xref: /linux/Documentation/admin-guide/sysctl/fs.rst (revision 7f71507851fc7764b36a3221839607d3a45c2025)
1===============================
2Documentation for /proc/sys/fs/
3===============================
4
5Copyright (c) 1998, 1999,  Rik van Riel <riel@nl.linux.org>
6
7Copyright (c) 2009,        Shen Feng<shen@cn.fujitsu.com>
8
9For general info and legal blurb, please look in intro.rst.
10
11------------------------------------------------------------------------------
12
13This file contains documentation for the sysctl files and directories
14in ``/proc/sys/fs/``.
15
16The files in this directory can be used to tune and monitor
17miscellaneous and general things in the operation of the Linux
18kernel. Since some of the files *can* be used to screw up your
19system, it is advisable to read both documentation and source
20before actually making adjustments.
21
221. /proc/sys/fs
23===============
24
25Currently, these files might (depending on your configuration)
26show up in ``/proc/sys/fs``:
27
28.. contents:: :local:
29
30
31aio-nr & aio-max-nr
32-------------------
33
34``aio-nr`` shows the current system-wide number of asynchronous io
35requests.  ``aio-max-nr`` allows you to change the maximum value
36``aio-nr`` can grow to.  If ``aio-nr`` reaches ``aio-nr-max`` then
37``io_setup`` will fail with ``EAGAIN``.  Note that raising
38``aio-max-nr`` does not result in the
39pre-allocation or re-sizing of any kernel data structures.
40
41dentry-negative
42----------------------------
43
44Policy for negative dentries. Set to 1 to to always delete the dentry when a
45file is removed, and 0 to disable it. By default, this behavior is disabled.
46
47dentry-state
48------------
49
50This file shows the values in ``struct dentry_stat_t``, as defined in
51``fs/dcache.c``::
52
53  struct dentry_stat_t dentry_stat {
54        long nr_dentry;
55        long nr_unused;
56        long age_limit;         /* age in seconds */
57        long want_pages;        /* pages requested by system */
58        long nr_negative;       /* # of unused negative dentries */
59        long dummy;             /* Reserved for future use */
60  };
61
62Dentries are dynamically allocated and deallocated.
63
64``nr_dentry`` shows the total number of dentries allocated (active
65+ unused). ``nr_unused shows`` the number of dentries that are not
66actively used, but are saved in the LRU list for future reuse.
67
68``age_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``nr_negative`` shows the number of unused dentries that are also
74negative dentries which do not map to any files. Instead,
75they help speeding up rejection of non-existing files provided
76by the users.
77
78
79file-max & file-nr
80------------------
81
82The value in ``file-max`` denotes the maximum number of file-
83handles that the Linux kernel will allocate. When you get lots
84of error messages about running out of file handles, you might
85want to increase this limit.
86
87Historically,the kernel was able to allocate file handles
88dynamically, but not to free them again. The three values in
89``file-nr`` denote the number of allocated file handles, the number
90of allocated but unused file handles, and the maximum number of
91file handles. Linux 2.6 and later always reports 0 as the number of free
92file handles -- this is not an error, it just means that the
93number of allocated file handles exactly matches the number of
94used file handles.
95
96Attempts to allocate more file descriptors than ``file-max`` are
97reported with ``printk``, look for::
98
99  VFS: file-max limit <number> reached
100
101in the kernel logs.
102
103
104inode-nr & inode-state
105----------------------
106
107As with file handles, the kernel allocates the inode structures
108dynamically, but can't free them yet.
109
110The file ``inode-nr`` contains the first two items from
111``inode-state``, so we'll skip to that file...
112
113``inode-state`` contains three actual numbers and four dummies.
114The actual numbers are, in order of appearance, ``nr_inodes``,
115``nr_free_inodes`` and ``preshrink``.
116
117``nr_inodes`` stands for the number of inodes the system has
118allocated.
119
120``nr_free_inodes`` represents the number of free inodes (?) and
121preshrink is nonzero when the
122system needs to prune the inode list instead of allocating
123more.
