1 .\" Copyright (C) 1994, 1995 by Daniel Quinlan (quinlan@yggdrasil.com)
2 .\" and Copyright (C) 2002-2008,2017 Michael Kerrisk <mtk.manpages@gmail.com>
3 .\" with networking additions from Alan Cox (A.Cox@swansea.ac.uk)
4 .\" and scsi additions from Michael Neuffer (neuffer@mail.uni-mainz.de)
5 .\" and sysctl additions from Andries Brouwer (aeb@cwi.nl)
6 .\" and System V IPC (as well as various other) additions from
7 .\" Michael Kerrisk <mtk.manpages@gmail.com>
9 .\" %%%LICENSE_START(GPLv2+_DOC_FULL)
10 .\" This is free documentation; you can redistribute it and/or
11 .\" modify it under the terms of the GNU General Public License as
12 .\" published by the Free Software Foundation; either version 2 of
13 .\" the License, or (at your option) any later version.
15 .\" The GNU General Public License's references to "object code"
16 .\" and "executables" are to be interpreted as the output of any
17 .\" document formatting or typesetting system, including
18 .\" intermediate and printed output.
20 .\" This manual is distributed in the hope that it will be useful,
21 .\" but WITHOUT ANY WARRANTY; without even the implied warranty of
22 .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
23 .\" GNU General Public License for more details.
25 .\" You should have received a copy of the GNU General Public
26 .\" License along with this manual; if not, see
27 .\" <http://www.gnu.org/licenses/>.
30 .\" Modified 1995-05-17 by faith@cs.unc.edu
31 .\" Minor changes by aeb and Marty Leisner (leisner@sdsp.mc.xerox.com).
32 .\" Modified 1996-04-13, 1996-07-22 by aeb@cwi.nl
33 .\" Modified 2001-12-16 by rwhron@earthlink.net
34 .\" Modified 2002-07-13 by jbelton@shaw.ca
35 .\" Modified 2002-07-22, 2003-05-27, 2004-04-06, 2004-05-25
36 .\" by Michael Kerrisk <mtk.manpages@gmail.com>
37 .\" 2004-11-17, mtk -- updated notes on /proc/loadavg
38 .\" 2004-12-01, mtk, rtsig-max and rtsig-nr went away in 2.6.8
39 .\" 2004-12-14, mtk, updated 'statm', and fixed error in order of list
40 .\" 2005-05-12, mtk, updated 'stat'
41 .\" 2005-07-13, mtk, added /proc/sys/fs/mqueue/*
42 .\" 2005-09-16, mtk, Added /proc/sys/fs/suid_dumpable
43 .\" 2005-09-19, mtk, added /proc/zoneinfo
44 .\" 2005-03-01, mtk, moved /proc/sys/fs/mqueue/* material to mq_overview.7.
45 .\" 2008-06-05, mtk, Added /proc/[pid]/oom_score, /proc/[pid]/oom_adj,
46 .\" /proc/[pid]/limits, /proc/[pid]/mountinfo, /proc/[pid]/mountstats,
47 .\" and /proc/[pid]/fdinfo/*.
48 .\" 2008-06-19, mtk, Documented /proc/[pid]/status.
49 .\" 2008-07-15, mtk, added /proc/config.gz
51 .\" FIXME cross check against Documentation/filesystems/proc.txt
52 .\" to see what information could be imported from that file
55 .TH PROC 5 2019-11-19 "Linux" "Linux Programmer's Manual"
57 proc \- process information pseudo-filesystem
61 filesystem is a pseudo-filesystem which provides an interface to
62 kernel data structures.
63 It is commonly mounted at
65 Typically, it is mounted automatically by the system,
66 but it can also be mounted manually using a command such as:
70 mount \-t proc proc /proc
74 Most of the files in the
76 filesystem are read-only,
77 but some files are writable, allowing kernel variables to be changed.
82 filesystem supports the following mount options:
84 .BR hidepid "=\fIn\fP (since Linux 3.3)"
85 .\" commit 0499680a42141d86417a8fbaa8c8db806bea1201
86 This option controls who can access the information in
91 is one of the following values:
95 Everybody may access all
98 This is the traditional behavior,
99 and the default if this mount option is not specified.
102 Users may not access files and subdirectories inside any
104 directories but their own (the
106 directories themselves remain visible).
107 Sensitive files such as
108 .IR /proc/[pid]/cmdline
110 .IR /proc/[pid]/status
111 are now protected against other users.
112 This makes it impossible to learn whether any user is running a
114 (so long as the program doesn't otherwise reveal itself by its behavior).
115 .\" As an additional bonus, since
116 .\" .IR /proc/[pid]/cmdline
117 .\" is unaccessible for other users,
118 .\" poorly written programs passing sensitive information via
119 .\" program arguments are now protected against local eavesdroppers.
122 As for mode 1, but in addition the
124 directories belonging to other users become invisible.
127 entries can no longer be used to discover the PIDs on the system.
128 This doesn't hide the fact that a process with a specific PID value exists
129 (it can be learned by other means, for example, by "kill \-0 $PID"),
130 but it hides a process's UID and GID,
131 which could otherwise be learned by employing
136 This greatly complicates an attacker's task of gathering
137 information about running processes (e.g., discovering whether
138 some daemon is running with elevated privileges,
139 whether another user is running some sensitive program,
140 whether other users are running any program at all, and so on).
143 .BR gid "=\fIgid\fP (since Linux 3.3)"
144 .\" commit 0499680a42141d86417a8fbaa8c8db806bea1201
145 Specifies the ID of a group whose members are authorized to
146 learn process information otherwise prohibited by
148 (i.e., users in this group behave as though
152 This group should be used instead of approaches such as putting
153 nonroot users into the
160 there are the following general groups of files and subdirectories:
162 .IR /proc/[pid] " subdirectories"
163 Each one of these subdirectories contains files and subdirectories
164 exposing information about the process with the corresponding process ID.
166 Underneath each of the
170 subdirectory contains subdirectories of the form
172 which contain corresponding information about each of the threads
173 in the process, where
175 is the kernel thread ID of the thread.
179 subdirectories are visible when iterating through
183 (and thus are visible when one uses
185 to view the contents of
188 .IR /proc/[tid] " subdirectories"
189 Each one of these subdirectories contains files and subdirectories
190 exposing information about the thread with the corresponding thread ID.
191 The contents of these directories are the same as the corresponding
192 .IR /proc/[pid]/task/[tid]
199 visible when iterating through
205 visible when one uses
207 to view the contents of
211 When a process accesses this magic symbolic link,
212 it resolves to the process's own
216 .I /proc/thread\-self
217 When a thread accesses this magic symbolic link,
218 it resolves to the process's own
219 .I /proc/self/task/[tid]
223 Various other files and subdirectories under
225 expose system-wide information.
227 All of the above are described in more detail below.
229 .SS Files and directories
230 The following list provides details of many of the files and directories
236 There is a numerical subdirectory for each running process; the
237 subdirectory is named by the process ID.
240 subdirectory contains the pseudo-files and directories described below.
242 The files inside each
244 directory are normally owned by the effective user and
245 effective group ID of the process.
246 However, as a security measure, the ownership is made
248 if the process's "dumpable" attribute is set to a value other than 1.
251 .\" commit 68eb94f16227336a5773b83ecfa8290f1d6b78ce
253 meant the "global" root user ID and group ID
254 (i.e., UID 0 and GID 0 in the initial user namespace).
256 if the process is in a noninitial user namespace that has a
257 valid mapping for user (group) ID 0 inside the namespace, then
258 the user (group) ownership of the files under
260 is instead made the same as the root user (group) ID of the namespace.
261 This means that inside a container,
262 things work as expected for the container "root" user.
264 The process's "dumpable" attribute may change for the following reasons:
267 The attribute was explicitly set via the
272 The attribute was reset to the value in the file
273 .IR /proc/sys/fs/suid_dumpable
274 (described below), for the reasons described in
278 Resetting the "dumpable" attribute to 1 reverts the ownership of the
280 files to the process's effective UID and GID.
283 .\" https://lwn.net/Articles/28222/
284 .\" From: Stephen Smalley <sds@epoch.ncsc.mil>
285 .\" To: LKML and others
286 .\" Subject: [RFC][PATCH] Process Attribute API for Security Modules
287 .\" Date: 08 Apr 2003 16:17:52 -0400
289 .\" http://www.nsa.gov/research/_files/selinux/papers/module/x362.shtml
291 The files in this directory provide an API for security modules.
292 The contents of this directory are files that can be read and written
293 in order to set security-related attributes.
294 This directory was added to support SELinux,
295 but the intention was that the API be general enough to support
296 other security modules.
297 For the purpose of explanation,
298 examples of how SELinux uses these files are provided below.
300 This directory is present only if the kernel was configured with
301 .BR CONFIG_SECURITY .
303 .IR /proc/[pid]/attr/current " (since Linux 2.6.0)"
304 The contents of this file represent the current
305 security attributes of the process.
307 In SELinux, this file is used to get the security context of a process.
308 Prior to Linux 2.6.11, this file could not be used to set the security
309 context (a write was always denied), since SELinux limited process security
312 (see the description of
313 .IR /proc/[pid]/attr/exec ,
315 Since Linux 2.6.11, SELinux lifted this restriction and began supporting
316 "set" operations via writes to this node if authorized by policy,
317 although use of this operation is only suitable for applications that are
318 trusted to maintain any desired separation between the old and new security
321 Prior to Linux 2.6.28, SELinux did not allow threads within a
322 multi-threaded process to set their security context via this node
323 as it would yield an inconsistency among the security contexts of the
324 threads sharing the same memory space.
325 Since Linux 2.6.28, SELinux lifted
326 this restriction and began supporting "set" operations for threads within
327 a multithreaded process if the new security context is bounded by the old
328 security context, where the bounded relation is defined in policy and
329 guarantees that the new security context has a subset of the permissions
330 of the old security context.
332 Other security modules may choose to support "set" operations via
335 .IR /proc/[pid]/attr/exec " (since Linux 2.6.0)"
336 This file represents the attributes to assign to the
337 process upon a subsequent
341 this is needed to support role/domain transitions, and
343 is the preferred point to make such transitions because it offers better
344 control over the initialization of the process in the new security label
345 and the inheritance of state.
346 In SELinux, this attribute is reset on
348 so that the new program reverts to the default behavior for any
350 calls that it may make.
351 In SELinux, a process can set
353 .I /proc/[pid]/attr/exec
356 .IR /proc/[pid]/attr/fscreate " (since Linux 2.6.0)"
357 This file represents the attributes to assign to files
358 created by subsequent calls to
365 SELinux employs this file to support creation of a file
366 (using the aforementioned system calls)
368 so that there is no risk of inappropriate access being obtained
369 between the time of creation and the time that attributes are set.
370 In SELinux, this attribute is reset on
372 so that the new program reverts to the default behavior for
373 any file creation calls it may make, but the attribute will persist
374 across multiple file creation calls within a program unless it is
376 In SELinux, a process can set only its own
377 .IR /proc/[pid]/attr/fscreate
380 .IR /proc/[pid]/attr/keycreate " (since Linux 2.6.18)"
381 .\" commit 4eb582cf1fbd7b9e5f466e3718a59c957e75254e
382 If a process writes a security context into this file,
383 all subsequently created keys
385 will be labeled with this context.
386 For further information, see the kernel source file
387 .I Documentation/security/keys/core.rst
389 .\" commit b68101a1e8f0263dbc7b8375d2a7c57c6216fb76
390 .I Documentation/security/keys.txt
391 on Linux between 3.0 and 4.13, or
392 .\" commit d410fa4ef99112386de5f218dd7df7b4fca910b4
393 .I Documentation/keys.txt
396 .IR /proc/[pid]/attr/prev " (since Linux 2.6.0)"
397 This file contains the security context of the process before the last
399 that is, the previous value of
400 .IR /proc/[pid]/attr/current .
402 .IR /proc/[pid]/attr/socketcreate " (since Linux 2.6.18)"
403 .\" commit 42c3e03ef6b298813557cdb997bd6db619cd65a2
404 If a process writes a security context into this file,
405 all subsequently created sockets will be labeled with this context.
407 .IR /proc/[pid]/autogroup " (since Linux 2.6.38)"
408 .\" commit 5091faa449ee0b7d73bc296a93bca9540fc51d0a
412 .IR /proc/[pid]/auxv " (since 2.6.0)"
413 .\" Precisely: Linux 2.6.0-test7
414 This contains the contents of the ELF interpreter information passed
415 to the process at exec time.
416 The format is one \fIunsigned long\fP ID
417 plus one \fIunsigned long\fP value for each entry.
418 The last entry contains two zeros.
422 Permission to access this file is governed by a ptrace access mode
423 .B PTRACE_MODE_READ_FSCREDS
427 .IR /proc/[pid]/cgroup " (since Linux 2.6.24)"
431 .IR /proc/[pid]/clear_refs " (since Linux 2.6.22)"
432 .\" commit b813e931b4c8235bb42e301096ea97dbdee3e8fe (2.6.22)
433 .\" commit 398499d5f3613c47f2143b8c54a04efb5d7a6da9 (2.6.32)
434 .\" commit 040fa02077de01c7e08fa75be6125e4ca5636011 (3.11)
436 .\" "Clears page referenced bits shown in smaps output"
437 .\" write-only, writable only by the owner of the process
439 This is a write-only file, writable only by owner of the process.
441 The following values may be written to the file:
444 1 (since Linux 2.6.22)
445 .\" Internally: CLEAR_REFS_ALL
446 Reset the PG_Referenced and ACCESSED/YOUNG
447 bits for all the pages associated with the process.
448 (Before kernel 2.6.32, writing any nonzero value to this file
451 2 (since Linux 2.6.32)
452 .\" Internally: CLEAR_REFS_ANON
453 Reset the PG_Referenced and ACCESSED/YOUNG
454 bits for all anonymous pages associated with the process.
456 3 (since Linux 2.6.32)
457 .\" Internally: CLEAR_REFS_MAPPED
458 Reset the PG_Referenced and ACCESSED/YOUNG
459 bits for all file-mapped pages associated with the process.
462 Clearing the PG_Referenced and ACCESSED/YOUNG bits provides a method
463 to measure approximately how much memory a process is using.
464 One first inspects the values in the "Referenced" fields
465 for the VMAs shown in
466 .IR /proc/[pid]/smaps
467 to get an idea of the memory footprint of the
469 One then clears the PG_Referenced and ACCESSED/YOUNG bits
470 and, after some measured time interval,
471 once again inspects the values in the "Referenced" fields
472 to get an idea of the change in memory footprint of the
473 process during the measured interval.
474 If one is interested only in inspecting the selected mapping types,
475 then the value 2 or 3 can be used instead of 1.
477 Further values can be written to affect different properties:
481 Clear the soft-dirty bit for all the pages associated with the process.
482 .\" Internally: CLEAR_REFS_SOFT_DIRTY
483 This is used (in conjunction with
484 .IR /proc/[pid]/pagemap )
485 by the check-point restore system to discover which pages of a process
486 have been dirtied since the file
487 .IR /proc/[pid]/clear_refs
491 .\" Internally: CLEAR_REFS_MM_HIWATER_RSS
492 Reset the peak resident set size ("high water mark") to the process's
493 current resident set size value.
497 .IR /proc/[pid]/clear_refs
498 other than those listed above has no effect.
501 .IR /proc/[pid]/clear_refs
502 file is present only if the
503 .B CONFIG_PROC_PAGE_MONITOR
504 kernel configuration option is enabled.
506 .I /proc/[pid]/cmdline
507 This read-only file holds the complete command line for the process,
508 unless the process is a zombie.
509 .\" In 2.3.26, this also used to be true if the process was swapped out.
510 In the latter case, there is nothing in this file:
511 that is, a read on this file will return 0 characters.
512 The command-line arguments appear in this file as a set of
513 strings separated by null bytes (\(aq\e0\(aq),
514 with a further null byte after the last string.
516 .IR /proc/[pid]/comm " (since Linux 2.6.33)"
517 .\" commit 4614a696bd1c3a9af3a08f0e5874830a85b889d4
518 This file exposes the process's
520 value\(emthat is, the command name associated with the process.
521 Different threads in the same process may have different
523 values, accessible via
524 .IR /proc/[pid]/task/[tid]/comm .
525 A thread may modify its
527 value, or that of any of other thread in the same thread group (see
532 by writing to the file
533 .IR /proc/self/task/[tid]/comm .
536 (16) characters are silently truncated.
538 This file provides a superset of the
543 operations, and is employed by
544 .BR pthread_setname_np (3)
545 when used to rename threads other than the caller.
547 .IR /proc/[pid]/coredump_filter " (since Linux 2.6.23)"
551 .IR /proc/[pid]/cpuset " (since Linux 2.6.12)"
552 .\" and/proc/[pid]/task/[tid]/cpuset
557 This is a symbolic link to the current working directory of the process.
558 To find out the current working directory of process 20,
559 for instance, you can do this:
563 .RB "$" " cd /proc/20/cwd; /bin/pwd"
569 command is often a shell built-in, and might
576 .\" The following was still true as at kernel 2.6.13
577 In a multithreaded process, the contents of this symbolic link
578 are not available if the main thread has already terminated
579 (typically by calling
580 .BR pthread_exit (3)).
582 Permission to dereference or read
584 this symbolic link is governed by a ptrace access mode
585 .B PTRACE_MODE_READ_FSCREDS
589 .I /proc/[pid]/environ
590 This file contains the initial environment that was set
591 when the currently executing program was started via
593 The entries are separated by null bytes (\(aq\e0\(aq),
594 and there may be a null byte at the end.
595 Thus, to print out the environment of process 1, you would do:
599 .RB "$" " cat /proc/1/environ | tr \(aq\e000\(aq \(aq\en\(aq"
605 the process modifies its environment
606 (e.g., by calling functions such as
613 reflect those changes.
615 Furthermore, a process may change the memory location that this file refers via
618 .BR PR_SET_MM_ENV_START .
620 Permission to access this file is governed by a ptrace access mode
621 .B PTRACE_MODE_READ_FSCREDS
626 Under Linux 2.2 and later, this file is a symbolic link
627 containing the actual pathname of the executed command.
628 This symbolic link can be dereferenced normally; attempting to open
629 it will open the executable.
632 to run another copy of the same executable that is being run by
634 If the pathname has been unlinked, the symbolic link will contain the
635 string \(aq(deleted)\(aq appended to the original pathname.
636 .\" The following was still true as at kernel 2.6.13
637 In a multithreaded process, the contents of this symbolic link
638 are not available if the main thread has already terminated
639 (typically by calling
640 .BR pthread_exit (3)).
642 Permission to dereference or read
644 this symbolic link is governed by a ptrace access mode
645 .B PTRACE_MODE_READ_FSCREDS
649 Under Linux 2.0 and earlier,
651 is a pointer to the binary which was executed,
652 and appears as a symbolic link.
655 call on this file under Linux 2.0 returns a string in the format:
659 For example, [0301]:1502 would be inode 1502 on device major 03 (IDE,
660 MFM, etc. drives) minor 01 (first partition on the first drive).
665 option can be used to locate the file.
668 This is a subdirectory containing one entry for each file which the
669 process has open, named by its file descriptor, and which is a
670 symbolic link to the actual file.
671 Thus, 0 is standard input, 1 standard output, 2 standard error, and so on.
673 For file descriptors for pipes and sockets,
674 the entries will be symbolic links whose content is the
675 file type with the inode.
678 call on this file returns a string in the format:
684 will be a socket and its inode is 2248868.
685 For sockets, that inode can be used to find more information
686 in one of the files under
689 For file descriptors that have no corresponding inode
690 (e.g., file descriptors produced by
692 .BR epoll_create (2),
694 .BR inotify_init (2),
695 .BR perf_event_open (2),
697 .BR timerfd_create (2),
699 .BR userfaultfd (2)),
700 the entry will be a symbolic link with contents of the form
702 anon_inode:<file-type>
704 In many cases (but not all), the
706 is surrounded by square brackets.
708 For example, an epoll file descriptor will have a symbolic link
709 whose content is the string
710 .IR "anon_inode:[eventpoll]" .
712 .\"The following was still true as at kernel 2.6.13
713 In a multithreaded process, the contents of this directory
714 are not available if the main thread has already terminated
715 (typically by calling
716 .BR pthread_exit (3)).
718 Programs that take a filename as a command-line argument,
719 but don't take input from standard input if no argument is supplied,
720 and programs that write to a file named as a command-line argument,
721 but don't send their output to standard output
722 if no argument is supplied, can nevertheless be made to use
723 standard input or standard output by using
725 files as command-line arguments.
726 For example, assuming that
728 is the flag designating an input file and
730 is the flag designating an output file:
734 .RB "$" " foobar \-i /proc/self/fd/0 \-o /proc/self/fd/1 ..."
738 and you have a working filter.
739 .\" The following is not true in my tests (MTK):
740 .\" Note that this will not work for
741 .\" programs that seek on their files, as the files in the fd directory
742 .\" are not seekable.
745 is approximately the same as
747 in some UNIX and UNIX-like systems.
748 Most Linux MAKEDEV scripts symbolically link
754 Most systems provide symbolic links
759 which respectively link to the files
766 Thus the example command above could be written as:
770 .RB "$" " foobar \-i /dev/stdin \-o /dev/stdout ..."
774 Permission to dereference or read
776 the symbolic links in this directory is governed by a ptrace access mode
777 .B PTRACE_MODE_READ_FSCREDS
781 Note that for file descriptors referring to inodes (pipes and sockets, see above),
782 those inodes still have permission bits and ownership information
783 distinct from those of the
786 and that the owner may differ from the user and group IDs of the process.
787 An unprivileged process may lack permissions to open them, as in this example:
791 .RB "$" " echo test | sudo \-u nobody cat"
793 .RB "$" " echo test | sudo \-u nobody cat /proc/self/fd/0"
794 cat: /proc/self/fd/0: Permission denied
798 File descriptor 0 refers to the pipe created by the shell
799 and owned by that shell's user, which is not
803 does not have permission to create a new file descriptor to read from that inode,
804 even though it can still read from its existing file descriptor 0.
806 .IR /proc/[pid]/fdinfo/ " (since Linux 2.6.22)"
807 This is a subdirectory containing one entry for each file which the
808 process has open, named by its file descriptor.
809 The files in this directory are readable only by the owner of the process.
810 The contents of each file can be read to obtain information
811 about the corresponding file descriptor.
812 The content depends on the type of file referred to by the
813 corresponding file descriptor.
815 For regular files and directories, we see something like:
819 .RB "$" " cat /proc/12015/fdinfo/4"
826 The fields are as follows:
830 This is a decimal number showing the file offset.
833 This is an octal number that displays the
834 file access mode and file status flags (see
836 If the close-on-exec file descriptor flag is set, then
838 will also include the value
842 .\" commit 1117f72ea0217ba0cc19f05adbbd8b9a397f5ab7
843 this field incorrectly displayed the setting of
845 at the time the file was opened,
846 rather than the current setting of the close-on-exec flag.
850 This field, present since Linux 3.15,
851 .\" commit 49d063cb353265c3af701bab215ac438ca7df36d
852 is the ID of the mount point containing this file.
853 See the description of
854 .IR /proc/[pid]/mountinfo .
857 For eventfd file descriptors (see
859 we see (since Linux 3.8)
860 .\" commit cbac5542d48127b546a23d816380a7926eee1c25
861 the following fields:
873 is the current value of the eventfd counter, in hexadecimal.
875 For epoll file descriptors (see
877 we see (since Linux 3.8)
878 .\" commit 138d22b58696c506799f8de759804083ff9effae
879 the following fields:
886 tfd: 9 events: 19 data: 74253d2500000009
887 tfd: 7 events: 19 data: 74253d2500000007
891 Each of the lines beginning
893 describes one of the file descriptors being monitored via
894 the epoll file descriptor (see
899 field is the number of the file descriptor.
