blob: 88fd7f5c8dcd61307171b3af852541d06a984380 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001Documentation for /proc/sys/fs/* kernel version 2.2.10
2 (c) 1998, 1999, Rik van Riel <riel@nl.linux.org>
Shen Feng760df932009-04-02 16:57:20 -07003 (c) 2009, Shen Feng<shen@cn.fujitsu.com>
Linus Torvalds1da177e2005-04-16 15:20:36 -07004
5For general info and legal blurb, please look in README.
6
7==============================================================
8
9This file contains documentation for the sysctl files in
10/proc/sys/fs/ and is valid for Linux kernel version 2.2.
11
12The files in this directory can be used to tune and monitor
13miscellaneous and general things in the operation of the Linux
14kernel. Since some of the files _can_ be used to screw up your
15system, it is advisable to read both documentation and source
16before actually making adjustments.
17
Shen Feng760df932009-04-02 16:57:20 -0700181. /proc/sys/fs
19----------------------------------------------------------
20
Linus Torvalds1da177e2005-04-16 15:20:36 -070021Currently, these files are in /proc/sys/fs:
Shen Feng760df932009-04-02 16:57:20 -070022- aio-max-nr
23- aio-nr
Linus Torvalds1da177e2005-04-16 15:20:36 -070024- dentry-state
25- dquot-max
26- dquot-nr
27- file-max
28- file-nr
29- inode-max
30- inode-nr
31- inode-state
Eric Dumazet9cfe0152008-02-06 01:37:16 -080032- nr_open
Linus Torvalds1da177e2005-04-16 15:20:36 -070033- overflowuid
34- overflowgid
Alexey Dobriyana2e0b562006-08-27 01:23:28 -070035- suid_dumpable
Linus Torvalds1da177e2005-04-16 15:20:36 -070036- super-max
37- super-nr
38
Shen Feng760df932009-04-02 16:57:20 -070039==============================================================
40
41aio-nr & aio-max-nr:
42
43aio-nr is the running total of the number of events specified on the
44io_setup system call for all currently active aio contexts. If aio-nr
45reaches aio-max-nr then io_setup will fail with EAGAIN. Note that
46raising aio-max-nr does not result in the pre-allocation or re-sizing
47of any kernel data structures.
Linus Torvalds1da177e2005-04-16 15:20:36 -070048
49==============================================================
50
51dentry-state:
52
53From linux/fs/dentry.c:
54--------------------------------------------------------------
55struct {
56 int nr_dentry;
57 int nr_unused;
58 int age_limit; /* age in seconds */
59 int want_pages; /* pages requested by system */
60 int dummy[2];
61} dentry_stat = {0, 0, 45, 0,};
62--------------------------------------------------------------
63
64Dentries are dynamically allocated and deallocated, and
65nr_dentry seems to be 0 all the time. Hence it's safe to
66assume that only nr_unused, age_limit and want_pages are
67used. Nr_unused seems to be exactly what its name says.
68Age_limit is the age in seconds after which dcache entries
69can be reclaimed when memory is short and want_pages is
70nonzero when shrink_dcache_pages() has been called and the
71dcache isn't pruned yet.
72
73==============================================================
74
75dquot-max & dquot-nr:
76
77The file dquot-max shows the maximum number of cached disk
78quota entries.
79
80The file dquot-nr shows the number of allocated disk quota
81entries and the number of free disk quota entries.
82
83If the number of free cached disk quotas is very low and
84you have some awesome number of simultaneous system users,
85you might want to raise the limit.
86
87==============================================================
88
89file-max & file-nr:
90
Linus Torvalds1da177e2005-04-16 15:20:36 -070091The value in file-max denotes the maximum number of file-
92handles that the Linux kernel will allocate. When you get lots
93of error messages about running out of file handles, you might
94want to increase this limit.
95
Federica Teodorica3b78a2011-03-15 16:12:05 -070096Historically,the kernel was able to allocate file handles
97dynamically, but not to free them again. The three values in
98file-nr denote the number of allocated file handles, the number
99of allocated but unused file handles, and the maximum number of
100file handles. Linux 2.6 always reports 0 as the number of free
101file handles -- this is not an error, it just means that the
102number of allocated file handles exactly matches the number of
103used file handles.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700104
Xiaotian Fengbcadbbd2009-09-23 15:56:13 -0700105Attempts to allocate more file descriptors than file-max are
106reported with printk, look for "VFS: file-max limit <number>
107reached".
Linus Torvalds1da177e2005-04-16 15:20:36 -0700108==============================================================
109
Eric Dumazet9cfe0152008-02-06 01:37:16 -0800110nr_open:
111
112This denotes the maximum number of file-handles a process can
113allocate. Default value is 1024*1024 (1048576) which should be
114enough for most machines. Actual limit depends on RLIMIT_NOFILE
115resource limit.