124
125
126mount-max
127---------
128
129This denotes the maximum number of mounts that may exist
130in a mount namespace.
131
132
133nr_open
134-------
135
136This denotes the maximum number of file-handles a process can
137allocate. Default value is 1024*1024 (1048576) which should be
138enough for most machines. Actual limit depends on ``RLIMIT_NOFILE``
139resource limit.
140
141
142overflowgid & overflowuid
143-------------------------
144
145Some filesystems only support 16-bit UIDs and GIDs, although in Linux
146UIDs and GIDs are 32 bits. When one of these filesystems is mounted
147with writes enabled, any UID or GID that would exceed 65535 is translated
148to a fixed value before being written to disk.
149
150These sysctls allow you to change the value of the fixed UID and GID.
151The default is 65534.
152
153
154pipe-user-pages-hard
155--------------------
156
157Maximum total number of pages a non-privileged user may allocate for pipes.
158Once this limit is reached, no new pipes may be allocated until usage goes
159below the limit again. When set to 0, no limit is applied, which is the default
160setting.
161
162
163pipe-user-pages-soft
164--------------------
165
166Maximum total number of pages a non-privileged user may allocate for pipes
167before the pipe size gets limited to a single page. Once this limit is reached,
168new pipes will be limited to a single page in size for this user in order to
169limit total memory usage, and trying to increase them using ``fcntl()`` will be
170denied until usage goes below the limit again. The default value allows to
171allocate up to 1024 pipes at their default size. When set to 0, no limit is
172applied.
173
174
175protected_fifos
176---------------
177
178The intent of this protection is to avoid unintentional writes to
179an attacker-controlled FIFO, where a program expected to create a regular
180file.
181
182When set to "0", writing to FIFOs is unrestricted.
183
184When set to "1" don't allow ``O_CREAT`` open on FIFOs that we don't own
185in world writable sticky directories, unless they are owned by the
186owner of the directory.
187
188When set to "2" it also applies to group writable sticky directories.
189
190This protection is based on the restrictions in Openwall.
191
192
193protected_hardlinks
194--------------------
195
196A long-standing class of security issues is the hardlink-based
197time-of-check-time-of-use race, most commonly seen in world-writable
198directories like ``/tmp``. The common method of exploitation of this flaw
199is to cross privilege boundaries when following a given hardlink (i.e. a
200root process follows a hardlink created by another user). Additionally,
201on systems without separated partitions, this stops unauthorized users
202from "pinning" vulnerable setuid/setgid files against being upgraded by
203the administrator, or linking to special files.
204
205When set to "0", hardlink creation behavior is unrestricted.
206
207When set to "1" hardlinks cannot be created by users if they do not
208already own the source file, or do not have read/write access to it.
209
210This protection is based on the restrictions in Openwall and grsecurity.
211
212
213protected_regular
214-----------------
215
216This protection is similar to `protected_fifos`_, but it
217avoids writes to an attacker-controlled regular file, where a program
218expected to create one.
219
220When set to "0", writing to regular files is unrestricted.
221
222When set to "1" don't allow ``O_CREAT`` open on regular files that we
223don't own in world writable sticky directories, unless they are
224owned by the owner of the directory.
225
226When set to "2" it also applies to group writable sticky directories.
227
228
229protected_symlinks
230------------------
231
232A long-standing class of security issues is the symlink-based
233time-of-check-time-of-use race, most commonly seen in world-writable
234directories like ``/tmp``. The common method of exploitation of this flaw
235is to cross privilege boundaries when following a given symlink (i.e. a
236root process follows a symlink belonging to another user). For a likely
237incomplete list of hundreds of examples across the years, please see:
238https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=/tmp
239
240When set to "0", symlink following behavior is unrestricted.
241
242When set to "1" symlinks are permitted to be followed only when outside
243a sticky world-writable directory, or when the uid of the symlink and
244follower match, or when the directory owner matches the symlink's owner.
245
246This protection is based on the restrictions in Openwall and grsecurity.