902 field is a hexadecimal mask of the events being monitored for this file
906 field is the data value associated with this file descriptor.
908 For signalfd file descriptors (see
910 we see (since Linux 3.8)
911 .\" commit 138d22b58696c506799f8de759804083ff9effae
912 the following fields:
919 sigmask: 0000000000000006
924 is the hexadecimal mask of signals that are accepted via this
925 signalfd file descriptor.
926 (In this example, bits 2 and 3 are set, corresponding to the signals
933 For inotify file descriptors (see
935 we see (since Linux 3.8)
936 the following fields:
943 inotify wd:2 ino:7ef82a sdev:800001 mask:800afff ignored_mask:0 fhandle\-bytes:8 fhandle\-type:1 f_handle:2af87e00220ffd73
944 inotify wd:1 ino:192627 sdev:800001 mask:800afff ignored_mask:0 fhandle\-bytes:8 fhandle\-type:1 f_handle:27261900802dfd73
948 Each of the lines beginning with "inotify" displays information about
949 one file or directory that is being monitored.
950 The fields in this line are as follows:
954 A watch descriptor number (in decimal).
957 The inode number of the target file (in hexadecimal).
960 The ID of the device where the target file resides (in hexadecimal).
963 The mask of events being monitored for the target file (in hexadecimal).
966 If the kernel was built with exportfs support, the path to the target
967 file is exposed as a file handle, via three hexadecimal fields:
973 For fanotify file descriptors (see
975 we see (since Linux 3.8)
976 the following fields:
983 fanotify flags:0 event\-flags:88002
984 fanotify ino:19264f sdev:800001 mflags:0 mask:1 ignored_mask:0 fhandle\-bytes:8 fhandle\-type:1 f_handle:4f261900a82dfd73
988 The fourth line displays information defined when the fanotify group
990 .BR fanotify_init (2):
997 .BR fanotify_init (2)
998 (expressed in hexadecimal).
1004 .BR fanotify_init (2)
1005 (expressed in hexadecimal).
1008 Each additional line shown in the file contains information
1009 about one of the marks in the fanotify group.
1010 Most of these fields are as for inotify, except:
1014 The flags associated with the mark
1015 (expressed in hexadecimal).
1018 The events mask for this mark
1019 (expressed in hexadecimal).
1022 The mask of events that are ignored for this mark
1023 (expressed in hexadecimal).
1026 For details on these fields, see
1027 .BR fanotify_mark (2).
1029 For timerfd file descriptors (see
1031 we see (since Linux 3.17)
1032 .\" commit af9c4957cf212ad9cf0bee34c95cb11de5426e85
1033 the following fields:
1043 it_value: (7695568592, 640020877)
1050 This is the numeric value of the clock ID
1051 (corresponding to one of the
1053 constants defined via
1055 that is used to mark the progress of the timer (in this example, 0 is
1056 .BR CLOCK_REALTIME ).
1059 This is the number of timer expirations that have occurred,
1060 (i.e., the value that
1062 on it would return).
1065 This field lists the flags with which the timerfd was last armed (see
1066 .BR timerfd_settime (2)),
1068 (in this example, both
1069 .B TFD_TIMER_ABSTIME
1071 .B TFD_TIMER_CANCEL_ON_SET
1075 This field contains the amount of time until the timer will next expire,
1076 expressed in seconds and nanoseconds.
1077 This is always expressed as a relative value,
1078 regardless of whether the timer was created using the
1079 .B TFD_TIMER_ABSTIME
1083 This field contains the interval of the timer,
1084 in seconds and nanoseconds.
1089 fields contain the values that
1090 .BR timerfd_gettime (2)
1091 on this file descriptor would return.)
1094 .IR /proc/[pid]/gid_map " (since Linux 3.5)"
1096 .BR user_namespaces (7).
1098 .IR /proc/[pid]/io " (since kernel 2.6.20)"
1099 .\" commit 7c3ab7381e79dfc7db14a67c6f4f3285664e1ec2
1100 This file contains I/O statistics for the process, for example:
1104 .RB "#" " cat /proc/3828/io"
1110 write_bytes: 323932160
1111 cancelled_write_bytes: 0
1115 The fields are as follows:
1118 .IR rchar ": characters read"
1119 The number of bytes which this task has caused to be read from storage.
1120 This is simply the sum of bytes which this process passed to
1122 and similar system calls.
1123 It includes things such as terminal I/O and
1124 is unaffected by whether or not actual
1125 physical disk I/O was required (the read might have been satisfied from
1128 .IR wchar ": characters written"
1129 The number of bytes which this task has caused, or shall cause to be written
1131 Similar caveats apply here as with
1134 .IR syscr ": read syscalls"
1135 Attempt to count the number of read I/O operations\(emthat is,
1136 system calls such as
1141 .IR syscw ": write syscalls"
1142 Attempt to count the number of write I/O operations\(emthat is,
1143 system calls such as
1148 .IR read_bytes ": bytes read"
1149 Attempt to count the number of bytes which this process really did cause to
1150 be fetched from the storage layer.
1151 This is accurate for block-backed filesystems.
1153 .IR write_bytes ": bytes written"
1154 Attempt to count the number of bytes which this process caused to be sent to
1157 .IR cancelled_write_bytes :
1158 The big inaccuracy here is truncate.
1159 If a process writes 1MB to a file and then deletes the file,
1160 it will in fact perform no writeout.
1161 But it will have been accounted as having caused 1MB of write.
1162 In other words: this field represents the number of bytes which this process
1163 caused to not happen, by truncating pagecache.
1164 A task can cause "negative" I/O too.
1165 If this task truncates some dirty pagecache,
1166 some I/O which another task has been accounted for
1169 will not be happening.
1173 In the current implementation, things are a bit racy on 32-bit systems:
1174 if process A reads process B's
1176 while process B is updating one of these 64-bit counters,
1177 process A could see an intermediate result.
1179 Permission to access this file is governed by a ptrace access mode
1180 .B PTRACE_MODE_READ_FSCREDS
1184 .IR /proc/[pid]/limits " (since Linux 2.6.24)"
1185 This file displays the soft limit, hard limit, and units of measurement
1186 for each of the process's resource limits (see
1188 Up to and including Linux 2.6.35,
1189 this file is protected to allow reading only by the real UID of the process.
1191 .\" commit 3036e7b490bf7878c6dae952eec5fb87b1106589
1192 this file is readable by all users on the system.
1193 .\" FIXME Describe /proc/[pid]/loginuid
1194 .\" Added in 2.6.11; updating requires CAP_AUDIT_CONTROL
1195 .\" CONFIG_AUDITSYSCALL
1197 .IR /proc/[pid]/map_files/ " (since kernel 3.3)
1198 .\" commit 640708a2cff7f81e246243b0073c66e6ece7e53e
1199 This subdirectory contains entries corresponding to memory-mapped
1202 Entries are named by memory region start and end
1203 address pair (expressed as hexadecimal numbers),
1204 and are symbolic links to the mapped files themselves.
1205 Here is an example, with the output wrapped and reformatted to fit on an 80-column display:
1209 .RB "#" " ls \-l /proc/self/map_files/"
1210 lr\-\-\-\-\-\-\-\-. 1 root root 64 Apr 16 21:31
1211 3252e00000\-3252e20000 \-> /usr/lib64/ld\-2.15.so
1216 Although these entries are present for memory regions that were
1219 flag, the way anonymous shared memory (regions created with the
1220 .B MAP_ANON | MAP_SHARED
1222 is implemented in Linux
1223 means that such regions also appear on this directory.
1224 Here is an example where the target file is the deleted
1230 lrw\-\-\-\-\-\-\-. 1 root root 64 Apr 16 21:33
1231 7fc075d2f000\-7fc075e6f000 \-> /dev/zero (deleted)
1235 This directory appears only if the
1236 .B CONFIG_CHECKPOINT_RESTORE
1237 kernel configuration option is enabled.
1239 .RB ( CAP_SYS_ADMIN )
1241 .\" This may change. See the mail thread
1242 .\" "[RFC][PATCH v2] procfs: Always expose /proc/<pid>/map_files/ and make it readable"
1244 is required to view the contents of this directory.
1247 A file containing the currently mapped memory regions and their access
1251 for some further information about memory mappings.
1253 Permission to access this file is governed by a ptrace access mode
1254 .B PTRACE_MODE_READ_FSCREDS
1258 The format of the file is:
1262 .I "address perms offset dev inode pathname"
1263 00400000\-00452000 r\-xp 00000000 08:02 173521 /usr/bin/dbus\-daemon
1264 00651000\-00652000 r\-\-p 00051000 08:02 173521 /usr/bin/dbus\-daemon
1265 00652000\-00655000 rw\-p 00052000 08:02 173521 /usr/bin/dbus\-daemon
1266 00e03000\-00e24000 rw\-p 00000000 00:00 0 [heap]
1267 00e24000\-011f7000 rw\-p 00000000 00:00 0 [heap]
1269 35b1800000\-35b1820000 r\-xp 00000000 08:02 135522 /usr/lib64/ld\-2.15.so
1270 35b1a1f000\-35b1a20000 r\-\-p 0001f000 08:02 135522 /usr/lib64/ld\-2.15.so
1271 35b1a20000\-35b1a21000 rw\-p 00020000 08:02 135522 /usr/lib64/ld\-2.15.so
1272 35b1a21000\-35b1a22000 rw\-p 00000000 00:00 0
1273 35b1c00000\-35b1dac000 r\-xp 00000000 08:02 135870 /usr/lib64/libc\-2.15.so
1274 35b1dac000\-35b1fac000 \-\-\-p 001ac000 08:02 135870 /usr/lib64/libc\-2.15.so
1275 35b1fac000\-35b1fb0000 r\-\-p 001ac000 08:02 135870 /usr/lib64/libc\-2.15.so
1276 35b1fb0000\-35b1fb2000 rw\-p 001b0000 08:02 135870 /usr/lib64/libc\-2.15.so
1278 f2c6ff8c000\-7f2c7078c000 rw\-p 00000000 00:00 0 [stack:986]
1280 7fffb2c0d000\-7fffb2c2e000 rw\-p 00000000 00:00 0 [stack]
1281 7fffb2d48000\-7fffb2d49000 r\-xp 00000000 00:00 0 [vdso]
1287 field is the address space in the process that the mapping occupies.
1290 field is a set of permissions:
1298 p = private (copy on write)
1304 field is the offset into the file/whatever;
1309 is the inode on that device.
1310 0 indicates that no inode is associated with the memory region,
1311 as would be the case with BSS (uninitialized data).
1315 field will usually be the file that is backing the mapping.
1317 you can easily coordinate with the
1319 field by looking at the
1320 Offset field in the ELF program headers
1321 .RI ( "readelf\ \-l" ).
1323 There are additional helpful pseudo-paths:
1327 The initial process's (also known as the main thread's) stack.
1329 .IR [stack:<tid>] " (from Linux 3.4 to 4.4)"
1330 .\" commit b76437579d1344b612cf1851ae610c636cec7db0 (added)
1331 .\" commit 65376df582174ffcec9e6471bf5b0dd79ba05e4a (removed)
1332 A thread's stack (where the
1335 It corresponds to the
1336 .IR /proc/[pid]/task/[tid]/
1338 This field was removed in Linux 4.5, since providing this information
1339 for a process with large numbers of threads is expensive.
1342 The virtual dynamically linked shared object.
1354 this is an anonymous mapping as obtained via
1356 There is no easy way to coordinate this back to a process's source,
1357 short of running it through
1363 is shown unescaped except for newline characters, which are replaced
1364 with an octal escape sequence.
1365 As a result, it is not possible to determine whether the original
1366 pathname contained a newline character or the literal
1370 If the mapping is file-backed and the file has been deleted, the string
1371 " (deleted)" is appended to the pathname.
1372 Note that this is ambiguous too.
1374 Under Linux 2.0, there is no field giving pathname.
1377 This file can be used to access the pages of a process's memory through
1383 Permission to access this file is governed by a ptrace access mode
1384 .B PTRACE_MODE_ATTACH_FSCREDS
1388 .IR /proc/[pid]/mountinfo " (since Linux 2.6.26)"
1389 .\" This info adapted from Documentation/filesystems/proc.txt
1390 .\" commit 2d4d4864ac08caff5c204a752bd004eed4f08760
1391 This file contains information about mount points
1392 in the process's mount namespace (see
1393 .BR mount_namespaces (7)).
1394 It supplies various information
1395 (e.g., propagation state, root of mount for bind mounts,
1396 identifier for each mount and its parent) that is missing from the (older)
1397 .IR /proc/[pid]/mounts
1398 file, and fixes various other problems with that file
1399 (e.g., nonextensibility,
1400 failure to distinguish per-mount versus per-superblock options).
1402 The file contains lines of the form:
1406 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 \- ext3 /dev/root rw,errors=continue
1407 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1411 The numbers in parentheses are labels for the descriptions below:
1415 mount ID: a unique ID for the mount (may be reused after
1419 parent ID: the ID of the parent mount
1420 (or of self for the root of this mount namespace's mount tree).
1422 If a new mount is stacked on top of a previous existing mount
1423 (so that it hides the existing mount) at pathname P,
1424 then the parent of the new mount is the previous mount at that location.
1425 Thus, when looking at all the mounts stacked at a particular location,
1426 the top-most mount is the one that is not the parent
1427 of any other mount at the same location.
1428 (Note, however, that this top-most mount will be accessible only if
1429 the longest path subprefix of P that is a mount point
1430 is not itself hidden by a stacked mount.)
1432 If the parent mount point lies outside the process's root directory (see
1434 the ID shown here won't have a corresponding record in
1436 whose mount ID (field 1) matches this parent mount ID
1437 (because mount points that lie outside the process's root directory
1440 As a special case of this point,
1441 the process's root mount point may have a parent mount
1442 (for the initramfs filesystem) that lies
1443 .\" Miklos Szeredi, Nov 2017: The hidden one is the initramfs, I believe
1444 .\" mtk: In the initial mount namespace, this hidden ID has the value 0
1445 outside the process's root directory,
1446 and an entry for that mount point will not appear in
1450 major:minor: the value of
1452 for files on this filesystem (see
1456 root: the pathname of the directory in the filesystem
1457 which forms the root of this mount.
1460 mount point: the pathname of the mount point relative
1461 to the process's root directory.
1464 mount options: per-mount options (see
1468 optional fields: zero or more fields of the form "tag[:value]"; see below.
1471 separator: the end of the optional fields is marked by a single hyphen.
1474 filesystem type: the filesystem type in the form "type[.subtype]".
1477 mount source: filesystem-specific information or "none".
1480 super options: per-superblock options (see
1484 Currently, the possible optional fields are
1487 .IR propagate_from ,
1491 .BR mount_namespaces (7)
1492 for a description of these fields.
1493 Parsers should ignore all unrecognized optional fields.
1495 For more information on mount propagation see:
1496 .I Documentation/filesystems/sharedsubtree.txt
1497 in the Linux kernel source tree.
1499 .IR /proc/[pid]/mounts " (since Linux 2.4.19)"
1500 This file lists all the filesystems currently mounted in the
1501 process's mount namespace (see
1502 .BR mount_namespaces (7)).
1503 The format of this file is documented in
1506 Since kernel version 2.6.15, this file is pollable:
1507 after opening the file for reading, a change in this file
1508 (i.e., a filesystem mount or unmount) causes
1510 to mark the file descriptor as having an exceptional condition, and
1514 mark the file as having a priority event
1516 (Before Linux 2.6.30,
1517 a change in this file was indicated by the file descriptor
1518 being marked as readable for
1520 and being marked as having an error condition for
1523 .BR epoll_wait (2).)
1525 .IR /proc/[pid]/mountstats " (since Linux 2.6.17)"
1526 This file exports information (statistics, configuration information)
1527 about the mount points in the process's mount namespace (see
1528 .BR mount_namespaces (7)).
1529 Lines in this file have the form:
1533 device /dev/sda7 mounted on /home with fstype ext3 [statistics]
1534 ( 1 ) ( 2 ) (3 ) (4)
1538 The fields in each line are:
1542 The name of the mounted device
1543 (or "nodevice" if there is no corresponding device).
1546 The mount point within the filesystem tree.
1549 The filesystem type.
1552 Optional statistics and configuration information.
1553 Currently (as at Linux 2.6.26), only NFS filesystems export
1554 information via this field.
1557 This file is readable only by the owner of the process.
1559 .IR /proc/[pid]/net " (since Linux 2.6.25)"
1560 See the description of
1563 .IR /proc/[pid]/ns/ " (since Linux 3.0)"
1564 .\" See commit 6b4e306aa3dc94a0545eb9279475b1ab6209a31f
1565 This is a subdirectory containing one entry for each namespace that
1566 supports being manipulated by
1568 For more information, see
1571 .IR /proc/[pid]/numa_maps " (since Linux 2.6.14)"
1575 .IR /proc/[pid]/oom_adj " (since Linux 2.6.11)"
1576 This file can be used to adjust the score used to select which process
1577 should be killed in an out-of-memory (OOM) situation.
1578 The kernel uses this value for a bit-shift operation of the process's
1581 valid values are in the range \-16 to +15,
1582 plus the special value \-17,
1583 which disables OOM-killing altogether for this process.
1584 A positive score increases the likelihood of this
1585 process being killed by the OOM-killer;
1586 a negative score decreases the likelihood.
1588 The default value for this file is 0;
1589 a new process inherits its parent's
1592 A process must be privileged
1593 .RB ( CAP_SYS_RESOURCE )
1594 to update this file.
1596 Since Linux 2.6.36, use of this file is deprecated in favor of
1597 .IR /proc/[pid]/oom_score_adj .
1599 .IR /proc/[pid]/oom_score " (since Linux 2.6.11)"
1600 .\" See mm/oom_kill.c::badness() in pre 2.6.36 sources
1601 .\" See mm/oom_kill.c::oom_badness() after 2.6.36
1602 .\" commit a63d83f427fbce97a6cea0db2e64b0eb8435cd10
1603 This file displays the current score that the kernel gives to
1604 this process for the purpose of selecting a process
1606 A higher score means that the process is more likely to be
1607 selected by the OOM-killer.
1608 The basis for this score is the amount of memory used by the process,
1609 with increases (+) or decreases (\-) for factors including:
1610 .\" See mm/oom_kill.c::badness() in pre 2.6.36 sources
1611 .\" See mm/oom_kill.c::oom_badness() after 2.6.36
1612 .\" commit a63d83f427fbce97a6cea0db2e64b0eb8435cd10
1615 whether the process is privileged (\-).
1616 .\" More precisely, if it has CAP_SYS_ADMIN or (pre 2.6.36) CAP_SYS_RESOURCE
1619 Before kernel 2.6.36 the following factors were also used in the calculation of oom_score:
1622 whether the process creates a lot of children using
1626 whether the process has been running a long time,
1627 or has used a lot of CPU time (\-);
1629 whether the process has a low nice value (i.e., > 0) (+); and
1631 whether the process is making direct hardware access (\-).
1632 .\" More precisely, if it has CAP_SYS_RAWIO
1637 also reflects the adjustment specified by the
1641 setting for the process.
1643 .IR /proc/[pid]/oom_score_adj " (since Linux 2.6.36)"
1644 .\" Text taken from 3.7 Documentation/filesystems/proc.txt
1645 This file can be used to adjust the badness heuristic used to select which
1646 process gets killed in out-of-memory conditions.
1648 The badness heuristic assigns a value to each candidate task ranging from 0
1649 (never kill) to 1000 (always kill) to determine which process is targeted.
1650 The units are roughly a proportion along that range of
1651 allowed memory the process may allocate from,
1652 based on an estimation of its current memory and swap use.
1653 For example, if a task is using all allowed memory,
1654 its badness score will be 1000.
1655 If it is using half of its allowed memory, its score will be 500.
1657 There is an additional factor included in the badness score: root
1658 processes are given 3% extra memory over other tasks.
1660 The amount of "allowed" memory depends on the context
1661 in which the OOM-killer was called.
1662 If it is due to the memory assigned to the allocating task's cpuset
1664 the allowed memory represents the set of mems assigned to that
1667 If it is due to a mempolicy's node(s) being exhausted,
1668 the allowed memory represents the set of mempolicy nodes.
1669 If it is due to a memory limit (or swap limit) being reached,
1670 the allowed memory is that configured limit.
1671 Finally, if it is due to the entire system being out of memory, the
1672 allowed memory represents all allocatable resources.
1676 is added to the badness score before it
1677 is used to determine which task to kill.
1678 Acceptable values range from \-1000
1679 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX).
1680 This allows user space to control the preference for OOM-killing,
1681 ranging from always preferring a certain
1682 task or completely disabling it from OOM killing.
1683 The lowest possible value, \-1000, is
1684 equivalent to disabling OOM-killing entirely for that task,
1685 since it will always report a badness score of 0.
1687 Consequently, it is very simple for user space to define
1688 the amount of memory to consider for each task.
1691 value of +500, for example,
1692 is roughly equivalent to allowing the remainder of tasks sharing the
1693 same system, cpuset, mempolicy, or memory controller resources
1694 to use at least 50% more memory.
1695 A value of \-500, on the other hand, would be roughly
1696 equivalent to discounting 50% of the task's
1697 allowed memory from being considered as scoring against the task.
1699 For backward compatibility with previous kernels,
1700 .I /proc/[pid]/oom_adj
1701 can still be used to tune the badness score.
1703 scaled linearly with
1707 .IR /proc/[pid]/oom_score_adj
1709 .IR /proc/[pid]/oom_adj
1710 will change the other with its scaled value.
1714 program provides a command-line interface for adjusting the
1716 value of a running process or a newly executed command.
1718 .IR /proc/[pid]/pagemap " (since Linux 2.6.25)"
1719 This file shows the mapping of each of the process's virtual pages
1720 into physical page frames or swap area.
1721 It contains one 64-bit value for each virtual page,
1722 with the bits set as follows:
1726 If set, the page is present in RAM.
1729 If set, the page is in swap space
1731 61 (since Linux 3.5)
1732 The page is a file-mapped page or a shared anonymous page.
1734 60\(en57 (since Linux 3.11)
1736 .\" Not quite true; see commit 541c237c0923f567c9c4cabb8a81635baadc713f
1738 56 (since Linux 4.2)
1739 .\" commit 77bb499bb60f4b79cca7d139c8041662860fcf87
1740 .\" commit 83b4b0bb635eee2b8e075062e4e008d1bc110ed7
1741 The page is exclusively mapped.
1743 55 (since Linux 3.11)
1745 (see the kernel source file
1746 .IR Documentation/admin\-guide/mm/soft\-dirty.rst ).
1749 If the page is present in RAM (bit 63), then these bits
1750 provide the page frame number, which can be used to index
1751 .IR /proc/kpageflags
1753 .IR /proc/kpagecount .
1754 If the page is present in swap (bit 62),
1755 then bits 4\(en0 give the swap type, and bits 54\(en5 encode the swap offset.
1758 Before Linux 3.11, bits 60\(en55 were
1759 used to encode the base-2 log of the page size.
1762 .IR /proc/[pid]/pagemap
1764 .IR /proc/[pid]/maps
1765 to determine which areas of memory are actually mapped and seek
1766 to skip over unmapped regions.
1769 .IR /proc/[pid]/pagemap
1770 file is present only if the
1771 .B CONFIG_PROC_PAGE_MONITOR
1772 kernel configuration option is enabled.
1774 Permission to access this file is governed by a ptrace access mode
1775 .B PTRACE_MODE_READ_FSCREDS
1779 .IR /proc/[pid]/personality " (since Linux 2.6.28)"
1780 .\" commit 478307230810d7e2a753ed220db9066dfdf88718
1781 This read-only file exposes the process's execution domain, as set by
1782 .BR personality (2).
