. It must be interpreted like this :
 136
 137            0 0 0 0 D D S S  (0..15)  : copy 2 bytes from <= 1kB distance
 138            length = 2
 139            state = S (copy S literals after this block)
 140          Always followed by exactly one byte : H H H H H H H H
 141            distance = (H << 2) + D + 1
 142
 143         If last instruction used to copy 4 or more literals (as detected by
 144         state == 4), the instruction becomes a copy of a 3-byte block from the
 145         dictionary from a 2..3kB distance, and must be interpreted like this :
 146
 147            0 0 0 0 D D S S  (0..15)  : copy 3 bytes from 2..3 kB distance
 148            length = 3
 149            state = S (copy S literals after this block)
 150          Always followed by exactly one byte : H H H H H H H H
 151            distance = (H << 2) + D + 2049
 152
 153       0 0 0 1 H L L L  (16..31)
 154            Copy of a block within 16..48kB distance (preferably less than 10B)
 155            length = 2 + (L ?: 7 + (zero_bytes * 255) + non_zero_byte)
 156         Always followed by exactly one LE16 :  D D D D D D D D : D D D D D D S S
 157            distance = 16384 + (H << 14) + D
 158            state = S (copy S literals after this block)
 159            End of stream is reached if distance == 16384
 160
 161         In version 1 only, this instruction is also used to encode a run of
 162         zeros if distance = 0xbfff, i.e. H = 1 and the D bits are all 1.
 163            In this case, it is followed by a fourth byte, X.
 164            run length = ((X << 3) | (0 0 0 0 0 L L L)) + 4.
 165
 166       0 0 1 L L L L L  (32..63)
 167            Copy of small block within 16kB distance (preferably less than 34B)
 168            length = 2 + (L ?: 31 + (zero_bytes * 255) + non_zero_byte)
 169         Always followed by exactly one LE16 :  D D D D D D D D : D D D D D D S S
 170            distance = D + 1
 171            state = S (copy S literals after this block)
 172
 173       0 1 L D D D S S  (64..127)
 174            Copy 3-4 bytes from block within 2kB distance
 175            state = S (copy S literals after this block)
 176            length = 3 + L
 177          Always followed by exactly one byte : H H H H H H H H
 178            distance = (H << 3) + D + 1
 179
 180       1 L L D D D S S  (128..255)
 181            Copy 5-8 bytes from block within 2kB distance
 182            state = S (copy S literals after this block)
 183            length = 5 + L
 184          Always followed by exactly one byte : H H H H H H H H
 185            distance = (H << 3) + D + 1
 186
 187 Authors
 188 =======
 189
 190   This document was written by Willy Tarreau <w@1wt.eu> on 2014/07/19 during an
 191   analysis of the decompression code available in Linux 3.16-rc5, and updated
 192   by Dave Rodgman <dave.rodgman@arm.com> on 2018/10/30 to introduce run-length
 193   encoding. The code is tricky, it is possible that this document contains
 194   mistakes or that a few corner cases were overlooked. In any case, please
 195   report any doubt, fix, or proposed updates to the author(s) so that the
 196   document can be updated.