5 Q: What do the letters XZ mean?
7 A: Nothing. They are just two letters, which come from the file format
8 suffix .xz. The .xz suffix was selected, because it seemed to be
9 pretty much unused. It has no deeper meaning.
12 Q: What are LZMA and LZMA2?
14 A: LZMA stands for Lempel-Ziv-Markov chain-Algorithm. It is the name
15 of the compression algorithm designed by Igor Pavlov for 7-Zip.
16 LZMA is based on LZ77 and range encoding.
18 LZMA2 is an updated version of the original LZMA to fix a couple of
19 practical issues. In context of XZ Utils, LZMA is called LZMA1 to
20 emphasize that LZMA is not the same thing as LZMA2. LZMA2 is the
21 primary compression algorithm in the .xz file format.
24 Q: There are many LZMA related projects. How does XZ Utils relate to them?
26 A: 7-Zip and LZMA SDK are the original projects. LZMA SDK is roughly
27 a subset of the 7-Zip source tree.
29 p7zip is 7-Zip's command-line tools ported to POSIX-like systems.
31 LZMA Utils provide a gzip-like lzma tool for POSIX-like systems.
32 LZMA Utils are based on LZMA SDK. XZ Utils are the successor to
35 There are several other projects using LZMA. Most are more or less
36 based on LZMA SDK. See <https://7-zip.org/links.html>.
39 Q: Why is liblzma named liblzma if its primary file format is .xz?
40 Shouldn't it be e.g. libxz?
42 A: When the designing of the .xz format began, the idea was to replace
43 the .lzma format and use the same .lzma suffix. It would have been
44 quite OK to reuse the suffix when there were very few .lzma files
45 around. However, the old .lzma format became popular before the
46 new format was finished. The new format was renamed to .xz but the
47 name of liblzma wasn't changed.
50 Q: Do XZ Utils support the .7z format?
52 A: No. Use 7-Zip (Windows) or p7zip (POSIX-like systems) to handle .7z
56 Q: I have many .tar.7z files. Can I convert them to .tar.xz without
57 spending hours recompressing the data?
59 A: In the "extra" directory, there is a script named 7z2lzma.bash which
60 is able to convert some .7z files to the .lzma format (not .xz). It
61 needs the 7za (or 7z) command from p7zip. The script may silently
62 produce corrupt output if certain assumptions are not met, so
63 decompress the resulting .lzma file and compare it against the
64 original before deleting the original file!
67 Q: I have many .lzma files. Can I quickly convert them to the .xz format?
69 A: For now, no. Since XZ Utils supports the .lzma format, it's usually
70 not too bad to keep the old files in the old format. If you want to
71 do the conversion anyway, you need to decompress the .lzma files and
72 then recompress to the .xz format.
74 Technically, there is a way to make the conversion relatively fast
75 (roughly twice the time that normal decompression takes). Writing
76 such a tool would take quite a bit of time though, and would probably
77 be useful to only a few people. If you really want such a conversion
78 tool, contact Lasse Collin and offer some money.
81 Q: I have installed xz, but my tar doesn't recognize .tar.xz files.
82 How can I extract .tar.xz files?
84 A: xz -dc foo.tar.xz | tar xf -
87 Q: Can I recover parts of a broken .xz file (e.g. a corrupted CD-R)?
89 A: It may be possible if the file consists of multiple blocks, which
90 typically is not the case if the file was created in single-threaded
91 mode. There is no recovery program yet.
94 Q: Is (some part of) XZ Utils patented?
96 A: Lasse Collin is not aware of any patents that could affect XZ Utils.
97 However, due to the nature of software patents, it's not possible to
98 guarantee that XZ Utils isn't affected by any third party patent(s).
101 Q: Where can I find documentation about the file format and algorithms?
103 A: The .xz format is documented in xz-file-format.txt. It is a container
104 format only, and doesn't include descriptions of any non-trivial
107 Documenting LZMA and LZMA2 is planned, but for now, there is no other
108 documentation than the source code. Before you begin, you should know
109 the basics of LZ77 and range-coding algorithms. LZMA is based on LZ77,
110 but LZMA is a lot more complex. Range coding is used to compress
111 the final bitstream like Huffman coding is used in Deflate.
114 Q: I cannot find BCJ and BCJ2 filters. Don't they exist in liblzma?
116 A: BCJ filter is called "x86" in liblzma. BCJ2 is not included,
117 because it requires using more than one encoded output stream.
120 Q: I need to use a script that runs "xz -9". On a system with 256 MiB
121 of RAM, xz says that it cannot allocate memory. Can I make the
122 script work without modifying it?