1783 The value is displayed in hexadecimal notation.
1785 Permission to access this file is governed by a ptrace access mode
1786 .B PTRACE_MODE_ATTACH_FSCREDS
1791 UNIX and Linux support the idea of a per-process root of the
1792 filesystem, set by the
1795 This file is a symbolic link that points to the process's
1796 root directory, and behaves in the same way as
1801 Note however that this file is not merely a symbolic link.
1802 It provides the same view of the filesystem (including namespaces and the
1803 set of per-process mounts) as the process itself.
1804 An example illustrates this point.
1805 In one terminal, we start a shell in new user and mount namespaces,
1806 and in that shell we create some new mount points:
1810 $ \fBPS1='sh1# ' unshare \-Urnm\fP
1811 sh1# \fBmount \-t tmpfs tmpfs /etc\fP # Mount empty tmpfs at /etc
1812 sh1# \fBmount \-\-bind /usr /dev\fP # Mount /usr at /dev
1818 In a second terminal window, in the initial mount namespace,
1819 we look at the contents of the corresponding mounts in
1820 the initial and new namespaces:
1824 $ \fBPS1='sh2# ' sudo sh\fP
1825 sh2# \fBls /etc | wc \-l\fP # In initial NS
1827 sh2# \fBls /proc/27123/root/etc | wc \-l\fP # /etc in other NS
1828 0 # The empty tmpfs dir
1829 sh2# \fBls /dev | wc \-l\fP # In initial NS
1831 sh2# \fBls /proc/27123/root/dev | wc \-l\fP # /dev in other NS
1834 sh2# \fBls /usr | wc \-l\fP # /usr in initial NS
1839 .\" The following was still true as at kernel 2.6.13
1840 In a multithreaded process, the contents of the
1842 symbolic link are not available if the main thread has already terminated
1843 (typically by calling
1844 .BR pthread_exit (3)).
1846 Permission to dereference or read
1848 this symbolic link is governed by a ptrace access mode
1849 .B PTRACE_MODE_READ_FSCREDS
1852 .\" FIXME Describe /proc/[pid]/projid_map
1854 .\" commit f76d207a66c3a53defea67e7d36c3eb1b7d6d61d
1856 .IR /proc/[pid]/seccomp " (Linux 2.6.12 to 2.6.22)"
1857 This file can be used to read and change the process's
1858 secure computing (seccomp) mode setting.
1859 It contains the value 0 if the process is not in seccomp mode,
1860 and 1 if the process is in strict seccomp mode (see
1862 Writing 1 to this file places the process irreversibly in strict seccomp mode.
1863 (Further attempts to write to the file fail with the
1868 this file went away, to be replaced by the
1873 operations (and later by
1878 .IR /proc/[pid]/status ).
1879 .\" FIXME Describe /proc/[pid]/sessionid
1880 .\" commit 1e0bd7550ea9cf474b1ad4c6ff5729a507f75fdc
1881 .\" CONFIG_AUDITSYSCALL
1882 .\" Added in 2.6.25; read-only; only readable by real UID
1884 .\" FIXME Describe /proc/[pid]/sched
1886 .\" CONFIG_SCHED_DEBUG, and additional fields if CONFIG_SCHEDSTATS
1887 .\" Displays various scheduling parameters
1888 .\" This file can be written, to reset stats
1889 .\" The set of fields exposed by this file have changed
1890 .\" significantly over time.
1891 .\" commit 43ae34cb4cd650d1eb4460a8253a8e747ba052ac
1893 .\" FIXME Describe /proc/[pid]/schedstats and
1894 .\" /proc/[pid]/task/[tid]/schedstats
1896 .\" CONFIG_SCHEDSTATS
1898 .IR /proc/[pid]/setgroups " (since Linux 3.19)"
1900 .BR user_namespaces (7).
1902 .IR /proc/[pid]/smaps " (since Linux 2.6.14)"
1903 This file shows memory consumption for each of the process's mappings.
1906 command displays similar information,
1907 in a form that may be easier for parsing.)
1908 For each mapping there is a series of lines such as the following:
1912 00400000\-0048a000 r\-xp 00000000 fd:03 960637 /bin/bash
1916 Shared_Clean: 452 kB
1923 ShmemHugePages: 0 kB
1924 ShmemPmdMapped: 0 kB
1926 KernelPageSize: 4 kB
1928 KernelPageSize: 4 kB
1932 VmFlags: rd ex mr mw me dw
1936 The first of these lines shows the same information as is displayed
1938 .IR /proc/[pid]/maps .
1939 The following lines show the size of the mapping,
1940 the amount of the mapping that is currently resident in RAM ("Rss"),
1941 the process's proportional share of this mapping ("Pss"),
1942 the number of clean and dirty shared pages in the mapping,
1943 and the number of clean and dirty private pages in the mapping.
1944 "Referenced" indicates the amount of memory currently marked as
1945 referenced or accessed.
1946 "Anonymous" shows the amount of memory
1947 that does not belong to any file.
1948 "Swap" shows how much
1949 would-be-anonymous memory is also used, but out on swap.
1951 The "KernelPageSize" line (available since Linux 2.6.29)
1952 is the page size used by the kernel to back the virtual memory area.
1953 This matches the size used by the MMU in the majority of cases.
1954 However, one counter-example occurs on PPC64 kernels
1955 whereby a kernel using 64kB as a base page size may still use 4kB
1956 pages for the MMU on older processors.
1957 To distinguish the two attributes, the "MMUPageSize" line
1958 (also available since Linux 2.6.29)
1959 reports the page size used by the MMU.
1961 The "Locked" indicates whether the mapping is locked in memory
1964 The "ProtectionKey" line (available since Linux 4.9, on x86 only)
1965 contains the memory protection key (see
1967 associated with the virtual memory area.
1968 This entry is present only if the kernel was built with the
1969 .B CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS
1970 configuration option.
1972 The "VmFlags" line (available since Linux 3.8)
1973 represents the kernel flags associated with the virtual memory area,
1974 encoded using the following two-letter codes:
1984 gd - stack segment grows down
1986 dw - disabled write to the mapped file
1987 lo - pages are locked in memory
1988 io - memory mapped I/O area
1989 sr - sequential read advise provided
1990 rr - random read advise provided
1991 dc - do not copy area on fork
1992 de - do not expand area on remapping
1993 ac - area is accountable
1994 nr - swap space is not reserved for the area
1995 ht - area uses huge tlb pages
1996 nl - non-linear mapping
1997 ar - architecture specific flag
1998 dd - do not include area into core dump
1999 sd - soft-dirty flag
2001 hg - huge page advise flag
2002 nh - no-huge page advise flag
2003 mg - mergeable advise flag
2005 "ProtectionKey" field contains the memory protection key (see
2007 associated with the virtual memory area.
2008 Present only if the kernel was built with the
2009 .B CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS
2010 configuration option. (since Linux 4.6)
2013 .IR /proc/[pid]/smaps
2014 file is present only if the
2015 .B CONFIG_PROC_PAGE_MONITOR
2016 kernel configuration option is enabled.
2018 .IR /proc/[pid]/stack " (since Linux 2.6.29)"
2019 .\" 2ec220e27f5040aec1e88901c1b6ea3d135787ad
2020 This file provides a symbolic trace of the function calls in this
2021 process's kernel stack.
2022 This file is provided only if the kernel was built with the
2023 .B CONFIG_STACKTRACE
2024 configuration option.
2026 Permission to access this file is governed by a ptrace access mode
2027 .B PTRACE_MODE_ATTACH_FSCREDS
2032 Status information about the process.
2035 It is defined in the kernel source file
2036 .IR fs/proc/array.c "."
2038 The fields, in order, with their proper
2040 format specifiers, are listed below.
2041 Whether or not certain of these fields display valid information is governed by
2042 a ptrace access mode
2043 .BR PTRACE_MODE_READ_FSCREDS "\ |\ " PTRACE_MODE_NOAUDIT
2046 If the check denies access, then the field value is displayed as 0.
2047 The affected fields are indicated with the marking [PT].
2056 The filename of the executable, in parentheses.
2057 This is visible whether or not the executable is swapped out.
2059 (3) \fIstate\fP \ %c
2060 One of the following characters, indicating process state:
2065 Sleeping in an interruptible wait
2067 Waiting in uninterruptible
2072 Stopped (on a signal) or (before Linux 2.6.33) trace stopped
2074 .\" commit 44d90df6b757c59651ddd55f1a84f28132b50d29
2075 Tracing stop (Linux 2.6.33 onward)
2077 Paging (only before Linux 2.6.0)
2079 Dead (from Linux 2.6.0 onward)
2081 .\" commit 44d90df6b757c59651ddd55f1a84f28132b50d29
2082 Dead (Linux 2.6.33 to
2083 .\" commit 74e37200de8e9c4e09b70c21c3f13c2071e77457
2086 .\" commit 44d90df6b757c59651ddd55f1a84f28132b50d29
2087 Wakekill (Linux 2.6.33 to
2088 .\" commit 74e37200de8e9c4e09b70c21c3f13c2071e77457
2091 .\" commit 44d90df6b757c59651ddd55f1a84f28132b50d29
2092 Waking (Linux 2.6.33 to
2093 .\" commit 74e37200de8e9c4e09b70c21c3f13c2071e77457
2096 .\" commit f2530dc71cf0822f90bb63ea4600caaef33a66bb
2097 Parked (Linux 3.9 to
2098 .\" commit 74e37200de8e9c4e09b70c21c3f13c2071e77457
2103 The PID of the parent of this process.
2106 The process group ID of the process.
2108 (6) \fIsession\fP \ %d
2109 The session ID of the process.
2111 (7) \fItty_nr\fP \ %d
2112 The controlling terminal of the process.
2113 (The minor device number is contained in the combination of bits
2114 31 to 20 and 7 to 0;
2115 the major device number is in bits 15 to 8.)
2117 (8) \fItpgid\fP \ %d
2118 .\" This field and following, up to and including wchan added 0.99.1
2119 The ID of the foreground process group of the controlling
2120 terminal of the process.
2122 (9) \fIflags\fP \ %u
2123 The kernel flags word of the process.
2125 see the PF_* defines in the Linux kernel source file
2126 .IR include/linux/sched.h .
2127 Details depend on the kernel version.
2129 The format for this field was %lu before Linux 2.6.
2131 (10) \fIminflt\fP \ %lu
2132 The number of minor faults the process has made which have not
2133 required loading a memory page from disk.
2135 (11) \fIcminflt\fP \ %lu
2136 The number of minor faults that the process's
2137 waited-for children have made.
2139 (12) \fImajflt\fP \ %lu
2140 The number of major faults the process has made which have
2141 required loading a memory page from disk.
2143 (13) \fIcmajflt\fP \ %lu
2144 The number of major faults that the process's
2145 waited-for children have made.
2147 (14) \fIutime\fP \ %lu
2148 Amount of time that this process has been scheduled in user mode,
2149 measured in clock ticks (divide by
2150 .IR sysconf(_SC_CLK_TCK) ).
2151 This includes guest time, \fIguest_time\fP
2152 (time spent running a virtual CPU, see below),
2153 so that applications that are not aware of the guest time field
2154 do not lose that time from their calculations.
2156 (15) \fIstime\fP \ %lu
2157 Amount of time that this process has been scheduled in kernel mode,
2158 measured in clock ticks (divide by
2159 .IR sysconf(_SC_CLK_TCK) ).
2161 (16) \fIcutime\fP \ %ld
2162 Amount of time that this process's
2163 waited-for children have been scheduled in user mode,
2164 measured in clock ticks (divide by
2165 .IR sysconf(_SC_CLK_TCK) ).
2168 This includes guest time, \fIcguest_time\fP
2169 (time spent running a virtual CPU, see below).
2171 (17) \fIcstime\fP \ %ld
2172 Amount of time that this process's
2173 waited-for children have been scheduled in kernel mode,
2174 measured in clock ticks (divide by
2175 .IR sysconf(_SC_CLK_TCK) ).
2177 (18) \fIpriority\fP \ %ld
2178 (Explanation for Linux 2.6)
2179 For processes running a real-time scheduling policy
2182 .BR sched_setscheduler (2)),
2183 this is the negated scheduling priority, minus one;
2184 that is, a number in the range \-2 to \-100,
2185 corresponding to real-time priorities 1 to 99.
2186 For processes running under a non-real-time scheduling policy,
2187 this is the raw nice value
2188 .RB ( setpriority (2))
2189 as represented in the kernel.
2190 The kernel stores nice values as numbers
2191 in the range 0 (high) to 39 (low),
2192 corresponding to the user-visible nice range of \-20 to 19.
2194 Before Linux 2.6, this was a scaled value based on
2195 the scheduler weighting given to this process.
2196 .\" And back in kernel 1.2 days things were different again.
2198 (19) \fInice\fP \ %ld
2200 .BR setpriority (2)),
2201 a value in the range 19 (low priority) to \-20 (high priority).
2202 .\" Back in kernel 1.2 days things were different.
2204 .\" \fIcounter\fP %ld
2205 .\" The current maximum size in jiffies of the process's next timeslice,
2206 .\" or what is currently left of its current timeslice, if it is the
2207 .\" currently running process.
2209 .\" \fItimeout\fP %u
2210 .\" The time in jiffies of the process's next timeout.
2211 .\" timeout was removed sometime around 2.1/2.2
2213 (20) \fInum_threads\fP \ %ld
2214 Number of threads in this process (since Linux 2.6).
2215 Before kernel 2.6, this field was hard coded to 0 as a placeholder
2216 for an earlier removed field.
2218 (21) \fIitrealvalue\fP \ %ld
2219 The time in jiffies before the next
2221 is sent to the process due to an interval timer.
2222 Since kernel 2.6.17, this field is no longer maintained,
2223 and is hard coded as 0.
2225 (22) \fIstarttime\fP \ %llu
2226 The time the process started after system boot.
2227 In kernels before Linux 2.6, this value was expressed in jiffies.
2228 Since Linux 2.6, the value is expressed in clock ticks (divide by
2229 .IR sysconf(_SC_CLK_TCK) ).
2231 The format for this field was %lu before Linux 2.6.
2233 (23) \fIvsize\fP \ %lu
2234 Virtual memory size in bytes.
2236 (24) \fIrss\fP \ %ld
2237 Resident Set Size: number of pages the process has in real memory.
2238 This is just the pages which
2239 count toward text, data, or stack space.
2240 This does not include pages
2241 which have not been demand-loaded in, or which are swapped out.
2243 (25) \fIrsslim\fP \ %lu
2244 Current soft limit in bytes on the rss of the process;
2245 see the description of
2250 (26) \fIstartcode\fP \ %lu \ [PT]
2251 The address above which program text can run.
2253 (27) \fIendcode\fP \ %lu \ [PT]
2254 The address below which program text can run.
2256 (28) \fIstartstack\fP \ %lu \ [PT]
2257 The address of the start (i.e., bottom) of the stack.
2259 (29) \fIkstkesp\fP \ %lu \ [PT]
2260 The current value of ESP (stack pointer), as found in the
2261 kernel stack page for the process.
2263 (30) \fIkstkeip\fP \ %lu \ [PT]
2264 The current EIP (instruction pointer).
2266 (31) \fIsignal\fP \ %lu
2267 The bitmap of pending signals, displayed as a decimal number.
2268 Obsolete, because it does not provide information on real-time signals; use
2269 .I /proc/[pid]/status
2272 (32) \fIblocked\fP \ %lu
2273 The bitmap of blocked signals, displayed as a decimal number.
2274 Obsolete, because it does not provide information on real-time signals; use
2275 .I /proc/[pid]/status
2278 (33) \fIsigignore\fP \ %lu
2279 The bitmap of ignored signals, displayed as a decimal number.
2280 Obsolete, because it does not provide information on real-time signals; use
2281 .I /proc/[pid]/status
2284 (34) \fIsigcatch\fP \ %lu
2285 The bitmap of caught signals, displayed as a decimal number.
2286 Obsolete, because it does not provide information on real-time signals; use
2287 .I /proc/[pid]/status
2290 (35) \fIwchan\fP \ %lu \ [PT]
2291 This is the "channel" in which the process is waiting.
2292 It is the address of a location in the kernel where the process is sleeping.
2293 The corresponding symbolic name can be found in
2294 .IR /proc/[pid]/wchan .
2296 (36) \fInswap\fP \ %lu
2297 .\" nswap was added in 2.0
2298 Number of pages swapped (not maintained).
2300 (37) \fIcnswap\fP \ %lu
2301 .\" cnswap was added in 2.0
2302 Cumulative \fInswap\fP for child processes (not maintained).
2304 (38) \fIexit_signal\fP \ %d \ (since Linux 2.1.22)
2305 Signal to be sent to parent when we die.
2307 (39) \fIprocessor\fP \ %d \ (since Linux 2.2.8)
2308 CPU number last executed on.
2310 (40) \fIrt_priority\fP \ %u \ (since Linux 2.5.19)
2311 Real-time scheduling priority, a number in the range 1 to 99 for
2312 processes scheduled under a real-time policy,
2313 or 0, for non-real-time processes (see
2314 .BR sched_setscheduler (2)).
2316 (41) \fIpolicy\fP \ %u \ (since Linux 2.5.19)
2317 Scheduling policy (see
2318 .BR sched_setscheduler (2)).
2319 Decode using the SCHED_* constants in
2322 The format for this field was %lu before Linux 2.6.22.
2324 (42) \fIdelayacct_blkio_ticks\fP \ %llu \ (since Linux 2.6.18)
2325 Aggregated block I/O delays, measured in clock ticks (centiseconds).
2327 (43) \fIguest_time\fP \ %lu \ (since Linux 2.6.24)
2328 Guest time of the process (time spent running a virtual CPU
2329 for a guest operating system), measured in clock ticks (divide by
2330 .IR sysconf(_SC_CLK_TCK) ).
2332 (44) \fIcguest_time\fP \ %ld \ (since Linux 2.6.24)
2333 Guest time of the process's children, measured in clock ticks (divide by
2334 .IR sysconf(_SC_CLK_TCK) ).
2336 (45) \fIstart_data\fP \ %lu \ (since Linux 3.3) \ [PT]
2337 .\" commit b3f7f573a20081910e34e99cbc91831f4f02f1ff
2338 Address above which program initialized and
2339 uninitialized (BSS) data are placed.
2341 (46) \fIend_data\fP \ %lu \ (since Linux 3.3) \ [PT]
2342 .\" commit b3f7f573a20081910e34e99cbc91831f4f02f1ff
2343 Address below which program initialized and
2344 uninitialized (BSS) data are placed.
2346 (47) \fIstart_brk\fP \ %lu \ (since Linux 3.3) \ [PT]
2347 .\" commit b3f7f573a20081910e34e99cbc91831f4f02f1ff
2348 Address above which program heap can be expanded with
2351 (48) \fIarg_start\fP \ %lu \ (since Linux 3.5) \ [PT]
2352 .\" commit 5b172087f99189416d5f47fd7ab5e6fb762a9ba3
2353 Address above which program command-line arguments
2357 (49) \fIarg_end\fP \ %lu \ (since Linux 3.5) \ [PT]
2358 .\" commit 5b172087f99189416d5f47fd7ab5e6fb762a9ba3
2359 Address below program command-line arguments
2363 (50) \fIenv_start\fP \ %lu \ (since Linux 3.5) \ [PT]
2364 .\" commit 5b172087f99189416d5f47fd7ab5e6fb762a9ba3
2365 Address above which program environment is placed.
2367 (51) \fIenv_end\fP \ %lu \ (since Linux 3.5) \ [PT]
2368 .\" commit 5b172087f99189416d5f47fd7ab5e6fb762a9ba3
2369 Address below which program environment is placed.
2371 (52) \fIexit_code\fP \ %d \ (since Linux 3.5) \ [PT]
2372 .\" commit 5b172087f99189416d5f47fd7ab5e6fb762a9ba3
2373 The thread's exit status in the form reported by
2377 .I /proc/[pid]/statm
2378 Provides information about memory usage, measured in pages.
2383 size (1) total program size
2384 (same as VmSize in \fI/proc/[pid]/status\fP)
2385 resident (2) resident set size
2386 (same as VmRSS in \fI/proc/[pid]/status\fP)
2387 shared (3) number of resident shared pages (i.e., backed by a file)
2388 (same as RssFile+RssShmem in \fI/proc/[pid]/status\fP)
2389 text (4) text (code)
2390 .\" (not including libs; broken, includes data segment)
2391 lib (5) library (unused since Linux 2.6; always 0)
2392 data (6) data + stack
2393 .\" (including libs; broken, includes library text)
2394 dt (7) dirty pages (unused since Linux 2.6; always 0)
2398 .I /proc/[pid]/status
2399 Provides much of the information in
2402 .I /proc/[pid]/statm
2403 in a format that's easier for humans to parse.
2408 .RB "$" " cat /proc/$$/status"
2417 Uid: 1000 1000 1000 1000
2418 Gid: 100 100 100 100
2441 HugetlbPages: 0 kB # 4.4
2442 CoreDumping: 0 # 4.15
2445 SigPnd: 0000000000000000
2446 ShdPnd: 0000000000000000
2447 SigBlk: 0000000000010000
2448 SigIgn: 0000000000384004
2449 SigCgt: 000000004b813efb
2450 CapInh: 0000000000000000
2451 CapPrm: 0000000000000000
2452 CapEff: 0000000000000000
2453 CapBnd: ffffffffffffffff
2454 CapAmb: 0000000000000000
2457 Speculation_Store_Bypass: vulnerable
2458 Cpus_allowed: 00000001
2459 Cpus_allowed_list: 0
2461 Mems_allowed_list: 0
2462 voluntary_ctxt_switches: 150
2463 nonvoluntary_ctxt_switches: 545
2467 The fields are as follows:
2471 Command run by this process.
2474 Process umask, expressed in octal with a leading zero; see
2479 Current state of the process.
2491 Thread group ID (i.e., Process ID).
2494 NUMA group ID (0 if none; since Linux 3.13).
2501 PID of parent process.
2504 PID of process tracing this process (0 if not being traced).
2507 Real, effective, saved set, and filesystem UIDs (GIDs).
2510 Number of file descriptor slots currently allocated.
2513 Supplementary group list.
2516 Thread group ID (i.e., PID) in each of the PID namespaces of which
2519 The leftmost entry shows the value with respect to the PID namespace
2520 of the process that mounted this procfs (or the root namespace
2521 if mounted by the kernel),
2522 followed by the value in successively nested inner namespaces.
2523 .\" commit e4bc33245124db69b74a6d853ac76c2976f472d5
2527 Thread ID in each of the PID namespaces of which
2530 The fields are ordered as for
2535 Process group ID in each of the PID namespaces of which
2538 The fields are ordered as for
2543 descendant namespace session ID hierarchy
2544 Session ID in each of the PID namespaces of which
2547 The fields are ordered as for
2552 Peak virtual memory size.
2555 Virtual memory size.
2558 Locked memory size (see
2563 .\" commit bc3e53f682d93df677dbd5006a404722b3adfe18
2565 These are pages that can't be moved because something needs to
2566 directly access physical memory.
2569 Peak resident set size ("high water mark").
2573 Note that the value here is the sum of
2580 Size of resident anonymous memory.
2581 .\" commit bf9683d6990589390b5178dafe8fd06808869293
2585 Size of resident file mappings.
2586 .\" commit bf9683d6990589390b5178dafe8fd06808869293
2590 Size of resident shared memory (includes System V shared memory,
2593 and shared anonymous mappings).
2594 .\" commit bf9683d6990589390b5178dafe8fd06808869293
2597 .IR VmData ", " VmStk ", " VmExe :
2598 Size of data, stack, and text segments.
2601 Shared library code size.
2604 Page table entries size (since Linux 2.6.10).
2607 .\" commit dc6c9a35b66b520cf67e05d8ca60ebecad3b0479
2608 Size of second-level page tables (added in Linux 4.0; removed in Linux 4.15).
2611 .\" commit b084d4353ff99d824d3bc5a5c2c22c70b1fba722
2612 Swapped-out virtual memory size by anonymous private pages;
2613 shmem swap usage is not included (since Linux 2.6.34).
