-
Notifications
You must be signed in to change notification settings - Fork 32
/
README
267 lines (192 loc) · 9.62 KB
/
README
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
erofs-utils
===========
Userspace tools for EROFS filesystem, currently including:
mkfs.erofs filesystem formatter
erofsfuse FUSE daemon alternative
dump.erofs filesystem analyzer
fsck.erofs filesystem compatibility & consistency checker as well
as extractor
EROFS filesystem overview
-------------------------
EROFS filesystem stands for Enhanced Read-Only File System. It aims to
form a generic read-only filesystem solution for various read-only use
cases instead of just focusing on storage space saving without
considering any side effects of runtime performance.
Typically EROFS could be considered in the following use scenarios:
- Firmwares in performance-sensitive systems, such as system
partitions of Android smartphones;
- Mountable immutable images such as container images for effective
metadata & data access compared with tar, cpio or other local
filesystems (e.g. ext4, XFS, btrfs, etc.)
- FSDAX-enabled rootfs for secure containers (Linux 5.15+);
- Live CDs which need a set of files with another high-performance
algorithm to optimize startup time; others files for archival
purposes only are not needed;
- and more.
Note that all EROFS metadata is uncompressed by design, so that you
could take EROFS as a drop-in read-only replacement of ext4, XFS,
btrfs, etc. without any compression-based dependencies and EROFS can
bring more effective filesystem accesses to users with reduced
metadata.
For more details of EROFS filesystem itself, please refer to:
https://www.kernel.org/doc/html/next/filesystems/erofs.html
For more details on how to build erofs-utils, see `docs/INSTALL.md`.
For more details about filesystem performance, see
`docs/PERFORMANCE.md`.
mkfs.erofs
----------
Two main kinds of EROFS images can be generated: (un)compressed images.
- For uncompressed images, there will be no compressed files in these
images. However, an EROFS image can contain files which consist of
various aligned data blocks and then a tail that is stored inline in
order to compact images [1].
- For compressed images, it will try to use the given algorithms first
for each regular file and see if storage space can be saved with
compression. If not, it will fall back to an uncompressed file.
Note that EROFS supports per-file compression configuration, proper
configuration options need to be enabled to parse compressed files by
the Linux kernel.
How to generate EROFS images
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Compression algorithms could be specified with the command-line option
`-z` to build a compressed EROFS image from a local directory:
$ mkfs.erofs -zlz4hc foo.erofs.img foo/
Supported algorithms by the Linux kernel:
- LZ4 (Linux 5.3+);
- LZMA (Linux 5.16+);
- DEFLATE (Linux 6.6+);
- Zstandard (Linux 6.10+).
Alternatively, generate an uncompressed EROFS from a local directory:
$ mkfs.erofs foo.erofs.img foo/
Additionally, you can specify a higher compression level to get a
(slightly) smaller image than the default level:
$ mkfs.erofs -zlz4hc,12 foo.erofs.img foo/
Multi-threaded support can be explicitly enabled with the ./configure
option `--enable-multithreading`; otherwise, single-threaded compression
will be used for now. It may take more time on multiprocessor platforms
if multi-threaded support is not enabled.
Currently, both `-Efragments` (not `-Eall-fragments`) and `-Ededupe`
don't support multi-threading due to time limitations.
Reproducible builds
~~~~~~~~~~~~~~~~~~~
Reproducible builds are typically used for verification and security,
ensuring the same binaries/distributions to be reproduced in a
deterministic way.
Images generated by the same version of `mkfs.erofs` will be identical
to previous runs if the same input is specified, and the same options
are used.
Specifically, variable timestamps and filesystem UUIDs can result in
unreproducible EROFS images. `-T` and `-U` can be used to fix them.
How to generate EROFS big pcluster images (Linux 5.13+)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
By default, EROFS formatter compresses data into separate one-block
(e.g. 4KiB) filesystem physical clusters for outstanding random read
performance. In other words, each EROFS filesystem block can be
independently decompressed. However, other similar filesystems
typically compress data into "blocks" of 128KiB or more for much smaller
images. Users may prefer smaller images for archiving purposes, even if
random performance is compromised with those configurations, and even
worse when using 4KiB blocks.
In order to fulfill users' needs, big plusters has been introduced
since Linux 5.13, in which each physical clusters will be more than one
blocks.
Specifically, `-C` is used to specify the maximum size of each pcluster
in bytes:
$ mkfs.erofs -zlz4hc -C65536 foo.erofs.img foo/
Thus, in this case, pcluster sizes can be up to 64KiB.