116
117==============================================================
118
Linus Torvalds1da177e2005-04-16 15:20:36 -0700119inode-max, inode-nr & inode-state:
120
121As with file handles, the kernel allocates the inode structures
122dynamically, but can't free them yet.
123
124The value in inode-max denotes the maximum number of inode
125handlers. This value should be 3-4 times larger than the value
126in file-max, since stdin, stdout and network sockets also
127need an inode struct to handle them. When you regularly run
128out of inodes, you need to increase this value.
129
130The file inode-nr contains the first two items from
131inode-state, so we'll skip to that file...
132
133Inode-state contains three actual numbers and four dummies.
134The actual numbers are, in order of appearance, nr_inodes,
135nr_free_inodes and preshrink.
136
137Nr_inodes stands for the number of inodes the system has
138allocated, this can be slightly more than inode-max because
139Linux allocates them one pageful at a time.
140
141Nr_free_inodes represents the number of free inodes (?) and
142preshrink is nonzero when the nr_inodes > inode-max and the
143system needs to prune the inode list instead of allocating
144more.
145
146==============================================================
147
148overflowgid & overflowuid:
149
150Some filesystems only support 16-bit UIDs and GIDs, although in Linux
151UIDs and GIDs are 32 bits. When one of these filesystems is mounted
152with writes enabled, any UID or GID that would exceed 65535 is translated
153to a fixed value before being written to disk.
154
155These sysctls allow you to change the value of the fixed UID and GID.
156The default is 65534.
157
158==============================================================
159
Alexey Dobriyana2e0b562006-08-27 01:23:28 -0700160suid_dumpable:
161
162This value can be used to query and set the core dump mode for setuid
163or otherwise protected/tainted binaries. The modes are
164
1650 - (default) - traditional behaviour. Any process which has changed
166 privilege levels or is execute only will not be dumped
1671 - (debug) - all processes dump core when possible. The core dump is
168 owned by the current user and no security is applied. This is
169 intended for system debugging situations only. Ptrace is unchecked.
1702 - (suidsafe) - any binary which normally would not be dumped is dumped
171 readable by root only. This allows the end user to remove
172 such a dump but not access it directly. For security reasons
173 core dumps in this mode will not overwrite one another or
Matt LaPlante5d3f0832006-11-30 05:21:10 +0100174 other files. This mode is appropriate when administrators are
Alexey Dobriyana2e0b562006-08-27 01:23:28 -0700175 attempting to debug problems in a normal environment.
176
177==============================================================
178
Linus Torvalds1da177e2005-04-16 15:20:36 -0700179super-max & super-nr:
180
181These numbers control the maximum number of superblocks, and
182thus the maximum number of mounted filesystems the kernel
183can have. You only need to increase super-max if you need to
184mount more filesystems than the current value in super-max
185allows you to.
186
187==============================================================
188
189aio-nr & aio-max-nr:
190
191aio-nr shows the current system-wide number of asynchronous io
192requests. aio-max-nr allows you to change the maximum value
193aio-nr can grow to.
194
195==============================================================
Shen Feng760df932009-04-02 16:57:20 -0700196
197
1982. /proc/sys/fs/binfmt_misc
199----------------------------------------------------------
200
201Documentation for the files in /proc/sys/fs/binfmt_misc is
202in Documentation/binfmt_misc.txt.
203
204
2053. /proc/sys/fs/mqueue - POSIX message queues filesystem
206----------------------------------------------------------
207
208The "mqueue" filesystem provides the necessary kernel features to enable the
209creation of a user space library that implements the POSIX message queues
210API (as noted by the MSG tag in the POSIX 1003.1-2001 version of the System
211Interfaces specification.)
212
213The "mqueue" filesystem contains values for determining/setting the amount of
214resources used by the file system.
215
216/proc/sys/fs/mqueue/queues_max is a read/write file for setting/getting the
217maximum number of message queues allowed on the system.
218
219/proc/sys/fs/mqueue/msg_max is a read/write file for setting/getting the
220maximum number of messages in a queue value. In fact it is the limiting value
221for another (user) limit which is set in mq_open invocation. This attribute of
222a queue must be less or equal then msg_max.
223
224/proc/sys/fs/mqueue/msgsize_max is a read/write file for setting/getting the
225maximum message size value (it is every message queue's attribute set during
226its creation).
227
228
2294. /proc/sys/fs/epoll - Configuration options for the epoll interface
230--------------------------------------------------------
231
232This directory contains configuration options for the epoll(7) interface.
233
Shen Feng760df932009-04-02 16:57:20 -0700234max_user_watches
235----------------
236
237Every epoll file descriptor can store a number of files to be monitored
238for event readiness. Each one of these monitored files constitutes a "watch".
239This configuration option sets the maximum number of "watches" that are
240allowed for each user.
241Each "watch" costs roughly 90 bytes on a 32bit kernel, and roughly 160 bytes
242on a 64bit one.
243The current default value for max_user_watches is the 1/32 of the available
244low memory, divided for the "watch" cost in bytes.
245