2616 Size of hugetlb memory portions
2617 .\" commit 5d317b2b6536592a9b51fe65faed43d65ca9158e
2621 Contains the value 1 if the process is currently dumping core,
2623 .\" commit c643401218be0f4ab3522e0c0a63016596d6e9ca
2625 This information can be used by a monitoring process to avoid killing
2626 a process that is currently dumping core,
2627 which could result in a corrupted core dump file.
2630 Number of threads in process containing this thread.
2633 This field contains two slash-separated numbers that relate to
2634 queued signals for the real user ID of this process.
2635 The first of these is the number of currently queued
2636 signals for this real user ID, and the second is the
2637 resource limit on the number of queued signals for this process
2638 (see the description of
2639 .BR RLIMIT_SIGPENDING
2643 .IR SigPnd ", " ShdPnd :
2644 Mask (expressed in hexadecimal)
2645 of signals pending for thread and for process as a whole (see
2650 .IR SigBlk ", " SigIgn ", " SigCgt :
2651 Masks (expressed in hexadecimal)
2652 indicating signals being blocked, ignored, and caught (see
2655 .IR CapInh ", " CapPrm ", " CapEff :
2656 Masks (expressed in hexadecimal)
2657 of capabilities enabled in inheritable, permitted, and effective sets
2659 .BR capabilities (7)).
2662 Capability bounding set, expressed in hexadecimal
2663 (since Linux 2.6.26, see
2664 .BR capabilities (7)).
2667 Ambient capability set, expressed in hexadecimal
2668 (since Linux 4.3, see
2669 .BR capabilities (7)).
2672 .\" commit af884cd4a5ae62fcf5e321fecf0ec1014730353d
2676 (since Linux 4.10, see
2680 .\" commit 2f4b3bf6b2318cfaa177ec5a802f4d8d6afbd816
2681 Seccomp mode of the process
2682 (since Linux 3.8, see
2685 .BR SECCOMP_MODE_DISABLED ;
2687 .BR SECCOMP_MODE_STRICT ;
2689 .BR SECCOMP_MODE_FILTER .
2690 This field is provided only if the kernel was built with the
2692 kernel configuration option enabled.
2694 .IR Speculation_Store_Bypass :
2695 .\" commit fae1fa0fc6cca8beee3ab8ed71d54f9a78fa3f64
2696 Speculation flaw mitigation state
2697 (since Linux 4.17, see
2701 Hexadecimal mask of CPUs on which this process may run
2702 (since Linux 2.6.24, see
2705 .IR Cpus_allowed_list :
2706 Same as previous, but in "list format"
2707 (since Linux 2.6.26, see
2711 Mask of memory nodes allowed to this process
2712 (since Linux 2.6.24, see
2715 .IR Mems_allowed_list :
2716 Same as previous, but in "list format"
2717 (since Linux 2.6.26, see
2720 .IR voluntary_ctxt_switches ", " nonvoluntary_ctxt_switches :
2721 Number of voluntary and involuntary context switches (since Linux 2.6.23).
2724 .IR /proc/[pid]/syscall " (since Linux 2.6.27)"
2725 .\" commit ebcb67341fee34061430f3367f2e507e52ee051b
2726 This file exposes the system call number and argument registers for the
2727 system call currently being executed by the process,
2728 followed by the values of the stack pointer and program counter registers.
2729 The values of all six argument registers are exposed,
2730 although most system calls use fewer registers.
2732 If the process is blocked, but not in a system call,
2733 then the file displays \-1 in place of the system call number,
2734 followed by just the values of the stack pointer and program counter.
2735 If process is not blocked, then the file contains just the string "running".
2737 This file is present only if the kernel was configured with
2738 .BR CONFIG_HAVE_ARCH_TRACEHOOK .
2740 Permission to access this file is governed by a ptrace access mode
2741 .B PTRACE_MODE_ATTACH_FSCREDS
2745 .IR /proc/[pid]/task " (since Linux 2.6.0)"
2746 .\" Precisely: Linux 2.6.0-test6
2747 This is a directory that contains one subdirectory
2748 for each thread in the process.
2749 The name of each subdirectory is the numerical thread ID
2754 Within each of these subdirectories, there is a set of
2755 files with the same names and contents as under the
2758 For attributes that are shared by all threads, the contents for
2759 each of the files under the
2761 subdirectories will be the same as in the corresponding
2765 (e.g., in a multithreaded process, all of the
2767 files will have the same value as the
2769 file in the parent directory, since all of the threads in a process
2770 share a working directory).
2771 For attributes that are distinct for each thread,
2772 the corresponding files under
2774 may have different values (e.g., various fields in each of the
2775 .I task/[tid]/status
2776 files may be different for each thread),
2777 .\" in particular: "children" :/
2778 or they might not exist in
2782 .\" The following was still true as at kernel 2.6.13
2783 In a multithreaded process, the contents of the
2785 directory are not available if the main thread has already terminated
2786 (typically by calling
2787 .BR pthread_exit (3)).
2790 .IR /proc/[pid]/task/[tid]/children " (since Linux 3.5)"
2791 .\" commit 818411616baf46ceba0cff6f05af3a9b294734f7
2792 A space-separated list of child tasks of this task.
2793 Each child task is represented by its TID.
2795 .\" see comments in get_children_pid() in fs/proc/array.c
2796 This option is intended for use by the checkpoint-restore (CRIU) system,
2797 and reliably provides a list of children only if all of the child processes
2798 are stopped or frozen.
2799 It does not work properly if children of the target task exit while
2800 the file is being read!
2801 Exiting children may cause non-exiting children to be omitted from the list.
2802 This makes this interface even more unreliable than classic PID-based
2803 approaches if the inspected task and its children aren't frozen,
2804 and most code should probably not use this interface.
2806 Until Linux 4.2, the presence of this file was governed by the
2807 .B CONFIG_CHECKPOINT_RESTORE
2808 kernel configuration option.
2810 .\" commit 2e13ba54a2682eea24918b87ad3edf70c2cf085b
2811 it is governed by the
2812 .B CONFIG_PROC_CHILDREN
2815 .IR /proc/[pid]/timers " (since Linux 3.10)"
2816 .\" commit 5ed67f05f66c41e39880a6d61358438a25f9fee5
2817 .\" commit 48f6a7a511ef8823fdff39afee0320092d43a8a0
2818 A list of the POSIX timers for this process.
2819 Each timer is listed with a line that starts with the string "ID:".
2825 signal: 60/00007fff86e452a8
2826 notify: signal/pid.2634
2829 signal: 60/00007fff86e452a8
2830 notify: signal/pid.2634
2835 The lines shown for each timer have the following meanings:
2839 The ID for this timer.
2840 This is not the same as the timer ID returned by
2841 .BR timer_create (2);
2842 rather, it is the same kernel-internal ID that is available via the
2850 This is the signal number that this timer uses to deliver notifications
2851 followed by a slash, and then the
2853 value supplied to the signal handler.
2854 Valid only for timers that notify via a signal.
2857 The part before the slash specifies the mechanism
2858 that this timer uses to deliver notifications,
2859 and is one of "thread", "signal", or "none".
2860 Immediately following the slash is either the string "tid" for timers
2863 notification, or "pid" for timers that notify by other mechanisms.
2864 Following the "." is the PID of the process
2865 (or the kernel thread ID of the thread) that will be delivered
2866 a signal if the timer delivers notifications via a signal.
2869 This field identifies the clock that the timer uses for measuring time.
2870 For most clocks, this is a number that matches one of the user-space
2872 constants exposed via
2874 .B CLOCK_PROCESS_CPUTIME_ID
2875 timers display with a value of \-6
2877 .B CLOCK_THREAD_CPUTIME_ID
2878 timers display with a value of \-2
2882 This file is available only when the kernel was configured with
2883 .BR CONFIG_CHECKPOINT_RESTORE .
2885 .IR /proc/[pid]/timerslack_ns " (since Linux 4.6)"
2886 .\" commit da8b44d5a9f8bf26da637b7336508ca534d6b319
2887 .\" commit 5de23d435e88996b1efe0e2cebe242074ce67c9e
2888 This file exposes the process's "current" timer slack value,
2889 expressed in nanoseconds.
2890 The file is writable,
2891 allowing the process's timer slack value to be changed.
2892 Writing 0 to this file resets the "current" timer slack to the
2893 "default" timer slack value.
2894 For further details, see the discussion of
2895 .BR PR_SET_TIMERSLACK
2900 permission to access this file was governed by a ptrace access mode
2901 .B PTRACE_MODE_ATTACH_FSCREDS
2904 However, this was subsequently deemed too strict a requirement
2905 (and had the side effect that requiring a process to have the
2907 capability would also allow it to view and change any process's memory).
2908 Therefore, since Linux 4.9,
2909 .\" commit 7abbaf94049914f074306d960b0f968ffe52e59f
2912 capability is required to access this file.
2914 .IR /proc/[pid]/uid_map ", " /proc/[pid]/gid_map " (since Linux 3.5)"
2916 .BR user_namespaces (7).
2918 .IR /proc/[pid]/wchan " (since Linux 2.6.0)"
2919 The symbolic name corresponding to the location
2920 in the kernel where the process is sleeping.
2922 Permission to access this file is governed by a ptrace access mode
2923 .B PTRACE_MODE_READ_FSCREDS
2928 There is a numerical subdirectory for each running thread
2929 that is not a thread group leader
2930 (i.e., a thread whose thread ID is not the same as its process ID);
2931 the subdirectory is named by the thread ID.
2932 Each one of these subdirectories contains files and subdirectories
2933 exposing information about the thread with the thread ID
2935 The contents of these directories are the same as the corresponding
2936 .IR /proc/[pid]/task/[tid]
2943 visible when iterating through
2949 visible when one uses
2951 to view the contents of
2953 However, the pathnames of these directories are visible to
2954 (i.e., usable as arguments in)
2955 system calls that operate on pathnames.
2958 Advanced power management version and battery information when
2960 is defined at kernel compilation time.
2963 This file contains information which is used for diagnosing memory
2964 fragmentation issues.
2965 Each line starts with the identification of the node and the name
2966 of the zone which together identify a memory region
2968 followed by the count of available chunks of a certain order in
2969 which these zones are split.
2970 The size in bytes of a certain order is given by the formula:
2972 (2^order)\ *\ PAGE_SIZE
2974 The binary buddy allocator algorithm inside the kernel will split
2975 one chunk into two chunks of a smaller order (thus with half the
2976 size) or combine two contiguous chunks into one larger chunk of
2977 a higher order (thus with double the size) to satisfy allocation
2978 requests and to counter memory fragmentation.
2979 The order matches the column number, when starting to count at zero.
2981 For example on an x86-64 system:
2985 Node 0, zone DMA 1 1 1 0 2 1 1 0 1 1 3
2986 Node 0, zone DMA32 65 47 4 81 52 28 13 10 5 1 404
2987 Node 0, zone Normal 216 55 189 101 84 38 37 27 5 3 587
2991 In this example, there is one node containing three zones and there
2992 are 11 different chunk sizes.
2993 If the page size is 4 kilobytes, then the first zone called
2995 (on x86 the first 16 megabyte of memory) has 1 chunk of 4 kilobytes
2996 (order 0) available and has 3 chunks of 4 megabytes (order 10) available.
2998 If the memory is heavily fragmented, the counters for higher
2999 order chunks will be zero and allocation of large contiguous areas
3002 Further information about the zones can be found in
3003 .IR /proc/zoneinfo .
3006 Contains subdirectories for installed busses.
3009 Subdirectory for PCMCIA devices when
3011 is set at kernel compilation time.
3013 .I /proc/bus/pccard/drivers
3016 Contains various bus subdirectories and pseudo-files containing
3017 information about PCI busses, installed devices, and device
3019 Some of these files are not ASCII.
3021 .I /proc/bus/pci/devices
3022 Information about PCI devices.
3023 They may be accessed through
3028 .IR /proc/cgroups " (since Linux 2.6.24)"
3033 Arguments passed to the Linux kernel at boot time.
3034 Often done via a boot manager such as
3039 .IR /proc/config.gz " (since Linux 2.6)"
3040 This file exposes the configuration options that were used
3041 to build the currently running kernel,
3042 in the same format as they would be shown in the
3044 file that resulted when configuring the kernel (using
3045 .IR "make xconfig" ,
3048 The file contents are compressed; view or search them using
3052 As long as no changes have been made to the following file,
3055 are the same as those provided by:
3059 cat /lib/modules/$(uname \-r)/build/.config
3064 is provided only if the kernel is configured with
3065 .BR CONFIG_IKCONFIG_PROC .
3068 A list of the ciphers provided by the kernel crypto API.
3069 For details, see the kernel
3070 .I "Linux Kernel Crypto API"
3071 documentation available under the kernel source directory
3072 .I Documentation/crypto/
3073 .\" commit 3b72c814a8e8cd638e1ba0da4dfce501e9dff5af
3075 .I Documentation/DocBook
3077 the documentation can be built using a command such as
3079 in the root directory of the kernel source tree).
3082 This is a collection of CPU and system architecture dependent items,
3083 for each supported architecture a different list.
3084 Two common entries are \fIprocessor\fP which gives CPU number and
3085 \fIbogomips\fP; a system constant that is calculated
3086 during kernel initialization.
3087 SMP machines have information for
3091 command gathers its information from this file.
3094 Text listing of major numbers and device groups.
3095 This can be used by MAKEDEV scripts for consistency with the kernel.
3097 .IR /proc/diskstats " (since Linux 2.5.69)"
3098 This file contains disk I/O statistics for each disk device.
3099 See the Linux kernel source file
3100 .I Documentation/iostats.txt
3101 for further information.
3104 This is a list of the registered \fIISA\fP DMA (direct memory access)
3110 .I /proc/execdomains
3111 List of the execution domains (ABI personalities).
3114 Frame buffer information when
3116 is defined during kernel compilation.
3118 .I /proc/filesystems
3119 A text listing of the filesystems which are supported by the kernel,
3120 namely filesystems which were compiled into the kernel or whose kernel
3121 modules are currently loaded.
3123 .BR filesystems (5).)
3124 If a filesystem is marked with "nodev",
3125 this means that it does not require a block device to be mounted
3126 (e.g., virtual filesystem, network filesystem).
3128 Incidentally, this file may be used by
3130 when no filesystem is specified and it didn't manage to determine the
3132 Then filesystems contained in this file are tried
3133 (excepted those that are marked with "nodev").
3136 .\" FIXME Much more needs to be said about /proc/fs
3138 Contains subdirectories that in turn contain files
3139 with information about (certain) mounted filesystems.
3143 exists on systems with the IDE bus.
3144 There are directories for each IDE channel and attached device.
3149 cache buffer size in KB
3150 capacity number of sectors
3151 driver driver version
3152 geometry physical and logical geometry
3153 identify in hexadecimal
3155 model manufacturer's model number
3156 settings drive settings
3157 smart_thresholds in hexadecimal
3158 smart_values in hexadecimal
3164 utility provides access to this information in a friendly format.
3167 This is used to record the number of interrupts per CPU per IO device.
3169 for the i386 and x86-64 architectures, at least, this also includes
3170 interrupts internal to the system (that is, not associated with a device
3171 as such), such as NMI (nonmaskable interrupt), LOC (local timer interrupt),
3172 and for SMP systems, TLB (TLB flush interrupt), RES (rescheduling
3173 interrupt), CAL (remote function call interrupt), and possibly others.
3174 Very easy to read formatting, done in ASCII.
3177 I/O memory map in Linux 2.4.
3180 This is a list of currently registered Input-Output port regions that
3183 .IR /proc/kallsyms " (since Linux 2.5.71)"
3184 This holds the kernel exported symbol definitions used by the
3186 tools to dynamically link and bind loadable modules.
3187 In Linux 2.5.47 and earlier, a similar file with slightly different syntax
3192 This file represents the physical memory of the system and is stored
3193 in the ELF core file format.
3194 With this pseudo-file, and an unstripped
3196 .RI ( /usr/src/linux/vmlinux )
3197 binary, GDB can be used to
3198 examine the current state of any kernel data structures.
3200 The total length of the file is the size of physical memory (RAM) plus
3203 .IR /proc/keys " (since Linux 2.6.10)"
3207 .IR /proc/key\-users " (since Linux 2.6.10)"
3212 This file can be used instead of the
3214 system call to read kernel messages.
3215 A process must have superuser
3216 privileges to read this file, and only one process should read this
3218 This file should not be read if a syslog process is running
3221 system call facility to log kernel messages.
3223 Information in this file is retrieved with the
3227 .IR /proc/kpagecgroup " (since Linux 4.3)"
3228 .\" commit 80ae2fdceba8313b0433f899bdd9c6c463291a17
3229 This file contains a 64-bit inode number of
3230 the memory cgroup each page is charged to,
3231 indexed by page frame number (see the discussion of
3232 .IR /proc/[pid]/pagemap ).
3235 .IR /proc/kpagecgroup
3236 file is present only if the
3238 kernel configuration option is enabled.
3240 .IR /proc/kpagecount " (since Linux 2.6.25)"
3241 This file contains a 64-bit count of the number of
3242 times each physical page frame is mapped,
3243 indexed by page frame number (see the discussion of
3244 .IR /proc/[pid]/pagemap ).
3247 .IR /proc/kpagecount
3248 file is present only if the
3249 .B CONFIG_PROC_PAGE_MONITOR
3250 kernel configuration option is enabled.
3252 .IR /proc/kpageflags " (since Linux 2.6.25)"
3253 This file contains 64-bit masks corresponding to each physical page frame;
3254 it is indexed by page frame number (see the discussion of
3255 .IR /proc/[pid]/pagemap ).
3256 The bits are as follows:
3269 11 - KPF_MMAP (since Linux 2.6.31)
3270 12 - KPF_ANON (since Linux 2.6.31)
3271 13 - KPF_SWAPCACHE (since Linux 2.6.31)
3272 14 - KPF_SWAPBACKED (since Linux 2.6.31)
3273 15 - KPF_COMPOUND_HEAD (since Linux 2.6.31)
3274 16 - KPF_COMPOUND_TAIL (since Linux 2.6.31)
3275 17 - KPF_HUGE (since Linux 2.6.31)
3276 18 - KPF_UNEVICTABLE (since Linux 2.6.31)
3277 19 - KPF_HWPOISON (since Linux 2.6.31)
3278 20 - KPF_NOPAGE (since Linux 2.6.31)
3279 21 - KPF_KSM (since Linux 2.6.32)
3280 22 - KPF_THP (since Linux 3.4)
3281 23 - KPF_BALLOON (since Linux 3.18)
3282 .\" KPF_BALLOON: commit 09316c09dde33aae14f34489d9e3d243ec0d5938
3283 24 - KPF_ZERO_PAGE (since Linux 4.0)
3284 .\" KPF_ZERO_PAGE: commit 56873f43abdcd574b25105867a990f067747b2f4
3285 25 - KPF_IDLE (since Linux 4.3)
3286 .\" KPF_IDLE: commit f074a8f49eb87cde95ac9d040ad5e7ea4f029738
3288 For further details on the meanings of these bits,
3289 see the kernel source file
3290 .IR Documentation/admin\-guide/mm/pagemap.rst .
3291 Before kernel 2.6.29,
3292 .\" commit ad3bdefe877afb47480418fdb05ecd42842de65e
3293 .\" commit e07a4b9217d1e97d2f3a62b6b070efdc61212110
3299 did not report correctly.
3302 .IR /proc/kpageflags
3303 file is present only if the
3304 .B CONFIG_PROC_PAGE_MONITOR
3305 kernel configuration option is enabled.
3307 .IR /proc/ksyms " (Linux 1.1.23\(en2.5.47)"
3309 .IR /proc/kallsyms .
3312 The first three fields in this file are load average figures
3313 giving the number of jobs in the run queue (state R)
3314 or waiting for disk I/O (state D) averaged over 1, 5, and 15 minutes.
3315 They are the same as the load average numbers given by
3318 The fourth field consists of two numbers separated by a slash (/).
3319 The first of these is the number of currently runnable kernel
3320 scheduling entities (processes, threads).
3321 The value after the slash is the number of kernel scheduling entities
3322 that currently exist on the system.
3323 The fifth field is the PID of the process that was most
3324 recently created on the system.
3327 This file shows current file locks
3328 .RB ( flock "(2) and " fcntl (2))
3332 An example of the content shown in this file is the following:
3336 1: POSIX ADVISORY READ 5433 08:01:7864448 128 128
3337 2: FLOCK ADVISORY WRITE 2001 08:01:7864554 0 EOF
3338 3: FLOCK ADVISORY WRITE 1568 00:2f:32388 0 EOF
3339 4: POSIX ADVISORY WRITE 699 00:16:28457 0 EOF
3340 5: POSIX ADVISORY WRITE 764 00:16:21448 0 0
3341 6: POSIX ADVISORY READ 3548 08:01:7867240 1 1
3342 7: POSIX ADVISORY READ 3548 08:01:7865567 1826 2335
3343 8: OFDLCK ADVISORY WRITE \-1 08:01:8713209 128 191
3347 The fields shown in each line are as follows:
3350 The ordinal position of the lock in the list.
3353 Values that may appear here include:
3357 This is a BSD file lock created using
3361 This is an open file description (OFD) lock created using
3365 This is a POSIX byte-range lock created using
3369 Among the strings that can appear here are the following:
3373 This is an advisory lock.
3376 This is a mandatory lock.
3380 Values that can appear here are:
3384 This is a POSIX or OFD read lock, or a BSD shared lock.
3387 This is a POSIX or OFD write lock, or a BSD exclusive lock.
3390 The PID of the process that owns the lock.
3392 Because OFD locks are not owned by a single process
3393 (since multiple processes may have file descriptors that
3394 refer to the same open file description),
3395 the value \-1 is displayed in this field for OFD locks.
3396 (Before kernel 4.14,
3397 .\" commit 9d5b86ac13c573795525ecac6ed2db39ab23e2a8
3398 a bug meant that the PID of the process that
3399 initially acquired the lock was displayed instead of the value \-1.)
3401 Three colon-separated subfields that identify the major and minor device
3402 ID of the device containing the filesystem where the locked file resides,
3403 followed by the inode number of the locked file.
3405 The byte offset of the first byte of the lock.
3406 For BSD locks, this value is always 0.
3408 The byte offset of the last byte of the lock.
3410 in this field means that the lock extends to the end of the file.
3411 For BSD locks, the value shown is always
3416 .\" commit d67fd44f697dff293d7cdc29af929241b669affe
3417 the list of locks shown in
3419 is filtered to show just the locks for the processes in the PID
3421 .BR pid_namespaces (7))
3424 filesystem was mounted.
3425 (In the initial PID namespace,
3426 there is no filtering of the records shown in this file.)
3430 command provides a bit more information about each lock.
3432 .IR /proc/malloc " (only up to and including Linux 2.2)"
3433 .\" It looks like this only ever did something back in 1.0 days
3434 This file is present only if
3435 .B CONFIG_DEBUG_MALLOC
3436 was defined during compilation.
3439 This file reports statistics about memory usage on the system.
3442 to report the amount of free and used memory (both physical and swap)
3443 on the system as well as the shared memory and buffers used by the
3445 Each line of the file consists of a parameter name, followed by a colon,
3446 the value of the parameter, and an option unit of measurement (e.g., "kB").
3447 The list below describes the parameter names and
3448 the format specifier required to read the field value.
3449 Except as noted below,
3450 all of the fields have been present since at least Linux 2.6.0.
3451 Some fields are displayed only if the kernel was configured
3452 with various options; those dependencies are noted in the list.
3456 Total usable RAM (i.e., physical RAM minus a few reserved
3457 bits and the kernel binary code).
3461 .IR LowFree + HighFree .
3463 .IR MemAvailable " %lu (since Linux 3.14)"
3464 An estimate of how much memory is available for starting new
3465 applications, without swapping.
3468 Relatively temporary storage for raw disk blocks that
3469 shouldn't get tremendously large (20MB or so).