124 A: Set a default memory usage limit for compression. You can do it e.g.
125 in a shell initialization script such as ~/.bashrc or /etc/profile:
127 XZ_DEFAULTS=--memlimit-compress=150MiB
130 xz will then scale the compression settings down so that the given
131 memory usage limit is not reached. This way xz shouldn't run out
134 Check also that memory-related resource limits are high enough.
135 On most systems, "ulimit -a" will show the current resource limits.
138 Q: How do I create files that can be decompressed with XZ Embedded?
140 A: See the documentation in XZ Embedded. In short, something like
141 this is a good start:
143 xz --check=crc32 --lzma2=preset=6e,dict=64KiB
145 Or if a BCJ filter is needed too, e.g. if compressing
146 a kernel image for PowerPC:
148 xz --check=crc32 --powerpc --lzma2=preset=6e,dict=64KiB
150 Adjust the dictionary size to get a good compromise between
151 compression ratio and decompressor memory usage. Note that
152 in single-call decompression mode of XZ Embedded, a big
153 dictionary doesn't increase memory usage.
156 Q: How is multi-threaded compression implemented in XZ Utils?
158 A: The simplest method is splitting the uncompressed data into blocks
159 and compressing them in parallel independent from each other.
160 This is currently the only threading method supported in XZ Utils.
161 Since the blocks are compressed independently, they can also be
162 decompressed independently. Together with the index feature in .xz,
163 this allows using threads to create .xz files for random-access
164 reading. This also makes threaded decompression possible.
166 The independent blocks method has a couple of disadvantages too. It
167 will compress worse than a single-block method. Often the difference
168 is not too big (maybe 1-2 %) but sometimes it can be too big. Also,
169 the memory usage of the compressor increases linearly when adding
172 At least two other threading methods are possible but these haven't
173 been implemented in XZ Utils:
175 Match finder parallelization has been in 7-Zip for ages. It doesn't
176 affect compression ratio or memory usage significantly. Among the
177 three threading methods, only this is useful when compressing small
178 files (files that are not significantly bigger than the dictionary).
179 Unfortunately this method scales only to about two CPU cores.
181 The third method is pigz-style threading (I use that name, because
182 pigz <https://www.zlib.net/pigz/> uses that method). It doesn't
183 affect compression ratio significantly and scales to many cores.
184 The memory usage scales linearly when threads are added. This isn't
185 significant with pigz, because Deflate uses only a 32 KiB dictionary,
186 but with LZMA2 the memory usage will increase dramatically just like
187 with the independent-blocks method. There is also a constant
188 computational overhead, which may make pigz-method a bit dull on
189 dual-core compared to the parallel match finder method, but with more
190 cores the overhead is not a big deal anymore.
192 Combining the threading methods will be possible and also useful.
193 For example, combining match finder parallelization with pigz-style
194 threading or independent-blocks-threading can cut the memory usage
198 Q: I told xz to use many threads but it is using only one or two
199 processor cores. What is wrong?
201 A: Since multi-threaded compression is done by splitting the data into
202 blocks that are compressed individually, if the input file is too
203 small for the block size, then many threads cannot be used. The
204 default block size increases when the compression level is
205 increased. For example, xz -6 uses 8 MiB LZMA2 dictionary and
206 24 MiB blocks, and xz -9 uses 64 MiB LZMA dictionary and 192 MiB
207 blocks. If the input file is 100 MiB, xz -6 can use five threads
208 of which one will finish quickly as it has only 4 MiB to compress.
209 However, for the same file, xz -9 can only use one thread.
211 One can adjust block size with --block-size=SIZE but making the
212 block size smaller than LZMA2 dictionary is waste of RAM: using
213 xz -9 with 6 MiB blocks isn't any better than using xz -6 with
214 6 MiB blocks. The default settings use a block size bigger than
215 the LZMA2 dictionary size because this was seen as a reasonable
216 compromise between RAM usage and compression ratio.
218 When decompressing, the ability to use threads depends on how the
219 file was created. If it was created in multi-threaded mode then
220 it can be decompressed in multi-threaded mode too if there are
221 multiple blocks in the file.
224 Q: How do I build a program that needs liblzmadec (lzmadec.h)?
226 A: liblzmadec is part of LZMA Utils. XZ Utils has liblzma, but no
227 liblzmadec. The code using liblzmadec should be ported to use
228 liblzma instead. If you cannot or don't want to do that, download
229 LZMA Utils from <https://tukaani.org/lzma/>.
232 Q: The default build of liblzma is too big. How can I make it smaller?
234 A: Give --enable-small to the configure script. Use also appropriate
235 --enable or --disable options to include only those filter encoders
236 and decoders and integrity checks that you actually need. Use
237 CFLAGS=-Os (with GCC) or equivalent to tell your compiler to optimize
238 for size. See INSTALL for information about configure options.
240 If the result is still too big, take a look at XZ Embedded. It is
241 a separate project, which provides a limited but significantly
242 smaller XZ decoder implementation than XZ Utils. You can find it
243 at <https://tukaani.org/xz/embedded.html>.