247
248
249suid_dumpable
250-------------
251
252This value can be used to query and set the core dump mode for setuid
253or otherwise protected/tainted binaries. The modes are
254
255=   ==========  ===============================================================
2560   (default)	Traditional behaviour. Any process which has changed
257		privilege levels or is execute only will not be dumped.
2581   (debug)	All processes dump core when possible. The core dump is
259		owned by the current user and no security is applied. This is
260		intended for system debugging situations only.
261		Ptrace is unchecked.
262		This is insecure as it allows regular users to examine the
263		memory contents of privileged processes.
2642   (suidsafe)	Any binary which normally would not be dumped is dumped
265		anyway, but only if the ``core_pattern`` kernel sysctl (see
266		:ref:`Documentation/admin-guide/sysctl/kernel.rst <core_pattern>`)
267		is set to
268		either a pipe handler or a fully qualified path. (For more
269		details on this limitation, see CVE-2006-2451.) This mode is
270		appropriate when administrators are attempting to debug
271		problems in a normal environment, and either have a core dump
272		pipe handler that knows to treat privileged core dumps with
273		care, or specific directory defined for catching core dumps.
274		If a core dump happens without a pipe handler or fully
275		qualified path, a message will be emitted to syslog warning
276		about the lack of a correct setting.
277=   ==========  ===============================================================
278
279
280
2812. /proc/sys/fs/binfmt_misc
282===========================
283
284Documentation for the files in ``/proc/sys/fs/binfmt_misc`` is
285in Documentation/admin-guide/binfmt-misc.rst.
286
287
2883. /proc/sys/fs/mqueue - POSIX message queues filesystem
289========================================================
290
291
292The "mqueue"  filesystem provides  the necessary kernel features to enable the
293creation of a  user space  library that  implements  the  POSIX message queues
294API (as noted by the  MSG tag in the  POSIX 1003.1-2001 version  of the System
295Interfaces specification.)
296
297The "mqueue" filesystem contains values for determining/setting the
298amount of resources used by the file system.
299
300``/proc/sys/fs/mqueue/queues_max`` is a read/write file for
301setting/getting the maximum number of message queues allowed on the
302system.
303
304``/proc/sys/fs/mqueue/msg_max`` is a read/write file for
305setting/getting the maximum number of messages in a queue value.  In
306fact it is the limiting value for another (user) limit which is set in
307``mq_open`` invocation.  This attribute of a queue must be less than
308or equal to ``msg_max``.
309
310``/proc/sys/fs/mqueue/msgsize_max`` is a read/write file for
311setting/getting the maximum message size value (it is an attribute of
312every message queue, set during its creation).
313
314``/proc/sys/fs/mqueue/msg_default`` is a read/write file for
315setting/getting the default number of messages in a queue value if the
316``attr`` parameter of ``mq_open(2)`` is ``NULL``. If it exceeds
317``msg_max``, the default value is initialized to ``msg_max``.
318
319``/proc/sys/fs/mqueue/msgsize_default`` is a read/write file for
320setting/getting the default message size value if the ``attr``
321parameter of ``mq_open(2)`` is ``NULL``. If it exceeds
322``msgsize_max``, the default value is initialized to ``msgsize_max``.
323
3244. /proc/sys/fs/epoll - Configuration options for the epoll interface
325=====================================================================
326
327This directory contains configuration options for the epoll(7) interface.
328
329max_user_watches
330----------------
331
332Every epoll file descriptor can store a number of files to be monitored
333for event readiness. Each one of these monitored files constitutes a "watch".
334This configuration option sets the maximum number of "watches" that are
335allowed for each user.
336Each "watch" costs roughly 90 bytes on a 32-bit kernel, and roughly 160 bytes
337on a 64-bit one.
338The current default value for ``max_user_watches`` is 4% of the
339available low memory, divided by the "watch" cost in bytes.
340
3415. /proc/sys/fs/fuse - Configuration options for FUSE filesystems
342=====================================================================
343
344This directory contains the following configuration options for FUSE
345filesystems:
346
347``/proc/sys/fs/fuse/max_pages_limit`` is a read/write file for
348setting/getting the maximum number of pages that can be used for servicing
349requests in FUSE.
350