3472 In-memory cache for files read from the disk (the page cache).
3476 .IR SwapCached " %lu"
3477 Memory that once was swapped out, is swapped back in but
3478 still also is in the swap file.
3479 (If memory pressure is high, these pages
3480 don't need to be swapped out again because they are already
3485 Memory that has been used more recently and usually not
3486 reclaimed unless absolutely necessary.
3489 Memory which has been less recently used.
3490 It is more eligible to be reclaimed for other purposes.
3492 .IR Active(anon) " %lu (since Linux 2.6.28)"
3495 .IR Inactive(anon) " %lu (since Linux 2.6.28)"
3498 .IR Active(file) " %lu (since Linux 2.6.28)"
3501 .IR Inactive(file) " %lu (since Linux 2.6.28)"
3504 .IR Unevictable " %lu (since Linux 2.6.28)"
3505 (From Linux 2.6.28 to 2.6.30,
3506 \fBCONFIG_UNEVICTABLE_LRU\fP was required.)
3509 .IR Mlocked " %lu (since Linux 2.6.28)"
3510 (From Linux 2.6.28 to 2.6.30,
3511 \fBCONFIG_UNEVICTABLE_LRU\fP was required.)
3514 .IR HighTotal " %lu"
3515 (Starting with Linux 2.6.19, \fBCONFIG_HIGHMEM\fP is required.)
3516 Total amount of highmem.
3517 Highmem is all memory above ~860MB of physical memory.
3518 Highmem areas are for use by user-space programs,
3519 or for the page cache.
3520 The kernel must use tricks to access
3521 this memory, making it slower to access than lowmem.
3524 (Starting with Linux 2.6.19, \fBCONFIG_HIGHMEM\fP is required.)
3525 Amount of free highmem.
3528 (Starting with Linux 2.6.19, \fBCONFIG_HIGHMEM\fP is required.)
3529 Total amount of lowmem.
3530 Lowmem is memory which can be used for everything that
3531 highmem can be used for, but it is also available for the
3532 kernel's use for its own data structures.
3533 Among many other things,
3534 it is where everything from
3537 Bad things happen when you're out of lowmem.
3540 (Starting with Linux 2.6.19, \fBCONFIG_HIGHMEM\fP is required.)
3541 Amount of free lowmem.
3543 .IR MmapCopy " %lu (since Linux 2.6.29)"
3548 .IR SwapTotal " %lu"
3549 Total amount of swap space available.
3552 Amount of swap space that is currently unused.
3555 Memory which is waiting to get written back to the disk.
3557 .IR Writeback " %lu"
3558 Memory which is actively being written back to the disk.
3560 .IR AnonPages " %lu (since Linux 2.6.18)"
3561 Non-file backed pages mapped into user-space page tables.
3564 Files which have been mapped into memory (with
3568 .IR Shmem " %lu (since Linux 2.6.32)"
3569 Amount of memory consumed in
3573 .IR KReclaimable " %lu (since Linux 4.20)"
3574 Kernel allocations that the kernel will attempt to reclaim
3575 under memory pressure.
3578 (below), and other direct allocations with a shrinker.
3581 In-kernel data structures cache.
3585 .IR SReclaimable " %lu (since Linux 2.6.19)"
3588 that might be reclaimed, such as caches.
3590 .IR SUnreclaim " %lu (since Linux 2.6.19)"
3593 that cannot be reclaimed on memory pressure.
3595 .IR KernelStack " %lu (since Linux 2.6.32)"
3596 Amount of memory allocated to kernel stacks.
3598 .IR PageTables " %lu (since Linux 2.6.18)"
3599 Amount of memory dedicated to the lowest level of page tables.
3601 .IR Quicklists " %lu (since Linux 2.6.27)"
3602 (\fBCONFIG_QUICKLIST\fP is required.)
3605 .IR NFS_Unstable " %lu (since Linux 2.6.18)"
3606 NFS pages sent to the server, but not yet committed to stable storage.
3608 .IR Bounce " %lu (since Linux 2.6.18)"
3609 Memory used for block device "bounce buffers".
3611 .IR WritebackTmp " %lu (since Linux 2.6.26)"
3612 Memory used by FUSE for temporary writeback buffers.
3614 .IR CommitLimit " %lu (since Linux 2.6.10)"
3615 This is the total amount of memory currently available to
3616 be allocated on the system, expressed in kilobytes.
3617 This limit is adhered to
3618 only if strict overcommit accounting is enabled (mode 2 in
3619 .IR /proc/sys/vm/overcommit_memory ).
3620 The limit is calculated according to the formula described under
3621 .IR /proc/sys/vm/overcommit_memory .
3622 For further details, see the kernel source file
3623 .IR Documentation/vm/overcommit\-accounting.rst .
3625 .IR Committed_AS " %lu"
3626 The amount of memory presently allocated on the system.
3627 The committed memory is a sum of all of the memory which
3628 has been allocated by processes, even if it has not been
3629 "used" by them as of yet.
3630 A process which allocates 1GB of memory (using
3632 or similar), but touches only 300MB of that memory will show up
3633 as using only 300MB of memory even if it has the address space
3634 allocated for the entire 1GB.
3636 This 1GB is memory which has been "committed" to by the VM
3637 and can be used at any time by the allocating application.
3638 With strict overcommit enabled on the system (mode 2 in
3639 .IR /proc/sys/vm/overcommit_memory ),
3640 allocations which would exceed the
3642 will not be permitted.
3643 This is useful if one needs to guarantee that processes will not
3644 fail due to lack of memory once that memory has been successfully allocated.
3646 .IR VmallocTotal " %lu"
3647 Total size of vmalloc memory area.
3649 .IR VmallocUsed " %lu"
3650 Amount of vmalloc area which is used.
3652 .\" commit a5ad88ce8c7fae7ddc72ee49a11a75aa837788e0
3653 this field is no longer calculated, and is hard coded as 0.
3655 .IR /proc/vmallocinfo .
3657 .IR VmallocChunk " %lu"
3658 Largest contiguous block of vmalloc area which is free.
3660 .\" commit a5ad88ce8c7fae7ddc72ee49a11a75aa837788e0
3661 this field is no longer calculated and is hard coded as 0.
3663 .IR /proc/vmallocinfo .
3665 .IR HardwareCorrupted " %lu (since Linux 2.6.32)"
3666 (\fBCONFIG_MEMORY_FAILURE\fP is required.)
3669 .IR LazyFree " %lu (since Linux 4.12)"
3670 Shows the amount of memory marked by
3674 .IR AnonHugePages " %lu (since Linux 2.6.38)"
3675 (\fBCONFIG_TRANSPARENT_HUGEPAGE\fP is required.)
3676 Non-file backed huge pages mapped into user-space page tables.
3678 .IR ShmemHugePages " %lu (since Linux 4.8)"
3679 (\fBCONFIG_TRANSPARENT_HUGEPAGE\fP is required.)
3680 Memory used by shared memory (shmem) and
3682 allocated with huge pages
3684 .IR ShmemPmdMapped " %lu (since Linux 4.8)"
3685 (\fBCONFIG_TRANSPARENT_HUGEPAGE\fP is required.)
3686 Shared memory mapped into user space with huge pages.
3688 .IR CmaTotal " %lu (since Linux 3.1)"
3689 Total CMA (Contiguous Memory Allocator) pages.
3690 (\fBCONFIG_CMA\fP is required.)
3692 .IR CmaFree " %lu (since Linux 3.1)"
3693 Free CMA (Contiguous Memory Allocator) pages.
3694 (\fBCONFIG_CMA\fP is required.)
3696 .IR HugePages_Total " %lu"
3697 (\fBCONFIG_HUGETLB_PAGE\fP is required.)
3698 The size of the pool of huge pages.
3700 .IR HugePages_Free " %lu"
3701 (\fBCONFIG_HUGETLB_PAGE\fP is required.)
3702 The number of huge pages in the pool that are not yet allocated.
3704 .IR HugePages_Rsvd " %lu (since Linux 2.6.17)"
3705 (\fBCONFIG_HUGETLB_PAGE\fP is required.)
3706 This is the number of huge pages for
3707 which a commitment to allocate from the pool has been made,
3708 but no allocation has yet been made.
3709 These reserved huge pages
3710 guarantee that an application will be able to allocate a
3711 huge page from the pool of huge pages at fault time.
3713 .IR HugePages_Surp " %lu (since Linux 2.6.24)"
3714 (\fBCONFIG_HUGETLB_PAGE\fP is required.)
3715 This is the number of huge pages in
3716 the pool above the value in
3717 .IR /proc/sys/vm/nr_hugepages .
3718 The maximum number of surplus huge pages is controlled by
3719 .IR /proc/sys/vm/nr_overcommit_hugepages .
3721 .IR Hugepagesize " %lu"
3722 (\fBCONFIG_HUGETLB_PAGE\fP is required.)
3723 The size of huge pages.
3725 .IR DirectMap4k " %lu (since Linux 2.6.27)"
3726 Number of bytes of RAM linearly mapped by kernel in 4kB pages.
3729 .IR DirectMap4M " %lu (since Linux 2.6.27)"
3730 Number of bytes of RAM linearly mapped by kernel in 4MB pages.
3737 .IR DirectMap2M " %lu (since Linux 2.6.27)"
3738 Number of bytes of RAM linearly mapped by kernel in 2MB pages.
3745 .IR DirectMap1G " %lu (since Linux 2.6.27)"
3749 .B CONFIG_X86_DIRECT_GBPAGES
3754 A text list of the modules that have been loaded by the system.
3759 Before kernel 2.4.19, this file was a list
3760 of all the filesystems currently mounted on the system.
3761 With the introduction of per-process mount namespaces in Linux 2.4.19 (see
3762 .BR mount_namespaces (7)),
3763 this file became a link to
3764 .IR /proc/self/mounts ,
3765 which lists the mount points of the process's own mount namespace.
3766 The format of this file is documented in
3770 Memory Type Range Registers.
3771 See the Linux kernel source file
3772 .I Documentation/x86/mtrr.txt
3773 .\" commit 7225e75144b9718cbbe1820d9c011c809d5773fd
3775 .I Documentation/mtrr.txt
3776 before Linux 2.6.28)
3780 This directory contains various files and subdirectories containing
3781 information about the networking layer.
3782 The files contain ASCII structures and are,
3783 therefore, readable with
3785 However, the standard
3787 suite provides much cleaner access to these files.
3789 With the advent of network namespaces,
3790 various information relating to the network stack is virtualized (see
3791 .BR network_namespaces (7)).
3792 Thus, since Linux 2.6.25,
3793 .\" commit e9720acd728a46cb40daa52c99a979f7c4ff195c
3795 is a symbolic link to the directory
3796 .IR /proc/self/net ,
3797 which contains the same files and directories as listed below.
3798 However, these files and directories now expose information
3799 for the network namespace of which the process is a member.
3802 This holds an ASCII readable dump of the kernel ARP table used for
3803 address resolutions.
3804 It will show both dynamically learned and preprogrammed ARP entries.
3809 IP address HW type Flags HW address Mask Device
3810 192.168.0.50 0x1 0x2 00:50:BF:25:68:F3 * eth0
3811 192.168.0.250 0x1 0xc 00:00:00:00:00:00 * eth0
3815 Here "IP address" is the IPv4 address of the machine and the "HW type"
3816 is the hardware type of the address from RFC\ 826.
3817 The flags are the internal
3818 flags of the ARP structure (as defined in
3819 .IR /usr/include/linux/if_arp.h )
3821 the "HW address" is the data link layer mapping for that IP address if
3825 The dev pseudo-file contains network device status information.
3827 the number of received and sent packets, the number of errors and
3829 and other basic statistics.
3830 These are used by the
3832 program to report device status.
3837 Inter\-| Receive | Transmit
3838 face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
3839 lo: 2776770 11307 0 0 0 0 0 0 2776770 11307 0 0 0 0 0 0
3840 eth0: 1215645 2751 0 0 0 0 0 0 1782404 4324 0 0 0 427 0 0
3841 ppp0: 1622270 5552 1 0 0 0 0 0 354130 5669 0 0 0 0 0 0
3842 tap0: 7714 81 0 0 0 0 0 0 7714 81 0 0 0 0 0 0
3846 .\" .I /proc/net/ipx
3849 .\" .I /proc/net/ipx_route
3852 .I /proc/net/dev_mcast
3854 .IR /usr/src/linux/net/core/dev_mcast.c :
3858 indx interface_name dmi_u dmi_g dmi_address
3859 2 eth0 1 0 01005e000001
3860 3 eth1 1 0 01005e000001
3861 4 eth2 1 0 01005e000001
3866 Internet Group Management Protocol.
3868 .IR /usr/src/linux/net/core/igmp.c .
3871 This file uses the same format as the
3873 file and contains the current reverse mapping database used to provide
3875 reverse address lookup services.
3876 If RARP is not configured into the
3878 this file will not be present.
3881 Holds a dump of the RAW socket table.
3882 Much of the information is not of
3884 apart from debugging.
3885 The "sl" value is the kernel hash slot for the
3887 the "local_address" is the local address and protocol number pair.
3889 the internal status of the socket.
3890 The "tx_queue" and "rx_queue" are the
3891 outgoing and incoming data queue in terms of kernel memory usage.
3892 The "tr", "tm\->when", and "rexmits" fields are not used by RAW.
3894 field holds the effective UID of the creator of the socket.
3896 .\" .I /proc/net/route
3897 .\" No information, but looks similar to
3901 This file holds the ASCII data needed for the IP, ICMP, TCP, and UDP
3903 information bases for an SNMP agent.
3906 Holds a dump of the TCP socket table.
3907 Much of the information is not
3908 of use apart from debugging.
3909 The "sl" value is the kernel hash slot
3910 for the socket, the "local_address" is the local address and port number pair.
3911 The "rem_address" is the remote address and port number pair
3913 \&"St" is the internal status of the socket.
3914 The "tx_queue" and "rx_queue" are the
3915 outgoing and incoming data queue in terms of kernel memory usage.
3916 The "tr", "tm\->when", and "rexmits" fields hold internal information of
3917 the kernel socket state and are useful only for debugging.
3919 field holds the effective UID of the creator of the socket.
3922 Holds a dump of the UDP socket table.
3923 Much of the information is not of
3924 use apart from debugging.
3925 The "sl" value is the kernel hash slot for the
3926 socket, the "local_address" is the local address and port number pair.
3927 The "rem_address" is the remote address and port number pair
3929 "St" is the internal status of the socket.
3930 The "tx_queue" and "rx_queue" are the outgoing and incoming data queue
3931 in terms of kernel memory usage.
3932 The "tr", "tm\->when", and "rexmits" fields
3933 are not used by UDP.
3935 field holds the effective UID of the creator of the socket.
3940 sl local_address rem_address st tx_queue rx_queue tr rexmits tm\->when uid
3941 1: 01642C89:0201 0C642C89:03FF 01 00000000:00000001 01:000071BA 00000000 0
3942 1: 00000000:0801 00000000:0000 0A 00000000:00000000 00:00000000 6F000100 0
3943 1: 00000000:0201 00000000:0000 0A 00000000:00000000 00:00000000 00000000 0
3949 Lists the UNIX domain sockets present within the system and their
3955 Num RefCount Protocol Flags Type St Inode Path
3956 0: 00000002 00000000 00000000 0001 03 42
3957 1: 00000001 00000000 00010000 0001 01 1948 /dev/printer
3961 The fields are as follows:
3965 the kernel table slot number.
3968 the number of users of the socket.
3974 the internal kernel flags holding the status of the socket.
3980 sockets, this is 0001; for
3982 sockets, it is 0002; and for
3984 sockets, it is 0005.
3987 the internal state of the socket.
3990 the inode number of the socket.
3993 the bound pathname (if any) of the socket.
3994 Sockets in the abstract namespace are included in the list,
3995 and are shown with a
3997 that commences with the character '@'.
4000 .I /proc/net/netfilter/nfnetlink_queue
4001 This file contains information about netfilter user-space queueing, if used.
4002 Each line represents a queue.
4003 Queues that have not been subscribed to
4004 by user space are not shown.
4008 1 4207 0 2 65535 0 0 0 1
4009 (1) (2) (3)(4) (5) (6) (7) (8)
4013 The fields in each line are:
4017 The ID of the queue.
4018 This matches what is specified in the
4021 .B \-\-queue\-balance
4026 .BR iptables\-extensions (8)
4027 for more information.
4030 The netlink port ID subscribed to the queue.
4033 The number of packets currently queued and waiting to be processed by
4037 The copy mode of the queue.
4038 It is either 1 (metadata only) or 2
4039 (also copy payload data to user space).
4042 Copy range; that is, how many bytes of packet payload should be copied to
4047 Number of packets that had to be dropped by the kernel because
4048 too many packets are already waiting for user space to send back the mandatory
4049 accept/drop verdicts.
4053 Number of packets that were dropped within the netlink
4055 Such drops usually happen when the corresponding socket buffer is
4056 full; that is, user space is not able to read messages fast enough.
4060 Every queued packet is associated with a (32-bit)
4061 monotonically-increasing sequence number.
4062 This shows the ID of the most recent packet queued.
4065 The last number exists only for compatibility reasons and is always 1.
4068 Contains the major and minor numbers of each partition as well as the number
4069 of 1024-byte blocks and the partition name.
4072 This is a listing of all PCI devices found during kernel initialization
4073 and their configuration.
4075 This file has been deprecated in favor of a new
4078 .RI ( /proc/bus/pci ).
4079 It became optional in Linux 2.2 (available with
4080 .B CONFIG_PCI_OLD_PROC
4081 set at kernel compilation).
4082 It became once more nonoptionally enabled in Linux 2.4.
4083 Next, it was deprecated in Linux 2.6 (still available with
4084 .B CONFIG_PCI_LEGACY_PROC
4085 set), and finally removed altogether since Linux 2.6.17.
4086 .\" FIXME Document /proc/sched_debug (since Linux 2.6.23)
4087 .\" See also /proc/[pid]/sched
4089 .IR /proc/profile " (since Linux 2.4)"
4090 This file is present only if the kernel was booted with the
4092 command-line option.
4093 It exposes kernel profiling information in a binary format for use by
4094 .BR readprofile (1).
4095 Writing (e.g., an empty string) to this file resets the profiling counters;
4096 on some architectures,
4097 writing a binary integer "profiling multiplier" of size
4099 sets the profiling interrupt frequency.
4102 A directory with the
4104 mid-level pseudo-file and various SCSI low-level
4106 which contain a file for each SCSI host in this system, all of
4107 which give the status of some part of the SCSI IO subsystem.
4108 These files contain ASCII structures and are, therefore, readable with
4111 You can also write to some of the files to reconfigure the subsystem or
4112 switch certain features on or off.
4115 This is a listing of all SCSI devices known to the kernel.
4116 The listing is similar to the one seen during bootup.
4117 scsi currently supports only the \fIadd\-single\-device\fP command which
4118 allows root to add a hotplugged device to the list of known devices.
4124 echo \(aqscsi add\-single\-device 1 0 5 0\(aq > /proc/scsi/scsi
4129 host scsi1 to scan on SCSI channel 0 for a device on ID 5 LUN 0.
4131 is already a device known on this address or the address is invalid, an
4132 error will be returned.
4134 .I /proc/scsi/[drivername]
4135 \fI[drivername]\fP can currently be NCR53c7xx, aha152x, aha1542, aha1740,
4136 aic7xxx, buslogic, eata_dma, eata_pio, fdomain, in2000, pas16, qlogic,
4137 scsi_debug, seagate, t128, u15\-24f, ultrastore, or wd7000.
4138 These directories show up for all drivers that registered at least one
4140 Every directory contains one file per registered host.
4141 Every host-file is named after the number the host was assigned during
4144 Reading these files will usually show driver and host configuration,
4145 statistics, and so on.
4147 Writing to these files allows different things on different hosts.
4148 For example, with the \fIlatency\fP and \fInolatency\fP commands,
4149 root can switch on and off command latency measurement code in the
4151 With the \fIlockup\fP and \fIunlock\fP commands,
4152 root can control bus lockups simulated by the scsi_debug driver.
4155 This directory refers to the process accessing the
4158 and is identical to the
4160 directory named by the process ID of the same process.
4163 Information about kernel caches.
4169 kernel/system statistics.
4170 Varies with architecture.
4175 .I cpu 10132153 290696 3084719 46828483 16683 0 25195 0 175628 0
4177 .I cpu0 1393280 32966 572056 13343292 6130 0 17875 0 23933 0
4178 The amount of time, measured in units of
4179 USER_HZ (1/100ths of a second on most architectures, use
4180 .IR sysconf(_SC_CLK_TCK)
4181 to obtain the right value),
4182 .\" 1024 on Alpha and ia64
4183 that the system ("cpu" line) or the specific CPU ("cpu\fIN\fR" line)
4184 spent in various states:
4188 (1) Time spent in user mode.
4191 (2) Time spent in user mode with low priority (nice).
4194 (3) Time spent in system mode.
4197 (4) Time spent in the idle task.
4198 .\" FIXME . Actually, the following info about the /proc/stat 'cpu' field
4199 .\" does not seem to be quite right (at least in 2.6.12 or 3.6):
4200 .\" the idle time in /proc/uptime does not quite match this value
4201 This value should be USER_HZ times the
4206 .IR iowait " (since Linux 2.5.41)"
4207 (5) Time waiting for I/O to complete.
4208 This value is not reliable, for the following reasons:
4209 .\" See kernel commit 9c240d757658a3ae9968dd309e674c61f07c7f48
4212 The CPU will not wait for I/O to complete;
4213 iowait is the time that a task is waiting for I/O to complete.
4214 When a CPU goes into idle state for outstanding task I/O,
4215 another task will be scheduled on this CPU.
4217 On a multi-core CPU,
4218 the task waiting for I/O to complete is not running on any CPU,
4219 so the iowait of each CPU is difficult to calculate.
4221 The value in this field may
4223 in certain conditions.
4226 .IR irq " (since Linux 2.6.0)"
4227 .\" Precisely: Linux 2.6.0-test4
4228 (6) Time servicing interrupts.
4230 .IR softirq " (since Linux 2.6.0"
4231 .\" Precisely: Linux 2.6.0-test4
4232 (7) Time servicing softirqs.
4234 .IR steal " (since Linux 2.6.11)"
4235 (8) Stolen time, which is the time spent in other operating systems when
4236 running in a virtualized environment
4238 .IR guest " (since Linux 2.6.24)"
4239 (9) Time spent running a virtual CPU for guest
4240 operating systems under the control of the Linux kernel.
4241 .\" See Changelog entry for 5e84cfde51cf303d368fcb48f22059f37b3872de
4243 .IR guest_nice " (since Linux 2.6.33)"
4244 .\" commit ce0e7b28fb75cb003cfc8d0238613aaf1c55e797
4245 (10) Time spent running a niced guest (virtual CPU for guest
4246 operating systems under the control of the Linux kernel).
4249 \fIpage 5741 1808\fP
4250 The number of pages the system paged in and the number that were paged
4254 The number of swap pages that have been brought in and out.
4256 .\" FIXME . The following is not the full picture for the 'intr' of
4257 .\" /proc/stat on 2.6:
4259 This line shows counts of interrupts serviced since boot time,
4260 for each of the possible system interrupts.
4261 The first column is the total of all interrupts serviced
4262 including unnumbered architecture specific interrupts;
4263 each subsequent column is the total for that particular numbered interrupt.
4264 Unnumbered interrupts are not shown, only summed into the total.
4266 \fIdisk_io: (2,0):(31,30,5764,1,2) (3,0):\fP...
4267 (major,disk_idx):(noinfo, read_io_ops, blks_read, write_io_ops, blks_written)
4272 The number of context switches that the system underwent.
4274 \fIbtime 769041601\fP
4275 boot time, in seconds since the Epoch, 1970-01-01 00:00:00 +0000 (UTC).