Note that large pcluster size can degrade random performance (though it
may improve sequential read performance for typical storage devices), so
please evaluate carefully in advance. Alternatively, you can make
per-(sub)file compression strategies according to file access patterns
if needed.
How to generate EROFS images with multiple algorithms
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
It's possible to generate an EROFS image with files in different
algorithms due to various purposes. For example, LZMA for archival
purposes and LZ4 for runtime purposes.
In order to use alternative algorithms, just specify two or more
compressing configurations together separated by ':' like below:
-zlzma:lz4hc,12:lzma,9 -C32768
Although mkfs still choose the first one by default, you could try to
write a compress-hints file like below:
4096 1 .*\.so$
32768 2 .*\.txt$
4096 sbin/.*$
16384 0 .*
and specify with `--compress-hints=` so that ".so" files will use
"lz4hc,12" compression with 4k pclusters, ".txt" files will use
"lzma,9" compression with 32k pclusters, files under "/sbin" will use
the default "lzma" compression with 4k plusters and other files will
use "lzma" compression with 16k pclusters.
Note that the largest pcluster size should be specified with the "-C"
option (here 32k pcluster size), otherwise all larger pclusters will be
limited.
How to generate well-compressed EROFS images
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Even if EROFS is not designed for such purposes in the beginning, it
could still produce some smaller images (not always) compared to other
approaches with better performance (see `docs/PERFORMANCE.md`). In
order to build well-compressed EROFS images, try the following options:
-C1048576 (5.13+)
-Eztailpacking (5.16+)
-Efragments / -Eall-fragments ( 6.1+);
-Ededupe ( 6.1+).
Also EROFS uses lz4hc level 9 by default, whereas some other approaches
use lz4hc level 12 by default. So please explicitly specify
`-zlz4hc,12 ` for comparison purposes.
How to generate legacy EROFS images (Linux 4.19+)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Decompression inplace and compacted indexes have been introduced in
Linux v5.3, which are not forward-compatible with older kernels.
In order to generate _legacy_ EROFS images for old kernels,
consider adding "-E legacy-compress" to the command line, e.g.
$ mkfs.erofs -E legacy-compress -zlz4hc foo.erofs.img foo/
For Linux kernel >= 5.3, legacy EROFS images are _NOT recommended_
due to runtime performance loss compared with non-legacy images.
Obsoleted erofs.mkfs
~~~~~~~~~~~~~~~~~~~~
There is an original erofs.mkfs version developed by Li Guifu,
which was replaced by the new erofs-utils implementation.
git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs-utils.git -b obsoleted_mkfs
PLEASE NOTE: This version is highly _NOT recommended_ now.
erofsfuse
---------
erofsfuse is introduced to support EROFS format for various platforms
(including older linux kernels) and new on-disk features iteration.
It can also be used as an unpacking tool for unprivileged users.
It supports fixed-sized output decompression *without* any in-place
I/O or in-place decompression optimization. Also like the other FUSE
implementations, it suffers from most common performance issues (e.g.
significant I/O overhead, double caching, etc.)
Therefore, NEVER use it if performance is the top concern.
How to mount an EROFS image with erofsfuse
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
As the other FUSE implementations, it's quite easy to mount by using
erofsfuse, e.g.:
$ erofsfuse foo.erofs.img foo/
Alternatively, to make it run in foreground (with debugging level 3):
$ erofsfuse -f --dbglevel=3 foo.erofs.img foo/
To debug erofsfuse (also automatically run in foreground):
$ erofsfuse -d foo.erofs.img foo/
To unmount an erofsfuse mountpoint as a non-root user:
$ fusermount -u foo/
dump.erofs and fsck.erofs
-------------------------
dump.erofs and fsck.erofs are used to analyze, check, and extract
EROFS filesystems. Note that extended attributes and ACLs are still
unsupported when extracting images with fsck.erofs.
Note that fragment extraction with fsck.erofs could be slow now and
it needs to be optimized later. If you are interested, contribution
is, as always, welcome.
Contribution
------------
erofs-utils is a part of EROFS filesystem project, which is completely
community-driven open source software. If you have interest in EROFS,
feel free to send feedback and/or patches to:
linux-erofs mailing list <[email protected]>
Comments
--------
[1] According to the EROFS on-disk format, the tail blocks of files
could be inlined aggressively with their metadata (called
tail-packing) in order to minimize the extra I/Os and the storage
space.