xref: /openbmc/linux/fs/squashfs/Kconfig (revision 84d517f3)
1config SQUASHFS
2	tristate "SquashFS 4.0 - Squashed file system support"
3	depends on BLOCK
4	help
5	  Saying Y here includes support for SquashFS 4.0 (a Compressed
6	  Read-Only File System).  Squashfs is a highly compressed read-only
7	  filesystem for Linux.  It uses zlib, lzo or xz compression to
8	  compress both files, inodes and directories.  Inodes in the system
9	  are very small and all blocks are packed to minimise data overhead.
10	  Block sizes greater than 4K are supported up to a maximum of 1 Mbytes
11	  (default block size 128K).  SquashFS 4.0 supports 64 bit filesystems
12	  and files (larger than 4GB), full uid/gid information, hard links and
13	  timestamps.
14
15	  Squashfs is intended for general read-only filesystem use, for
16	  archival use (i.e. in cases where a .tar.gz file may be used), and in
17	  embedded systems where low overhead is needed.  Further information
18	  and tools are available from http://squashfs.sourceforge.net.
19
20	  If you want to compile this as a module ( = code which can be
21	  inserted in and removed from the running kernel whenever you want),
22	  say M here.  The module will be called squashfs.  Note that the root
23	  file system (the one containing the directory /) cannot be compiled
24	  as a module.
25
26	  If unsure, say N.
27
28choice
29	prompt "File decompression options"
30	depends on SQUASHFS
31	help
32	  Squashfs now supports two options for decompressing file
33	  data.  Traditionally Squashfs has decompressed into an
34	  intermediate buffer and then memcopied it into the page cache.
35	  Squashfs now supports the ability to decompress directly into
36	  the page cache.
37
38	  If unsure, select "Decompress file data into an intermediate buffer"
39
40config SQUASHFS_FILE_CACHE
41	bool "Decompress file data into an intermediate buffer"
42	help
43	  Decompress file data into an intermediate buffer and then
44	  memcopy it into the page cache.
45
46config SQUASHFS_FILE_DIRECT
47	bool "Decompress files directly into the page cache"
48	help
49	  Directly decompress file data into the page cache.
50	  Doing so can significantly improve performance because
51	  it eliminates a memcpy and it also removes the lock contention
52	  on the single buffer.
53
54endchoice
55
56choice
57	prompt "Decompressor parallelisation options"
58	depends on SQUASHFS
59	help
60	  Squashfs now supports three parallelisation options for
61	  decompression.  Each one exhibits various trade-offs between
62	  decompression performance and CPU and memory usage.
63
64	  If in doubt, select "Single threaded compression"
65
66config SQUASHFS_DECOMP_SINGLE
67	bool "Single threaded compression"
68	help
69	  Traditionally Squashfs has used single-threaded decompression.
70	  Only one block (data or metadata) can be decompressed at any
71	  one time.  This limits CPU and memory usage to a minimum.
72
73config SQUASHFS_DECOMP_MULTI
74	bool "Use multiple decompressors for parallel I/O"
75	help
76	  By default Squashfs uses a single decompressor but it gives
77	  poor performance on parallel I/O workloads when using multiple CPU
78	  machines due to waiting on decompressor availability.
79
80	  If you have a parallel I/O workload and your system has enough memory,
81	  using this option may improve overall I/O performance.
82
83	  This decompressor implementation uses up to two parallel
84	  decompressors per core.  It dynamically allocates decompressors
85	  on a demand basis.
86
87config SQUASHFS_DECOMP_MULTI_PERCPU
88	bool "Use percpu multiple decompressors for parallel I/O"
89	help
90	  By default Squashfs uses a single decompressor but it gives
91	  poor performance on parallel I/O workloads when using multiple CPU
92	  machines due to waiting on decompressor availability.
93
94	  This decompressor implementation uses a maximum of one
95	  decompressor per core.  It uses percpu variables to ensure
96	  decompression is load-balanced across the cores.
97
98endchoice
99
100config SQUASHFS_XATTR
101	bool "Squashfs XATTR support"
102	depends on SQUASHFS
103	help
104	  Saying Y here includes support for extended attributes (xattrs).
105	  Xattrs are name:value pairs associated with inodes by
106	  the kernel or by users (see the attr(5) manual page).
107
108	  If unsure, say N.
109
110config SQUASHFS_ZLIB
111	bool "Include support for ZLIB compressed file systems"
112	depends on SQUASHFS
113	select ZLIB_INFLATE
114	default y
115	help
116	  ZLIB compression is the standard compression used by Squashfs
117	  file systems.  It offers a good trade-off between compression
118	  achieved and the amount of CPU time and memory necessary to
119	  compress and decompress.
120
121	  If unsure, say Y.
122
123config SQUASHFS_LZO
124	bool "Include support for LZO compressed file systems"
125	depends on SQUASHFS
126	select LZO_DECOMPRESS
127	help
128	  Saying Y here includes support for reading Squashfs file systems
129	  compressed with LZO compression.  LZO compression is mainly
130	  aimed at embedded systems with slower CPUs where the overheads
131	  of zlib are too high.
132
133	  LZO is not the standard compression used in Squashfs and so most
134	  file systems will be readable without selecting this option.
135
136	  If unsure, say N.
137
138config SQUASHFS_XZ
139	bool "Include support for XZ compressed file systems"
140	depends on SQUASHFS
141	select XZ_DEC
142	help
143	  Saying Y here includes support for reading Squashfs file systems
144	  compressed with XZ compression.  XZ gives better compression than
145	  the default zlib compression, at the expense of greater CPU and
146	  memory overhead.
147
148	  XZ is not the standard compression used in Squashfs and so most
149	  file systems will be readable without selecting this option.
150
151	  If unsure, say N.
152
153config SQUASHFS_4K_DEVBLK_SIZE
154	bool "Use 4K device block size?"
155	depends on SQUASHFS
156	help
157	  By default Squashfs sets the dev block size (sb_min_blocksize)
158	  to 1K or the smallest block size supported by the block device
159	  (if larger).  This, because blocks are packed together and
160	  unaligned in Squashfs, should reduce latency.
161
162	  This, however, gives poor performance on MTD NAND devices where
163	  the optimal I/O size is 4K (even though the devices can support
164	  smaller block sizes).
165
166	  Using a 4K device block size may also improve overall I/O
167	  performance for some file access patterns (e.g. sequential
168	  accesses of files in filesystem order) on all media.
169
170	  Setting this option will force Squashfs to use a 4K device block
171	  size by default.
172
173	  If unsure, say N.
174
175config SQUASHFS_EMBEDDED
176	bool "Additional option for memory-constrained systems"
177	depends on SQUASHFS
178	help
179	  Saying Y here allows you to specify cache size.
180
181	  If unsure, say N.
182
183config SQUASHFS_FRAGMENT_CACHE_SIZE
184	int "Number of fragments cached" if SQUASHFS_EMBEDDED
185	depends on SQUASHFS
186	default "3"
187	help
188	  By default SquashFS caches the last 3 fragments read from
189	  the filesystem.  Increasing this amount may mean SquashFS
190	  has to re-read fragments less often from disk, at the expense
191	  of extra system memory.  Decreasing this amount will mean
192	  SquashFS uses less memory at the expense of extra reads from disk.
193
194	  Note there must be at least one cached fragment.  Anything
195	  much more than three will probably not make much difference.
196