4277 \fIprocesses 86031\fP
4278 Number of forks since boot.
4280 \fIprocs_running 6\fP
4281 Number of processes in runnable state.
4282 (Linux 2.5.45 onward.)
4284 \fIprocs_blocked 2\fP
4285 Number of processes blocked waiting for I/O to complete.
4286 (Linux 2.5.45 onward.)
4288 .I softirq 229245889 94 60001584 13619 5175704 2471304 28 51212741 59130143 0 51240672
4289 .\" commit d3d64df21d3d0de675a0d3ffa7c10514f3644b30
4290 This line shows the number of softirq for all CPUs.
4291 The first column is the total of all softirqs and
4292 each subsequent column is the total for particular softirq.
4293 (Linux 2.6.31 onward.)
4302 This directory (present since 1.3.57) contains a number of files
4303 and subdirectories corresponding to kernel variables.
4304 These variables can be read and sometimes modified using
4305 the \fI/proc\fP filesystem, and the (deprecated)
4309 String values may be terminated by either \(aq\e0\(aq or \(aq\en\(aq.
4311 Integer and long values may be written either in decimal or in
4312 hexadecimal notation (e.g. 0x3FFF).
4313 When writing multiple integer or long values, these may be separated
4314 by any of the following whitespace characters:
4315 \(aq\ \(aq, \(aq\et\(aq, or \(aq\en\(aq.
4316 Using other separators leads to the error
4319 .IR /proc/sys/abi " (since Linux 2.4.10)"
4320 This directory may contain files with application binary information.
4321 .\" On some systems, it is not present.
4322 See the Linux kernel source file
4323 .I Documentation/sysctl/abi.txt
4324 for more information.
4327 This directory may be empty.
4330 This directory contains device-specific information (e.g.,
4331 .IR dev/cdrom/info ).
4333 some systems, it may be empty.
4336 This directory contains the files and subdirectories for kernel variables
4337 related to filesystems.
4339 .I /proc/sys/fs/binfmt_misc
4340 Documentation for files in this directory can be found
4341 in the Linux kernel source in the file
4342 .IR Documentation/admin\-guide/binfmt\-misc.rst
4344 .IR Documentation/binfmt_misc.txt
4347 .IR /proc/sys/fs/dentry\-state " (since Linux 2.2)"
4348 This file contains information about the status of the
4349 directory cache (dcache).
4350 The file contains six numbers,
4351 .IR nr_dentry ", " nr_unused ", " age_limit " (age in seconds), "
4353 (pages requested by system) and two dummy values.
4357 is the number of allocated dentries (dcache entries).
4358 This field is unused in Linux 2.2.
4361 is the number of unused dentries.
4364 .\" looks like this is unused in kernels 2.2 to 2.6
4365 is the age in seconds after which dcache entries
4366 can be reclaimed when memory is short.
4369 .\" looks like this is unused in kernels 2.2 to 2.6
4370 is nonzero when the kernel has called shrink_dcache_pages() and the
4371 dcache isn't pruned yet.
4374 .I /proc/sys/fs/dir\-notify\-enable
4375 This file can be used to disable or enable the
4377 interface described in
4379 on a system-wide basis.
4380 A value of 0 in this file disables the interface,
4381 and a value of 1 enables it.
4383 .I /proc/sys/fs/dquot\-max
4384 This file shows the maximum number of cached disk quota entries.
4385 On some (2.4) systems, it is not present.
4386 If the number of free cached disk quota entries is very low and
4387 you have some awesome number of simultaneous system users,
4388 you might want to raise the limit.
4390 .I /proc/sys/fs/dquot\-nr
4391 This file shows the number of allocated disk quota
4392 entries and the number of free disk quota entries.
4394 .IR /proc/sys/fs/epoll " (since Linux 2.6.28)"
4395 This directory contains the file
4396 .IR max_user_watches ,
4397 which can be used to limit the amount of kernel memory consumed by the
4400 For further details, see
4403 .I /proc/sys/fs/file\-max
4405 a system-wide limit on the number of open files for all processes.
4406 System calls that fail when encountering this limit fail with the error
4410 which can be used by a process to set the per-process limit,
4412 on the number of files it may open.)
4414 of error messages in the kernel log about running out of file handles
4415 (look for "VFS: file\-max limit <number> reached"),
4416 try increasing this value:
4420 echo 100000 > /proc/sys/fs/file\-max
4424 Privileged processes
4425 .RB ( CAP_SYS_ADMIN )
4430 .I /proc/sys/fs/file\-nr
4431 This (read-only) file contains three numbers:
4432 the number of allocated file handles
4433 (i.e., the number of files presently opened);
4434 the number of free file handles;
4435 and the maximum number of file handles (i.e., the same value as
4436 .IR /proc/sys/fs/file\-max ).
4437 If the number of allocated file handles is close to the
4438 maximum, you should consider increasing the maximum.
4440 the kernel allocated file handles dynamically,
4441 but it didn't free them again.
4442 Instead the free file handles were kept in a list for reallocation;
4443 the "free file handles" value indicates the size of that list.
4444 A large number of free file handles indicates that there was
4445 a past peak in the usage of open file handles.
4446 Since Linux 2.6, the kernel does deallocate freed file handles,
4447 and the "free file handles" value is always zero.
4449 .IR /proc/sys/fs/inode\-max " (only present until Linux 2.2)"
4450 This file contains the maximum number of in-memory inodes.
4451 This value should be 3\(en4 times larger
4454 since \fIstdin\fP, \fIstdout\fP
4455 and network sockets also need an inode to handle them.
4456 When you regularly run out of inodes, you need to increase this value.
4458 Starting with Linux 2.4,
4459 there is no longer a static limit on the number of inodes,
4460 and this file is removed.
4462 .I /proc/sys/fs/inode\-nr
4463 This file contains the first two values from
4466 .I /proc/sys/fs/inode\-state
4468 contains seven numbers:
4470 .IR nr_free_inodes ,
4472 and four dummy values (always zero).
4475 is the number of inodes the system has allocated.
4476 .\" This can be slightly more than
4478 .\" because Linux allocates them one page full at a time.
4480 represents the number of free inodes.
4487 and the system needs to prune the inode list instead of allocating more;
4488 since Linux 2.4, this field is a dummy value (always zero).
4490 .IR /proc/sys/fs/inotify " (since Linux 2.6.13)"
4491 This directory contains files
4492 .IR max_queued_events ", " max_user_instances ", and " max_user_watches ,
4493 that can be used to limit the amount of kernel memory consumed by the
4496 For further details, see
4499 .I /proc/sys/fs/lease\-break\-time
4500 This file specifies the grace period that the kernel grants to a process
4501 holding a file lease
4503 after it has sent a signal to that process notifying it
4504 that another process is waiting to open the file.
4505 If the lease holder does not remove or downgrade the lease within
4506 this grace period, the kernel forcibly breaks the lease.
4508 .I /proc/sys/fs/leases\-enable
4509 This file can be used to enable or disable file leases
4511 on a system-wide basis.
4512 If this file contains the value 0, leases are disabled.
4513 A nonzero value enables leases.
4515 .IR /proc/sys/fs/mount\-max " (since Linux 4.9)"
4516 .\" commit d29216842a85c7970c536108e093963f02714498
4517 The value in this file specifies the maximum number of mounts that may exist
4518 in a mount namespace.
4519 The default value in this file is 100,000.
4521 .IR /proc/sys/fs/mqueue " (since Linux 2.6.6)"
4522 This directory contains files
4523 .IR msg_max ", " msgsize_max ", and " queues_max ,
4524 controlling the resources used by POSIX message queues.
4529 .IR /proc/sys/fs/nr_open " (since Linux 2.6.25)
4530 .\" commit 9cfe015aa424b3c003baba3841a60dd9b5ad319b
4531 This file imposes ceiling on the value to which the
4533 resource limit can be raised (see
4535 This ceiling is enforced for both unprivileged and privileged process.
4536 The default value in this file is 1048576.
4537 (Before Linux 2.6.25, the ceiling for
4539 was hard-coded to the same value.)
4541 .IR /proc/sys/fs/overflowgid " and " /proc/sys/fs/overflowuid
4543 allow you to change the value of the fixed UID and GID.
4544 The default is 65534.
4545 Some filesystems support only 16-bit UIDs and GIDs, although in Linux
4546 UIDs and GIDs are 32 bits.
4547 When one of these filesystems is mounted
4548 with writes enabled, any UID or GID that would exceed 65535 is translated
4549 to the overflow value before being written to disk.
4551 .IR /proc/sys/fs/pipe\-max\-size " (since Linux 2.6.35)"
4555 .IR /proc/sys/fs/pipe\-user\-pages\-hard " (since Linux 4.5)"
4559 .IR /proc/sys/fs/pipe\-user\-pages\-soft " (since Linux 4.5)"
4563 .IR /proc/sys/fs/protected_hardlinks " (since Linux 3.6)"
4564 .\" commit 800179c9b8a1e796e441674776d11cd4c05d61d7
4565 When the value in this file is 0,
4566 no restrictions are placed on the creation of hard links
4567 (i.e., this is the historical behavior before Linux 3.6).
4568 When the value in this file is 1,
4569 a hard link can be created to a target file
4570 only if one of the following conditions is true:
4573 The calling process has the
4575 capability in its user namespace
4576 and the file UID has a mapping in the namespace.
4578 The filesystem UID of the process creating the link matches
4579 the owner (UID) of the target file
4581 .BR credentials (7),
4582 a process's filesystem UID is normally the same as its effective UID).
4584 All of the following conditions are true:
4587 the target is a regular file;
4589 the target file does not have its set-user-ID mode bit enabled;
4591 the target file does not have both its set-group-ID and
4592 group-executable mode bits enabled; and
4594 the caller has permission to read and write the target file
4595 (either via the file's permissions mask or because it has
4596 suitable capabilities).
4600 The default value in this file is 0.
4601 Setting the value to 1
4602 prevents a longstanding class of security issues caused by
4603 hard-link-based time-of-check, time-of-use races,
4604 most commonly seen in world-writable directories such as
4606 The common method of exploiting this flaw
4607 is to cross privilege boundaries when following a given hard link
4608 (i.e., a root process follows a hard link created by another user).
4609 Additionally, on systems without separated partitions,
4610 this stops unauthorized users from "pinning" vulnerable set-user-ID and
4611 set-group-ID files against being upgraded by
4612 the administrator, or linking to special files.
4614 .IR /proc/sys/fs/protected_symlinks " (since Linux 3.6)"
4615 .\" commit 800179c9b8a1e796e441674776d11cd4c05d61d7
4616 When the value in this file is 0,
4617 no restrictions are placed on following symbolic links
4618 (i.e., this is the historical behavior before Linux 3.6).
4619 When the value in this file is 1, symbolic links are followed only
4620 in the following circumstances:
4623 the filesystem UID of the process following the link matches
4624 the owner (UID) of the symbolic link
4626 .BR credentials (7),
4627 a process's filesystem UID is normally the same as its effective UID);
4629 the link is not in a sticky world-writable directory; or
4631 the symbolic link and its parent directory have the same owner (UID)
4634 A system call that fails to follow a symbolic link
4635 because of the above restrictions returns the error
4640 The default value in this file is 0.
4641 Setting the value to 1 avoids a longstanding class of security issues
4642 based on time-of-check, time-of-use races when accessing symbolic links.
4644 .IR /proc/sys/fs/suid_dumpable " (since Linux 2.6.13)"
4645 .\" The following is based on text from Documentation/sysctl/kernel.txt
4646 The value in this file is assigned to a process's "dumpable" flag
4647 in the circumstances described in
4650 the value in this file determines whether core dump files are
4651 produced for set-user-ID or otherwise protected/tainted binaries.
4652 The "dumpable" setting also affects the ownership of files in a process's
4654 directory, as described above.
4656 Three different integer values can be specified:
4660 .\" In kernel source: SUID_DUMP_DISABLE
4661 This provides the traditional (pre-Linux 2.6.13) behavior.
4662 A core dump will not be produced for a process which has
4663 changed credentials (by calling
4666 or similar, or by executing a set-user-ID or set-group-ID program)
4667 or whose binary does not have read permission enabled.
4670 .\" In kernel source: SUID_DUMP_USER
4671 All processes dump core when possible.
4672 (Reasons why a process might nevertheless not dump core are described in
4674 The core dump is owned by the filesystem user ID of the dumping process
4675 and no security is applied.
4676 This is intended for system debugging situations only:
4677 this mode is insecure because it allows unprivileged users to
4678 examine the memory contents of privileged processes.
4680 \fI2\ ("suidsafe")\fP
4681 .\" In kernel source: SUID_DUMP_ROOT
4682 Any binary which normally would not be dumped (see "0" above)
4683 is dumped readable by root only.
4684 This allows the user to remove the core dump file but not to read it.
4685 For security reasons core dumps in this mode will not overwrite one
4686 another or other files.
4687 This mode is appropriate when administrators are
4688 attempting to debug problems in a normal environment.
4690 Additionally, since Linux 3.6,
4691 .\" 9520628e8ceb69fa9a4aee6b57f22675d9e1b709
4692 .I /proc/sys/kernel/core_pattern
4693 must either be an absolute pathname
4694 or a pipe command, as detailed in
4696 Warnings will be written to the kernel log if
4698 does not follow these rules, and no core dump will be produced.
4699 .\" 54b501992dd2a839e94e76aa392c392b55080ce8
4702 For details of the effect of a process's "dumpable" setting
4703 on ptrace access mode checking, see
4706 .I /proc/sys/fs/super\-max
4708 controls the maximum number of superblocks, and
4709 thus the maximum number of mounted filesystems the kernel
4711 You need increase only
4713 if you need to mount more filesystems than the current value in
4717 .I /proc/sys/fs/super\-nr
4719 contains the number of filesystems currently mounted.
4722 This directory contains files controlling a range of kernel parameters,
4725 .I /proc/sys/kernel/acct
4727 contains three numbers:
4732 If BSD-style process accounting is enabled, these values control
4734 If free space on filesystem where the log lives goes below
4736 percent, accounting suspends.
4737 If free space gets above
4739 percent, accounting resumes.
4742 how often the kernel checks the amount of free space (value is in
4744 Default values are 4, 2 and 30.
4745 That is, suspend accounting if 2% or less space is free; resume it
4746 if 4% or more space is free; consider information about amount of free space
4747 valid for 30 seconds.
4749 .IR /proc/sys/kernel/auto_msgmni " (Linux 2.6.27 to 3.18)"
4750 .\" commit 9eefe520c814f6f62c5d36a2ddcd3fb99dfdb30e (introduces feature)
4751 .\" commit 0050ee059f7fc86b1df2527aaa14ed5dc72f9973 (rendered redundant)
4752 From Linux 2.6.27 to 3.18,
4753 this file was used to control recomputing of the value in
4754 .IR /proc/sys/kernel/msgmni
4755 upon the addition or removal of memory or upon IPC namespace creation/removal.
4756 Echoing "1" into this file enabled
4758 automatic recomputing (and triggered a recomputation of
4760 based on the current amount of available memory and number of IPC namespaces).
4761 Echoing "0" disabled automatic recomputing.
4762 (Automatic recomputing was also disabled if a value was explicitly assigned to
4763 .IR /proc/sys/kernel/msgmni .)
4764 The default value in
4768 Since Linux 3.19, the content of this file has no effect (because
4770 .\" FIXME Must document the 3.19 'msgmni' changes.
4771 defaults to near the maximum value possible),
4772 and reads from this file always return the value "0".
4774 .IR /proc/sys/kernel/cap_last_cap " (since Linux 3.2)"
4776 .BR capabilities (7).
4778 .IR /proc/sys/kernel/cap\-bound " (from Linux 2.2 to 2.6.24)"
4779 This file holds the value of the kernel
4780 .I "capability bounding set"
4781 (expressed as a signed decimal number).
4782 This set is ANDed against the capabilities permitted to a process
4785 Starting with Linux 2.6.25,
4786 the system-wide capability bounding set disappeared,
4787 and was replaced by a per-thread bounding set; see
4788 .BR capabilities (7).
4790 .I /proc/sys/kernel/core_pattern
4794 .I /proc/sys/kernel/core_pipe_limit
4798 .I /proc/sys/kernel/core_uses_pid
4802 .I /proc/sys/kernel/ctrl\-alt\-del
4804 controls the handling of Ctrl-Alt-Del from the keyboard.
4805 When the value in this file is 0, Ctrl-Alt-Del is trapped and
4808 program to handle a graceful restart.
4809 When the value is greater than zero, Linux's reaction to a Vulcan
4810 Nerve Pinch (tm) will be an immediate reboot, without even
4811 syncing its dirty buffers.
4812 Note: when a program (like dosemu) has the keyboard in "raw"
4813 mode, the ctrl-alt-del is intercepted by the program before it
4814 ever reaches the kernel tty layer, and it's up to the program
4815 to decide what to do with it.
4817 .IR /proc/sys/kernel/dmesg_restrict " (since Linux 2.6.37)"
4818 The value in this file determines who can see kernel syslog contents.
4819 A value of 0 in this file imposes no restrictions.
4820 If the value is 1, only privileged users can read the kernel syslog.
4825 .\" commit 620f6e8e855d6d447688a5f67a4e176944a084e8
4828 capability may change the value in this file.
4830 .IR /proc/sys/kernel/domainname " and " /proc/sys/kernel/hostname
4831 can be used to set the NIS/YP domainname and the
4832 hostname of your box in exactly the same way as the commands
4840 .RB "#" " echo \(aqdarkstar\(aq > /proc/sys/kernel/hostname"
4841 .RB "#" " echo \(aqmydomain\(aq > /proc/sys/kernel/domainname"
4845 has the same effect as
4849 .RB "#" " hostname \(aqdarkstar\(aq"
4850 .RB "#" " domainname \(aqmydomain\(aq"
4854 Note, however, that the classic darkstar.frop.org has the
4855 hostname "darkstar" and DNS (Internet Domain Name Server)
4856 domainname "frop.org", not to be confused with the NIS (Network
4857 Information Service) or YP (Yellow Pages) domainname.
4859 domain names are in general different.
4860 For a detailed discussion
4865 .I /proc/sys/kernel/hotplug
4867 contains the pathname for the hotplug policy agent.
4868 The default value in this file is
4871 .\" Removed in commit 87f504e5c78b910b0c1d6ffb89bc95e492322c84 (tglx/history.git)
4872 .IR /proc/sys/kernel/htab\-reclaim " (before Linux 2.4.9.2)"
4873 (PowerPC only) If this file is set to a nonzero value,
4875 .\" removed in commit 1b483a6a7b2998e9c98ad985d7494b9b725bd228, before 2.6.28
4877 .IR Documentation/powerpc/ppc_htab.txt )
4879 each time the system hits the idle loop.
4881 .IR /proc/sys/kernel/keys/*
4882 This directory contains various files that define parameters and limits
4883 for the key-management facility.
4884 These files are described in
4887 .IR /proc/sys/kernel/kptr_restrict " (since Linux 2.6.38)"
4888 .\" 455cd5ab305c90ffc422dd2e0fb634730942b257
4889 The value in this file determines whether kernel addresses are exposed via
4891 files and other interfaces.
4892 A value of 0 in this file imposes no restrictions.
4893 If the value is 1, kernel pointers printed using the
4895 format specifier will be replaced with zeros unless the user has the
4898 If the value is 2, kernel pointers printed using the
4900 format specifier will be replaced with zeros regardless
4901 of the user's capabilities.
4902 The initial default value for this file was 1,
4903 but the default was changed
4904 .\" commit 411f05f123cbd7f8aa1edcae86970755a6e2a9d9
4905 to 0 in Linux 2.6.39.
4907 .\" commit 620f6e8e855d6d447688a5f67a4e176944a084e8
4910 capability can change the value in this file.
4912 .I /proc/sys/kernel/l2cr
4913 (PowerPC only) This file
4914 contains a flag that controls the L2 cache of G3 processor
4916 If 0, the cache is disabled.
4919 .I /proc/sys/kernel/modprobe
4920 This file contains the pathname for the kernel module loader.
4921 The default value is
4922 .IR /sbin/modprobe .
4923 The file is present only if the kernel is built with the
4926 in Linux 2.6.26 and earlier)
4928 It is described by the Linux kernel source file
4929 .I Documentation/kmod.txt
4930 (present only in kernel 2.4 and earlier).
4932 .IR /proc/sys/kernel/modules_disabled " (since Linux 2.6.31)"
4933 .\" 3d43321b7015387cfebbe26436d0e9d299162ea1
4934 .\" From Documentation/sysctl/kernel.txt
4935 A toggle value indicating if modules are allowed to be loaded
4936 in an otherwise modular kernel.
4937 This toggle defaults to off (0), but can be set true (1).
4938 Once true, modules can be neither loaded nor unloaded,
4939 and the toggle cannot be set back to false.
4940 The file is present only if the kernel is built with the
4944 .IR /proc/sys/kernel/msgmax " (since Linux 2.2)"
4946 a system-wide limit specifying the maximum number of bytes in
4947 a single message written on a System V message queue.
4949 .IR /proc/sys/kernel/msgmni " (since Linux 2.4)"
4950 This file defines the system-wide limit on the number of
4951 message queue identifiers.
4953 .IR /proc/sys/kernel/auto_msgmni .
4955 .IR /proc/sys/kernel/msgmnb " (since Linux 2.2)"
4956 This file defines a system-wide parameter used to initialize the
4958 setting for subsequently created message queues.
4961 setting specifies the maximum number of bytes that may be written to the
4964 .IR /proc/sys/kernel/ngroups_max " (since Linux 2.6.4)"
4965 This is a read-only file that displays the upper limit on the
4966 number of a process's group memberships.
4968 .IR /proc/sys/kernel/ns_last_pid " (since Linux 3.3)"
4970 .BR pid_namespaces (7).
4972 .IR /proc/sys/kernel/ostype " and " /proc/sys/kernel/osrelease
4977 .IR /proc/sys/kernel/overflowgid " and " /proc/sys/kernel/overflowuid
4978 These files duplicate the files
4979 .I /proc/sys/fs/overflowgid
4981 .IR /proc/sys/fs/overflowuid .
4983 .I /proc/sys/kernel/panic
4984 This file gives read/write access to the kernel variable
4986 If this is zero, the kernel will loop on a panic; if nonzero,
4987 it indicates that the kernel should autoreboot after this number
4990 software watchdog device driver, the recommended setting is 60.
4992 .IR /proc/sys/kernel/panic_on_oops " (since Linux 2.5.68)"
4993 This file controls the kernel's behavior when an oops
4994 or BUG is encountered.
4995 If this file contains 0, then the system
4996 tries to continue operation.
4997 If it contains 1, then the system
4998 delays a few seconds (to give klogd time to record the oops output)
5001 .I /proc/sys/kernel/panic
5002 file is also nonzero, then the machine will be rebooted.
5004 .IR /proc/sys/kernel/pid_max " (since Linux 2.5.34)"
5005 This file specifies the value at which PIDs wrap around
5006 (i.e., the value in this file is one greater than the maximum PID).
5007 PIDs greater than this value are not allocated;
5008 thus, the value in this file also acts as a system-wide limit
5009 on the total number of processes and threads.
5010 The default value for this file, 32768,
5011 results in the same range of PIDs as on earlier kernels.
5012 On 32-bit platforms, 32768 is the maximum value for
5016 can be set to any value up to 2^22
5017 .RB ( PID_MAX_LIMIT ,
5018 approximately 4 million).
5019 .\" Prior to 2.6.10, pid_max could also be raised above 32768 on 32-bit
5020 .\" platforms, but this broke /proc/[pid]
5021 .\" See http://marc.theaimsgroup.com/?l=linux-kernel&m=109513010926152&w=2
5023 .IR /proc/sys/kernel/powersave\-nap " (PowerPC only)"
5024 This file contains a flag.
5025 If set, Linux-PPC will use the "nap" mode of
5027 otherwise the "doze" mode will be used.
5029 .I /proc/sys/kernel/printk
5033 .IR /proc/sys/kernel/pty " (since Linux 2.6.4)"
5034 This directory contains two files relating to the number of UNIX 98
5035 pseudoterminals (see
5039 .I /proc/sys/kernel/pty/max
5040 This file defines the maximum number of pseudoterminals.
5041 .\" FIXME Document /proc/sys/kernel/pty/reserve
5042 .\" New in Linux 3.3
5043 .\" commit e9aba5158a80098447ff207a452a3418ae7ee386
5045 .I /proc/sys/kernel/pty/nr
5047 indicates how many pseudoterminals are currently in use.
5049 .I /proc/sys/kernel/random
5051 contains various parameters controlling the operation of the file
5055 for further information.
5057 .IR /proc/sys/kernel/random/uuid " (since Linux 2.4)"
5058 Each read from this read-only file returns a randomly generated 128-bit UUID,
5059 as a string in the standard UUID format.
5061 .IR /proc/sys/kernel/randomize_va_space " (since Linux 2.6.12)"
5062 .\" Some further details can be found in Documentation/sysctl/kernel.txt
5063 Select the address space layout randomization (ASLR) policy for the system
5064 (on architectures that support ASLR).
5065 Three values are supported for this file:
5069 This is the default for architectures that don't support ASLR,
5070 and when the kernel is booted with the
5074 Make the addresses of
5076 allocations, the stack, and the VDSO page randomized.
5077 Among other things, this means that shared libraries will be
5078 loaded at randomized addresses.
5079 The text segment of PIE-linked binaries will also be loaded
5080 at a randomized address.
5081 This value is the default if the kernel was configured with
5082 .BR CONFIG_COMPAT_BRK .
5084 (Since Linux 2.6.25)
5085 .\" commit c1d171a002942ea2d93b4fbd0c9583c56fce0772
5086 Also support heap randomization.
5087 This value is the default if the kernel was not configured with
5088 .BR CONFIG_COMPAT_BRK .
5091 .I /proc/sys/kernel/real\-root\-dev
5092 This file is documented in the Linux kernel source file
5093 .I Documentation/admin\-guide/initrd.rst
5094 .\" commit 9d85025b0418163fae079c9ba8f8445212de8568
5096 .I Documentation/initrd.txt
5099 .IR /proc/sys/kernel/reboot\-cmd " (Sparc only) "
5100 This file seems to be a way to give an argument to the SPARC
5101 ROM/Flash boot loader.
5102 Maybe to tell it what to do after
5105 .I /proc/sys/kernel/rtsig\-max
5106 (Only in kernels up to and including 2.6.7; see
5108 This file can be used to tune the maximum number
5109 of POSIX real-time (queued) signals that can be outstanding
5112 .I /proc/sys/kernel/rtsig\-nr
5113 (Only in kernels up to and including 2.6.7.)
5114 This file shows the number of POSIX real-time signals currently queued.
5116 .IR /proc/[pid]/sched_autogroup_enabled " (since Linux 2.6.38)"
5117 .\" commit 5091faa449ee0b7d73bc296a93bca9540fc51d0a
5121 .IR /proc/sys/kernel/sched_child_runs_first " (since Linux 2.6.23)"
5122 If this file contains the value zero, then, after a
5124 the parent is first scheduled on the CPU.
5125 If the file contains a nonzero value,
5126 then the child is scheduled first on the CPU.
5127 (Of course, on a multiprocessor system,
5128 the parent and the child might both immediately be scheduled on a CPU.)
5130 .IR /proc/sys/kernel/sched_rr_timeslice_ms " (since Linux 3.9)"
5132 .BR sched_rr_get_interval (2).
5134 .IR /proc/sys/kernel/sched_rt_period_us " (since Linux 2.6.25)"
5138 .IR /proc/sys/kernel/sched_rt_runtime_us " (since Linux 2.6.25)"
5142 .IR /proc/sys/kernel/seccomp " (since Linux 4.14)"
5143 .\" commit 8e5f1ad116df6b0de65eac458d5e7c318d1c05af
5144 This directory provides additional seccomp information and
5148 for further details.
5150 .IR /proc/sys/kernel/sem " (since Linux 2.4)"
5151 This file contains 4 numbers defining limits for System V IPC semaphores.
5152 These fields are, in order:
5155 The maximum semaphores per semaphore set.
5157 A system-wide limit on the number of semaphores in all semaphore sets.
5159 The maximum number of operations that may be specified in a
5163 A system-wide limit on the maximum number of semaphore identifiers.
5166 .I /proc/sys/kernel/sg\-big\-buff
5168 shows the size of the generic SCSI device (sg) buffer.
5169 You can't tune it just yet, but you could change it at
5170 compile time by editing
5171 .I include/scsi/sg.h
5175 However, there shouldn't be any reason to change this value.
5177 .IR /proc/sys/kernel/shm_rmid_forced " (since Linux 3.1)"
5178 .\" commit b34a6b1da371ed8af1221459a18c67970f7e3d53
5179 .\" See also Documentation/sysctl/kernel.txt
5180 If this file is set to 1, all System V shared memory segments will
5181 be marked for destruction as soon as the number of attached processes
5183 in other words, it is no longer possible to create shared memory segments
5184 that exist independently of any attached process.
5186 The effect is as though a
5189 is performed on all existing segments as well as all segments
5190 created in the future (until this file is reset to 0).
5191 Note that existing segments that are attached to no process will be
5192 immediately destroyed when this file is set to 1.
5193 Setting this option will also destroy segments that were created,
5195 upon termination of the process that created the segment with
5198 Setting this file to 1 provides a way of ensuring that
5199 all System V shared memory segments are counted against the
5200 resource usage and resource limits (see the description of
5204 of at least one process.
5206 Because setting this file to 1 produces behavior that is nonstandard
5207 and could also break existing applications,
5208 the default value in this file is 0.
5209 Set this file to 1 only if you have a good understanding
5210 of the semantics of the applications using
5211 System V shared memory on your system.
5213 .IR /proc/sys/kernel/shmall " (since Linux 2.2)"
5215 contains the system-wide limit on the total number of pages of
5216 System V shared memory.
5218 .IR /proc/sys/kernel/shmmax " (since Linux 2.2)"
5220 can be used to query and set the run-time limit
5221 on the maximum (System V IPC) shared memory segment size that can be
5223 Shared memory segments up to 1GB are now supported in the
5225 This value defaults to
5228 .IR /proc/sys/kernel/shmmni " (since Linux 2.4)"
5230 specifies the system-wide maximum number of System V shared memory
5231 segments that can be created.
5233 .IR /proc/sys/kernel/sysctl_writes_strict " (since Linux 3.16)"
5234 .\" commit f88083005ab319abba5d0b2e4e997558245493c8
5235 .\" commit 2ca9bb456ada8bcbdc8f77f8fc78207653bbaa92
5236 .\" commit f4aacea2f5d1a5f7e3154e967d70cf3f711bcd61
5237 .\" commit 24fe831c17ab8149413874f2fd4e5c8a41fcd294
5238 The value in this file determines how the file offset affects
5239 the behavior of updating entries in files under
5241 The file has three possible values:
5245 This provides legacy handling, with no printk warnings.
5248 must fully contain the value to be written,
5249 and multiple writes on the same file descriptor
5250 will overwrite the entire value, regardless of the file position.
5253 (default) This provides the same behavior as for \-1,
5254 but printk warnings are written for processes that
5255 perform writes when the file offset is not 0.
5258 Respect the file offset when writing strings into
5261 Multiple writes will
5263 to the value buffer.
5264 Anything written beyond the maximum length
5265 of the value buffer will be ignored.
5268 entries must always be at file offset 0 and the value must be
5269 fully contained in the buffer provided to
5272 .\" With /proc/sys/kernel/sysctl_writes_strict==1, writes at an
5273 .\" offset other than 0 do not generate an error. Instead, the
5274 .\" write() succeeds, but the file is left unmodified.
5275 .\" This is surprising. The behavior may change in the future.
5276 .\" See thread.gmane.org/gmane.linux.man/9197
5277 .\" From: Michael Kerrisk (man-pages <mtk.manpages@...>
5278 .\" Subject: sysctl_writes_strict documentation + an oddity?
5279 .\" Newsgroups: gmane.linux.man, gmane.linux.kernel
5280 .\" Date: 2015-05-09 08:54:11 GMT
5283 .I /proc/sys/kernel/sysrq
5284 This file controls the functions allowed to be invoked by the SysRq key.
5286 the file contains 1 meaning that every possible SysRq request is allowed
5287 (in older kernel versions, SysRq was disabled by default,
5288 and you were required to specifically enable it at run-time,
5289 but this is not the case any more).
5290 Possible values in this file are:
5294 Disable sysrq completely
5297 Enable all functions of sysrq
5300 Bit mask of allowed sysrq functions, as follows:
5305 Enable control of console logging level
5308 Enable control of keyboard (SAK, unraw)
5311 Enable debugging dumps of processes etc.
5317 Enable remount read-only
5320 Enable signaling of processes (term, kill, oom-kill)
5323 Allow reboot/poweroff
5326 Allow nicing of all real-time tasks
5331 This file is present only if the
5332 .B CONFIG_MAGIC_SYSRQ
5333 kernel configuration option is enabled.
5334 For further details see the Linux kernel source file
5335 .I Documentation/admin\-guide/sysrq.rst
5336 .\" commit 9d85025b0418163fae079c9ba8f8445212de8568
5338 .I Documentation/sysrq.txt
5341 .I /proc/sys/kernel/version
5342 This file contains a string such as:
5344 #5 Wed Feb 25 21:49:24 MET 1998
5347 this is the fifth kernel built from this source base and the
5348 date following it indicates the time the kernel was built.
5350 .IR /proc/sys/kernel/threads\-max " (since Linux 2.3.11)"
5351 .\" The following is based on Documentation/sysctl/kernel.txt
5352 This file specifies the system-wide limit on the number of
5353 threads (tasks) that can be created on the system.
5356 .\" commit 230633d109e35b0a24277498e773edeb79b4a331
5357 the value that can be written to
5360 The minimum value that can be written is 20.
5361 The maximum value that can be written is given by the
5365 If a value outside of this range is written to
5371 The value written is checked against the available RAM pages.
5372 If the thread structures would occupy too much (more than 1/8th)
5373 of the available RAM pages,
5375 is reduced accordingly.
5377 .IR /proc/sys/kernel/yama/ptrace_scope " (since Linux 3.5)"
5381 .IR /proc/sys/kernel/zero\-paged " (PowerPC only) "
5384 When enabled (nonzero), Linux-PPC will pre-zero pages in
5385 the idle loop, possibly speeding up get_free_pages.
5388 This directory contains networking stuff.
5389 Explanations for some of the files under this directory can be found in
5394 .I /proc/sys/net/core/bpf_jit_enable
5398 .I /proc/sys/net/core/somaxconn
5399 This file defines a ceiling value for the
5405 manual page for details.
5408 This directory may be empty.
5411 This directory supports Sun remote procedure call for network filesystem
5413 On some systems, it is not present.
5415 .IR /proc/sys/user " (since Linux 4.9)"
5420 This directory contains files for memory management tuning, buffer and
5423 .IR /proc/sys/vm/admin_reserve_kbytes " (since Linux 3.10)"
5424 .\" commit 4eeab4f5580d11bffedc697684b91b0bca0d5009
5425 This file defines the amount of free memory (in KiB) on the system that
5426 should be reserved for users with the capability
5429 The default value in this file is the minimum of [3% of free pages, 8MiB]
5431 The default is intended to provide enough for the superuser
5432 to log in and kill a process, if necessary,
5433 under the default overcommit 'guess' mode (i.e., 0 in
5434 .IR /proc/sys/vm/overcommit_memory ).
5436 Systems running in "overcommit never" mode (i.e., 2 in
5437 .IR /proc/sys/vm/overcommit_memory )
5438 should increase the value in this file to account
5439 for the full virtual memory size of the programs used to recover (e.g.,
5444 Otherwise, the superuser may not be able to log in to recover the system.
5445 For example, on x86-64 a suitable value is 131072 (128MiB reserved).
5447 Changing the value in this file takes effect whenever
5448 an application requests memory.
5450 .IR /proc/sys/vm/compact_memory " (since Linux 2.6.35)"
5451 When 1 is written to this file, all zones are compacted such that free
5452 memory is available in contiguous blocks where possible.
5453 The effect of this action can be seen by examining
5454 .IR /proc/buddyinfo .
5456 Present only if the kernel was configured with
5457 .BR CONFIG_COMPACTION .
5459 .IR /proc/sys/vm/drop_caches " (since Linux 2.6.16)"
5460 Writing to this file causes the kernel to drop clean caches, dentries, and
5461 inodes from memory, causing that memory to become free.
5462 This can be useful for memory management testing and
5463 performing reproducible filesystem benchmarks.
5464 Because writing to this file causes the benefits of caching to be lost,
5465 it can degrade overall system performance.
5467 To free pagecache, use:
5469 echo 1 > /proc/sys/vm/drop_caches
5471 To free dentries and inodes, use:
5473 echo 2 > /proc/sys/vm/drop_caches
5475 To free pagecache, dentries and inodes, use:
5477 echo 3 > /proc/sys/vm/drop_caches
5479 Because writing to this file is a nondestructive operation and dirty objects
5480 are not freeable, the
5485 .IR /proc/sys/vm/legacy_va_layout " (since Linux 2.6.9)"
5486 .\" The following is from Documentation/filesystems/proc.txt
5487 If nonzero, this disables the new 32-bit memory-mapping layout;
5488 the kernel will use the legacy (2.4) layout for all processes.
5490 .IR /proc/sys/vm/memory_failure_early_kill " (since Linux 2.6.32)"
5491 .\" The following is based on the text in Documentation/sysctl/vm.txt
5492 Control how to kill processes when an uncorrected memory error
5493 (typically a 2-bit error in a memory module)
5494 that cannot be handled by the kernel
5495 is detected in the background by hardware.
5496 In some cases (like the page still having a valid copy on disk),
5497 the kernel will handle the failure
5498 transparently without affecting any applications.
5499 But if there is no other up-to-date copy of the data,
5500 it will kill processes to prevent any data corruptions from propagating.
5502 The file has one of the following values:
5505 Kill all processes that have the corrupted-and-not-reloadable page mapped
5506 as soon as the corruption is detected.
5507 Note that this is not supported for a few types of pages,
5508 such as kernel internally
5509 allocated data or the swap cache, but works for the majority of user pages.
5511 Unmap the corrupted page from all processes and kill a process
5512 only if it tries to access the page.
5515 The kill is performed using a
5521 Processes can handle this if they want to; see
5525 This feature is active only on architectures/platforms with advanced machine
5526 check handling and depends on the hardware capabilities.
5528 Applications can override the
5529 .I memory_failure_early_kill
5530 setting individually with the
5535 Present only if the kernel was configured with
5536 .BR CONFIG_MEMORY_FAILURE .
5538 .IR /proc/sys/vm/memory_failure_recovery " (since Linux 2.6.32)"
5539 .\" The following is based on the text in Documentation/sysctl/vm.txt
5540 Enable memory failure recovery (when supported by the platform)
5545 Always panic on a memory failure.
5548 Present only if the kernel was configured with
5549 .BR CONFIG_MEMORY_FAILURE .
5551 .IR /proc/sys/vm/oom_dump_tasks " (since Linux 2.6.25)"
5552 .\" The following is from Documentation/sysctl/vm.txt
5553 Enables a system-wide task dump (excluding kernel threads) to be
5554 produced when the kernel performs an OOM-killing.
5555 The dump includes the following information
5556 for each task (thread, process):
5557 thread ID, real user ID, thread group ID (process ID),
5558 virtual memory size, resident set size,
5559 the CPU that the task is scheduled on,
5560 oom_adj score (see the description of
5561 .IR /proc/[pid]/oom_adj ),
5563 This is helpful to determine why the OOM-killer was invoked
5564 and to identify the rogue task that caused it.
5566 If this contains the value zero, this information is suppressed.
5567 On very large systems with thousands of tasks,
5568 it may not be feasible to dump the memory state information for each one.
5569 Such systems should not be forced to incur a performance penalty in
5570 OOM situations when the information may not be desired.
5572 If this is set to nonzero, this information is shown whenever the
5573 OOM-killer actually kills a memory-hogging task.
5575 The default value is 0.
5577 .IR /proc/sys/vm/oom_kill_allocating_task " (since Linux 2.6.24)"
5578 .\" The following is from Documentation/sysctl/vm.txt
5579 This enables or disables killing the OOM-triggering task in
5580 out-of-memory situations.
5582 If this is set to zero, the OOM-killer will scan through the entire
5583 tasklist and select a task based on heuristics to kill.
5584 This normally selects a rogue memory-hogging task that
5585 frees up a large amount of memory when killed.
5587 If this is set to nonzero, the OOM-killer simply kills the task that
5588 triggered the out-of-memory condition.
5589 This avoids a possibly expensive tasklist scan.
5592 .I /proc/sys/vm/panic_on_oom
5593 is nonzero, it takes precedence over whatever value is used in
5594 .IR /proc/sys/vm/oom_kill_allocating_task .
5596 The default value is 0.
5598 .IR /proc/sys/vm/overcommit_kbytes " (since Linux 3.14)"
5599 .\" commit 49f0ce5f92321cdcf741e35f385669a421013cb7
5600 This writable file provides an alternative to
5601 .IR /proc/sys/vm/overcommit_ratio
5605 .IR /proc/sys/vm/overcommit_memory
5607 It allows the amount of memory overcommitting to be specified as
5608 an absolute value (in kB),
5609 rather than as a percentage, as is done with
5610 .IR overcommit_ratio .
5611 This allows for finer-grained control of
5613 on systems with extremely large memory sizes.
5616 .IR overcommit_kbytes
5618 .IR overcommit_ratio
5621 .IR overcommit_kbytes
5622 has a nonzero value, then it is used to calculate
5625 .IR overcommit_ratio
5627 Writing a value to either of these files causes the
5628 value in the other file to be set to zero.
5630 .I /proc/sys/vm/overcommit_memory
5631 This file contains the kernel virtual memory accounting mode.
5635 0: heuristic overcommit (this is the default)
5637 1: always overcommit, never check
5639 2: always check, never overcommit
5646 are not checked, and the default check is very weak,
5647 leading to the risk of getting a process "OOM-killed".
5649 In mode 1, the kernel pretends there is always enough memory,
5650 until memory actually runs out.
5651 One use case for this mode is scientific computing applications
5652 that employ large sparse arrays.
5653 In Linux kernel versions before 2.6.0, any nonzero value implies mode 1.
5655 In mode 2 (available since Linux 2.6), the total virtual address space
5656 that can be allocated
5662 CommitLimit = (total_RAM \- total_huge_TLB) *
5663 overcommit_ratio / 100 + total_swap
5669 is the total amount of RAM on the system;
5672 is the amount of memory set aside for huge pages;
5676 .IR /proc/sys/vm/overcommit_ratio ;
5680 is the amount of swap space.
5683 For example, on a system with 16GB of physical RAM, 16GB
5684 of swap, no space dedicated to huge pages, and an
5686 of 50, this formula yields a
5690 Since Linux 3.14, if the value in
5691 .I /proc/sys/vm/overcommit_kbytes
5694 is instead calculated as:
5696 CommitLimit = overcommit_kbytes + total_swap
5698 See also the description of
5699 .IR /proc/sys/vm/admin_reserve_kbytes
5701 .IR /proc/sys/vm/user_reserve_kbytes .
5703 .IR /proc/sys/vm/overcommit_ratio " (since Linux 2.6.0)"
5704 This writable file defines a percentage by which memory
5705 can be overcommitted.
5706 The default value in the file is 50.
5707 See the description of
5708 .IR /proc/sys/vm/overcommit_memory .
5710 .IR /proc/sys/vm/panic_on_oom " (since Linux 2.6.18)"
5711 .\" The following is adapted from Documentation/sysctl/vm.txt
5712 This enables or disables a kernel panic in
5713 an out-of-memory situation.
5715 If this file is set to the value 0,
5716 the kernel's OOM-killer will kill some rogue process.
5717 Usually, the OOM-killer is able to kill a rogue process and the
5718 system will survive.
5720 If this file is set to the value 1,
5721 then the kernel normally panics when out-of-memory happens.
5722 However, if a process limits allocations to certain nodes
5723 using memory policies
5728 and those nodes reach memory exhaustion status,
5729 one process may be killed by the OOM-killer.
5730 No panic occurs in this case:
5731 because other nodes' memory may be free,
5732 this means the system as a whole may not have reached
5733 an out-of-memory situation yet.
5735 If this file is set to the value 2,
5736 the kernel always panics when an out-of-memory condition occurs.
5738 The default value is 0.
5739 1 and 2 are for failover of clustering.
5740 Select either according to your policy of failover.
5742 .IR /proc/sys/vm/swappiness
5743 .\" The following is from Documentation/sysctl/vm.txt
5744 The value in this file controls how aggressively the kernel will swap
5746 Higher values increase aggressiveness, lower values
5747 decrease aggressiveness.
5748 The default value is 60.
5750 .IR /proc/sys/vm/user_reserve_kbytes " (since Linux 3.10)"
5751 .\" commit c9b1d0981fcce3d9976d7b7a56e4e0503bc610dd
5752 Specifies an amount of memory (in KiB) to reserve for user processes,
5753 This is intended to prevent a user from starting a single memory hogging
5754 process, such that they cannot recover (kill the hog).
5755 The value in this file has an effect only when
5756 .IR /proc/sys/vm/overcommit_memory
5757 is set to 2 ("overcommit never" mode).
5758 In this case, the system reserves an amount of memory that is the minimum
5759 of [3% of current process size,
5760 .IR user_reserve_kbytes ].
5762 The default value in this file is the minimum of [3% of free pages, 128MiB]
5765 If the value in this file is set to zero,
5766 then a user will be allowed to allocate all free memory with a single process
5767 (minus the amount reserved by
5768 .IR /proc/sys/vm/admin_reserve_kbytes ).
5769 Any subsequent attempts to execute a command will result in
5770 "fork: Cannot allocate memory".
5772 Changing the value in this file takes effect whenever
5773 an application requests memory.
5775 .IR /proc/sys/vm/unprivileged_userfaultfd " (since Linux 5.2)"
5776 .\" cefdca0a86be517bc390fc4541e3674b8e7803b0
5777 This (writable) file exposes a flag that controls whether
5778 unprivileged processes are allowed to employ
5779 .BR userfaultfd (2).
5780 If this file has the value 1, then unprivileged processes may use
5781 .BR userfaultfd (2).
5782 If this file has the value 0, then only processes that have the
5784 capability may employ
5785 .BR userfaultfd (2).
5786 The default value in this file is 1.
5788 .IR /proc/sysrq\-trigger " (since Linux 2.4.21)"
5789 Writing a character to this file triggers the same SysRq function as
5790 typing ALT-SysRq-<character> (see the description of
5791 .IR /proc/sys/kernel/sysrq ).
5792 This file is normally writable only by
5794 For further details see the Linux kernel source file
5795 .I Documentation/admin\-guide/sysrq.rst
5796 .\" commit 9d85025b0418163fae079c9ba8f8445212de8568
5798 .I Documentation/sysrq.txt
5802 Subdirectory containing the pseudo-files
5803 .IR msg ", " sem " and " shm "."
5804 These files list the System V Interprocess Communication (IPC) objects
5805 (respectively: message queues, semaphores, and shared memory)
5806 that currently exist on the system,
5807 providing similar information to that available via
5809 These files have headers and are formatted (one IPC object per line)
5810 for easy understanding.
5812 provides further background on the information shown by these files.
5814 .IR /proc/thread\-self " (since Linux 3.17)"
5815 .\" commit 0097875bd41528922fb3bb5f348c53f17e00e2fd
5816 This directory refers to the thread accessing the
5819 and is identical to the
5820 .I /proc/self/task/[tid]
5821 directory named by the process thread ID
5825 .IR /proc/timer_list " (since Linux 2.6.21)"
5826 .\" commit 289f480af87e45f7a6de6ba9b4c061c2e259fe98
5827 This read-only file exposes a list of all currently pending
5828 (high-resolution) timers,
5829 all clock-event sources, and their parameters in a human-readable form.
5831 .IR /proc/timer_stats " (from Linux 2.6.21 until Linux 4.10)"
5832 .\" commit 82f67cd9fca8c8762c15ba7ed0d5747588c1e221
5833 .\" Date: Fri Feb 16 01:28:13 2007 -0800
5834 .\" Text largely derived from Documentation/timers/timer_stats.txt
5835 .\" removed in commit dfb4357da6ddbdf57d583ba64361c9d792b0e0b1
5836 .\" Date: Wed Feb 8 11:26:59 2017 -0800
5837 This is a debugging facility to make timer (ab)use in a Linux
5838 system visible to kernel and user-space developers.
5839 It can be used by kernel and user-space developers to verify that
5840 their code does not make undue use of timers.
5841 The goal is to avoid unnecessary wakeups,
5842 thereby optimizing power consumption.
5844 If enabled in the kernel
5845 .RB ( CONFIG_TIMER_STATS ),
5847 it has almost zero run-time overhead and a relatively small
5848 data-structure overhead.
5849 Even if collection is enabled at run time, overhead is low:
5850 all the locking is per-CPU and lookup is hashed.
5853 .I /proc/timer_stats
5854 file is used both to control sampling facility and to read out the
5855 sampled information.
5859 functionality is inactive on bootup.
5860 A sampling period can be started using the following command:
5864 # echo 1 > /proc/timer_stats
5868 The following command stops a sampling period:
5872 # echo 0 > /proc/timer_stats
5876 The statistics can be retrieved by:
5880 $ cat /proc/timer_stats
5884 While sampling is enabled, each readout from
5885 .I /proc/timer_stats
5887 newly updated statistics.
5888 Once sampling is disabled, the sampled information
5889 is kept until a new sample period is started.
5890 This allows multiple readouts.
5893 .IR /proc/timer_stats :
5897 .RB $ " cat /proc/timer_stats"
5898 Timer Stats Version: v0.3
5899 Sample period: 1.764 s
5901 255, 0 swapper/3 hrtimer_start_range_ns (tick_sched_timer)
5902 71, 0 swapper/1 hrtimer_start_range_ns (tick_sched_timer)
5903 58, 0 swapper/0 hrtimer_start_range_ns (tick_sched_timer)
5904 4, 1694 gnome\-shell mod_delayed_work_on (delayed_work_timer_fn)
5905 17, 7 rcu_sched rcu_gp_kthread (process_timeout)
5907 1, 4911 kworker/u16:0 mod_delayed_work_on (delayed_work_timer_fn)
5908 1D, 2522 kworker/0:0 queue_delayed_work_on (delayed_work_timer_fn)
5909 1029 total events, 583.333 events/sec
5913 The output columns are:
5916 a count of the number of events,
5917 optionally (since Linux 2.6.23) followed by the letter \(aqD\(aq
5918 .\" commit c5c061b8f9726bc2c25e19dec227933a13d1e6b7 deferrable timers
5919 if this is a deferrable timer;
5921 the PID of the process that initialized the timer;
5923 the name of the process that initialized the timer;
5925 the function where the timer was initialized; and
5928 the callback function that is associated with the timer.
5931 During the Linux 4.11 development cycle,
5932 this file was removed because of security concerns,
5933 as it exposes information across namespaces.
5934 Furthermore, it is possible to obtain
5935 the same information via in-kernel tracing facilities such as ftrace.
5938 Subdirectory containing the pseudo-files and subdirectories for
5939 tty drivers and line disciplines.
5942 This file contains two numbers (values in seconds): the uptime of the
5943 system (including time spent in suspend) and the amount of time spent
5944 in the idle process.
5947 This string identifies the kernel version that is currently running.
5948 It includes the contents of
5949 .IR /proc/sys/kernel/ostype ,
5950 .I /proc/sys/kernel/osrelease
5952 .IR /proc/sys/kernel/version .
5957 Linux version 1.0.9 (quinlan@phaze) #1 Sat May 14 01:51:54 EDT 1994
5960 .\" FIXME 2.6.13 seems to have /proc/vmcore implemented; document this
5961 .\" See Documentation/kdump/kdump.txt
5962 .\" commit 666bfddbe8b8fd4fd44617d6c55193d5ac7edb29
5963 .\" Needs CONFIG_VMCORE
5966 .IR /proc/vmstat " (since Linux 2.6.0)"
5967 This file displays various virtual memory statistics.
5968 Each line of this file contains a single name-value pair,
5969 delimited by white space.
5970 Some lines are present only if the kernel was configured with
5972 (In some cases, the options required for particular files have changed
5973 across kernel versions, so they are not listed here.
5974 Details can be found by consulting the kernel source code.)
5975 The following fields may be present:
5976 .\" FIXME We need explanations for each of the following fields...
5979 .IR nr_free_pages " (since Linux 2.6.31)"
5980 .\" commit d23ad42324cc4378132e51f2fc5c9ba6cbe75182
5982 .IR nr_alloc_batch " (since Linux 3.12)"
5983 .\" commit 81c0a2bb515fd4daae8cab64352877480792b515
5985 .IR nr_inactive_anon " (since Linux 2.6.28)"
5986 .\" commit 4f98a2fee8acdb4ac84545df98cccecfd130f8db
5988 .IR nr_active_anon " (since Linux 2.6.28)"
5989 .\" commit 4f98a2fee8acdb4ac84545df98cccecfd130f8db
5991 .IR nr_inactive_file " (since Linux 2.6.28)"
5992 .\" commit 4f98a2fee8acdb4ac84545df98cccecfd130f8db
5994 .IR nr_active_file " (since Linux 2.6.28)"
5995 .\" commit 4f98a2fee8acdb4ac84545df98cccecfd130f8db
5997 .IR nr_unevictable " (since Linux 2.6.28)"
5998 .\" commit 7b854121eb3e5ba0241882ff939e2c485228c9c5
6000 .IR nr_mlock " (since Linux 2.6.28)"
6001 .\" commit 5344b7e648980cc2ca613ec03a56a8222ff48820
6003 .IR nr_anon_pages " (since Linux 2.6.18)"
6004 .\" commit f3dbd34460ff54962d3e3244b6bcb7f5295356e6
6006 .IR nr_mapped " (since Linux 2.6.0)"
6008 .IR nr_file_pages " (since Linux 2.6.18)"
6009 .\" commit 347ce434d57da80fd5809c0c836f206a50999c26
6011 .IR nr_dirty " (since Linux 2.6.0)"
6013 .IR nr_writeback " (since Linux 2.6.0)"
6015 .IR nr_slab_reclaimable " (since Linux 2.6.19)"
6016 .\" commit 972d1a7b140569084439a81265a0f15b74e924e0
6017 .\" Linux 2.6.0 had nr_slab
6019 .IR nr_slab_unreclaimable " (since Linux 2.6.19)"
6020 .\" commit 972d1a7b140569084439a81265a0f15b74e924e0
6022 .IR nr_page_table_pages " (since Linux 2.6.0)"
6024 .IR nr_kernel_stack " (since Linux 2.6.32)"
6025 .\" commit c6a7f5728a1db45d30df55a01adc130b4ab0327c
6026 Amount of memory allocated to kernel stacks.
6028 .IR nr_unstable " (since Linux 2.6.0)"
6030 .IR nr_bounce " (since Linux 2.6.12)"
6031 .\" commit edfbe2b0038723e5699ab22695ccd62b5542a5c1
6033 .IR nr_vmscan_write " (since Linux 2.6.19)"
6034 .\" commit e129b5c23c2b471d47f1c5d2b8b193fc2034af43
6036 .IR nr_vmscan_immediate_reclaim " (since Linux 3.2)"
6037 .\" commit 49ea7eb65e7c5060807fb9312b1ad4c3eab82e2c
6039 .IR nr_writeback_temp " (since Linux 2.6.26)"
6040 .\" commit fc3ba692a4d19019387c5acaea63131f9eab05dd
6042 .IR nr_isolated_anon " (since Linux 2.6.32)"
6043 .\" commit a731286de62294b63d8ceb3c5914ac52cc17e690
6045 .IR nr_isolated_file " (since Linux 2.6.32)"
6046 .\" commit a731286de62294b63d8ceb3c5914ac52cc17e690
6048 .IR nr_shmem " (since Linux 2.6.32)"
6049 .\" commit 4b02108ac1b3354a22b0d83c684797692efdc395
6050 Pages used by shmem and
6053 .IR nr_dirtied " (since Linux 2.6.37)"
6054 .\" commit ea941f0e2a8c02ae876cd73deb4e1557248f258c
6056 .IR nr_written " (since Linux 2.6.37)"
6057 .\" commit ea941f0e2a8c02ae876cd73deb4e1557248f258c
6059 .IR nr_pages_scanned " (since Linux 3.17)"
6060 .\" commit 0d5d823ab4e608ec7b52ac4410de4cb74bbe0edd
6062 .IR numa_hit " (since Linux 2.6.18)"
6063 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6064 .\" Present only if the kernel was configured with
6065 .\" .BR CONFIG_NUMA .
6067 .IR numa_miss " (since Linux 2.6.18)"
6068 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6069 .\" Present only if the kernel was configured with
6070 .\" .BR CONFIG_NUMA .
6072 .IR numa_foreign " (since Linux 2.6.18)"
6073 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6074 .\" Present only if the kernel was configured with
6075 .\" .BR CONFIG_NUMA .
6077 .IR numa_interleave " (since Linux 2.6.18)"
6078 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6079 .\" Present only if the kernel was configured with
6080 .\" .BR CONFIG_NUMA .
6082 .IR numa_local " (since Linux 2.6.18)"
6083 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6084 .\" Present only if the kernel was configured with
6085 .\" .BR CONFIG_NUMA .
6087 .IR numa_other " (since Linux 2.6.18)"
6088 .\" commit ca889e6c45e0b112cb2ca9d35afc66297519b5d5
6089 .\" Present only if the kernel was configured with
6090 .\" .BR CONFIG_NUMA .
6092 .IR workingset_refault " (since Linux 3.15)"
6093 .\" commit a528910e12ec7ee203095eb1711468a66b9b60b0
6094 .\" Present only if the kernel was configured with
6095 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6097 .IR workingset_activate " (since Linux 3.15)"
6098 .\" commit a528910e12ec7ee203095eb1711468a66b9b60b0
6099 .\" Present only if the kernel was configured with
6100 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6102 .IR workingset_nodereclaim " (since Linux 3.15)"
6103 .\" commit 449dd6984d0e47643c04c807f609dd56d48d5bcc
6104 .\" Present only if the kernel was configured with
6105 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6107 .IR nr_anon_transparent_hugepages " (since Linux 2.6.38)"
6108 .\" Present only if the kernel was configured with
6109 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6111 .IR nr_free_cma " (since Linux 3.7)"
6112 .\" commit d1ce749a0db12202b711d1aba1d29e823034648d
6113 Number of free CMA (Contiguous Memory Allocator) pages.
6114 .\" Present only if the kernel was configured with
6115 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6117 .IR nr_dirty_threshold " (since Linux 2.6.37)"
6118 .\" commit 79da826aee6a10902ef411bc65864bd02102fa83
6119 .\" Present only if the kernel was configured with
6120 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6122 .IR nr_dirty_background_threshold " (since Linux 2.6.37)"
6123 .\" commit 79da826aee6a10902ef411bc65864bd02102fa83
6124 .\" Present only if the kernel was configured with
6125 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6127 .IR pgpgin " (since Linux 2.6.0)"
6128 .\" Present only if the kernel was configured with
6129 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6131 .IR pgpgout " (since Linux 2.6.0)"
6132 .\" Present only if the kernel was configured with
6133 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6135 .IR pswpin " (since Linux 2.6.0)"
6136 .\" Present only if the kernel was configured with
6137 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6139 .IR pswpout " (since Linux 2.6.0)"
6140 .\" Present only if the kernel was configured with
6141 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6143 .IR pgalloc_dma " (since Linux 2.6.5)"
6144 .\" Linux 2.6.0 had pgalloc
6145 .\" Present only if the kernel was configured with
6146 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6148 .IR pgalloc_dma32 " (since Linux 2.6.16)"
6149 .\" commit 9328b8faae922e52073785ed6c1eaa8565648a0e
6150 .\" Present only if the kernel was configured with
6151 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6153 .IR pgalloc_normal " (since Linux 2.6.5)"
6154 .\" Present only if the kernel was configured with
6155 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6157 .IR pgalloc_high " (since Linux 2.6.5)"
6158 .\" Present only if the kernel was configured with
6159 .\" .BR CONFIG_VM_EVENT_COUNTERS
6161 .\" .BR CONFIG_HIGHMEM .
6163 .IR pgalloc_movable " (since Linux 2.6.23)"
6164 .\" commit 2a1e274acf0b1c192face19a4be7c12d4503eaaf
6165 .\" Present only if the kernel was configured with
6166 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6168 .IR pgfree " (since Linux 2.6.0)"
6169 .\" Present only if the kernel was configured with
6170 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6172 .IR pgactivate " (since Linux 2.6.0)"
6173 .\" Present only if the kernel was configured with
6174 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6176 .IR pgdeactivate " (since Linux 2.6.0)"
6177 .\" Present only if the kernel was configured with
6178 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6180 .IR pgfault " (since Linux 2.6.0)"
6181 .\" Present only if the kernel was configured with
6182 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6184 .IR pgmajfault " (since Linux 2.6.0)"
6185 .\" Present only if the kernel was configured with
6186 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6188 .IR pgrefill_dma " (since Linux 2.6.5)"
6189 .\" Linux 2.6.0 had pgrefill
6190 .\" Present only if the kernel was configured with
6191 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6193 .IR pgrefill_dma32 " (since Linux 2.6.16)"
6194 .\" commit 9328b8faae922e52073785ed6c1eaa8565648a0e
6195 .\" Present only if the kernel was configured with
6196 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6198 .IR pgrefill_normal " (since Linux 2.6.5)"
6199 .\" Present only if the kernel was configured with
6200 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6202 .IR pgrefill_high " (since Linux 2.6.5)"
6203 .\" Present only if the kernel was configured with
6204 .\" .BR CONFIG_VM_EVENT_COUNTERS
6206 .\" .BR CONFIG_HIGHMEM .
6208 .IR pgrefill_movable " (since Linux 2.6.23)"
6209 .\" commit 2a1e274acf0b1c192face19a4be7c12d4503eaaf
6210 .\" Present only if the kernel was configured with
6211 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6212 .\" Formerly there were
6217 .\" These were split out into pgsteal_kswapd* and pgsteal_direct*
6218 .\" in commit 904249aa68010c8e223263c922fcbb840a3f42e4
6220 .IR pgsteal_kswapd_dma " (since Linux 3.4)"
6221 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6222 .\" Linux 2.6.0 had pgsteal
6223 .\" Present only if the kernel was configured with
6224 .\" .\" .BR CONFIG_VM_EVENT_COUNTERS .
6226 .IR pgsteal_kswapd_dma32 " (since Linux 3.4)"
6227 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6228 .\" commit 9328b8faae922e52073785ed6c1eaa8565648a0e
6229 .\" Present only if the kernel was configured with
6230 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6232 .IR pgsteal_kswapd_normal " (since Linux 3.4)"
6233 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6234 .\" Present only if the kernel was configured with
6235 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6237 .IR pgsteal_kswapd_high " (since Linux 3.4)"
6238 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6239 .\" Present only if the kernel was configured with
6240 .\" .BR CONFIG_VM_EVENT_COUNTERS
6242 .\" .BR CONFIG_HIGHMEM .
6244 .IR pgsteal_kswapd_movable " (since Linux 3.4)"
6245 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6246 .\" Present only if the kernel was configured with
6247 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6249 .IR pgsteal_direct_dma
6250 .\" Present only if the kernel was configured with
6251 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6253 .IR pgsteal_direct_dma32 " (since Linux 3.4)"
6254 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6255 .\" Present only if the kernel was configured with
6256 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6258 .IR pgsteal_direct_normal " (since Linux 3.4)"
6259 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6260 .\" Present only if the kernel was configured with
6261 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6263 .IR pgsteal_direct_high " (since Linux 3.4)"
6264 .\" commit 904249aa68010c8e223263c922fcbb840a3f42e4
6265 .\" Present only if the kernel was configured with
6266 .\" .BR CONFIG_VM_EVENT_COUNTERS
6268 .\" .BR CONFIG_HIGHMEM .
6270 .IR pgsteal_direct_movable " (since Linux 2.6.23)"
6271 .\" commit 2a1e274acf0b1c192face19a4be7c12d4503eaaf
6272 .\" Present only if the kernel was configured with
6273 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6275 .IR pgscan_kswapd_dma
6276 .\" Linux 2.6.0 had pgscan
6277 .\" Present only if the kernel was configured with
6278 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6280 .IR pgscan_kswapd_dma32 " (since Linux 2.6.16)"
6281 .\" commit 9328b8faae922e52073785ed6c1eaa8565648a0e
6282 .\" Present only if the kernel was configured with
6283 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6285 .IR pgscan_kswapd_normal " (since Linux 2.6.5)"
6286 .\" Present only if the kernel was configured with
6287 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6289 .IR pgscan_kswapd_high
6290 .\" Present only if the kernel was configured with
6291 .\" .BR CONFIG_VM_EVENT_COUNTERS
6293 .\" .BR CONFIG_HIGHMEM .
6295 .IR pgscan_kswapd_movable " (since Linux 2.6.23)"
6296 .\" commit 2a1e274acf0b1c192face19a4be7c12d4503eaaf
6297 .\" Present only if the kernel was configured with
6298 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6300 .IR pgscan_direct_dma
6301 .\" Present only if the kernel was configured with
6302 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6304 .IR pgscan_direct_dma32 " (since Linux 2.6.16)"
6305 .\" commit 9328b8faae922e52073785ed6c1eaa8565648a0e
6306 .\" Present only if the kernel was configured with
6307 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6309 .IR pgscan_direct_normal
6310 .\" Present only if the kernel was configured with
6311 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6313 .IR pgscan_direct_high
6314 .\" Present only if the kernel was configured with
6315 .\" .BR CONFIG_VM_EVENT_COUNTERS
6317 .\" .BR CONFIG_HIGHMEM .
6319 .IR pgscan_direct_movable " (since Linux 2.6.23)"
6320 .\" commit 2a1e274acf0b1c192face19a4be7c12d4503eaaf
6321 .\" Present only if the kernel was configured with
6322 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6324 .IR pgscan_direct_throttle " (since Linux 3.6)"
6325 .\" commit 68243e76ee343d63c6cf76978588a885951e2818
6326 .\" Present only if the kernel was configured with
6327 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6329 .IR zone_reclaim_failed " (since linux 2.6.31)"
6330 .\" commit 24cf72518c79cdcda486ed26074ff8151291cf65
6331 .\" Present only if the kernel was configured with
6332 .\" .BR CONFIG_VM_EVENT_COUNTERS
6334 .\" .BR CONFIG_NUMA .
6336 .IR pginodesteal " (since linux 2.6.0)"
6337 .\" Present only if the kernel was configured with
6338 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6340 .IR slabs_scanned " (since linux 2.6.5)"
6341 .\" Present only if the kernel was configured with
6342 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6344 .IR kswapd_inodesteal " (since linux 2.6.0)"
6345 .\" Present only if the kernel was configured with
6346 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6348 .IR kswapd_low_wmark_hit_quickly " (since 2.6.33)"
6349 .\" commit bb3ab596832b920c703d1aea1ce76d69c0f71fb7
6350 .\" Present only if the kernel was configured with
6351 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6353 .IR kswapd_high_wmark_hit_quickly " (since 2.6.33)"
6354 .\" commit bb3ab596832b920c703d1aea1ce76d69c0f71fb7
6355 .\" Present only if the kernel was configured with
6356 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6358 .IR pageoutrun " (since Linux 2.6.0)"
6359 .\" Present only if the kernel was configured with
6360 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6362 .IR allocstall " (since Linux 2.6.0)"
6363 .\" Present only if the kernel was configured with
6364 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6366 .IR pgrotated " (since Linux 2.6.0)"
6367 .\" Present only if the kernel was configured with
6368 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6370 .IR drop_pagecache " (since Linux 3.15)"
6371 .\" commit 5509a5d27b971a90b940e148ca9ca53312e4fa7a
6372 .\" Present only if the kernel was configured with
6373 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6375 .IR drop_slab " (since Linux 3.15)"
6376 .\" commit 5509a5d27b971a90b940e148ca9ca53312e4fa7a
6377 .\" Present only if the kernel was configured with
6378 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6380 .IR numa_pte_updates " (since Linux 3.8)"
6381 .\" commit 03c5a6e16322c997bf8f264851bfa3f532ad515f
6382 .\" Present only if the kernel was configured with
6383 .\" .BR CONFIG_VM_EVENT_COUNTERS
6385 .\" .BR CONFIG_NUMA_BALANCING .
6387 .IR numa_huge_pte_updates " (since Linux 3.13)"
6388 .\" commit 72403b4a0fbdf433c1fe0127e49864658f6f6468
6389 .\" Present only if the kernel was configured with
6390 .\" .BR CONFIG_VM_EVENT_COUNTERS
6392 .\" .BR CONFIG_NUMA_BALANCING .
6394 .IR numa_hint_faults " (since Linux 3.8)"
6395 .\" commit 03c5a6e16322c997bf8f264851bfa3f532ad515f
6396 .\" Present only if the kernel was configured with
6397 .\" .BR CONFIG_VM_EVENT_COUNTERS
6399 .\" .BR CONFIG_NUMA_BALANCING .
6401 .IR numa_hint_faults_local " (since Linux 3.8)"
6402 .\" commit 03c5a6e16322c997bf8f264851bfa3f532ad515f
6403 .\" Present only if the kernel was configured with
6404 .\" .BR CONFIG_VM_EVENT_COUNTERS
6406 .\" .BR CONFIG_NUMA_BALANCING .
6408 .IR numa_pages_migrated " (since Linux 3.8)"
6409 .\" commit 03c5a6e16322c997bf8f264851bfa3f532ad515f
6410 .\" Present only if the kernel was configured with
6411 .\" .BR CONFIG_VM_EVENT_COUNTERS
6413 .\" .BR CONFIG_NUMA_BALANCING
6415 .\" .BR CONFIG_NUMA_BALANCING .
6417 .IR pgmigrate_success " (since Linux 3.8)"
6418 .\" commit 5647bc293ab15f66a7b1cda850c5e9d162a6c7c2
6419 .\" Present only if the kernel was configured with
6420 .\" .BR CONFIG_VM_EVENT_COUNTERS
6422 .\" .BR CONFIG_MIGRATION .
6424 .IR pgmigrate_fail " (since Linux 3.8)"
6425 .\" commit 5647bc293ab15f66a7b1cda850c5e9d162a6c7c2
6426 .\" Present only if the kernel was configured with
6427 .\" .BR CONFIG_VM_EVENT_COUNTERS
6429 .\" .BR CONFIG_MIGRATION .
6431 .IR compact_migrate_scanned " (since Linux 3.8)"
6432 .\" commit 397487db696cae0b026a474a5cd66f4e372995e6
6433 .\" Linux 3.8 dropped compact_blocks_moved, compact_pages_moved, and
6434 .\" compact_pagemigrate_failed
6435 .\" Present only if the kernel was configured with
6436 .\" .BR CONFIG_VM_EVENT_COUNTERS
6438 .\" .BR CONFIG_COMPACTION .
6440 .IR compact_free_scanned " (since Linux 3.8)"
6441 .\" commit 397487db696cae0b026a474a5cd66f4e372995e6
6442 .\" Present only if the kernel was configured with
6443 .\" .BR CONFIG_VM_EVENT_COUNTERS
6445 .\" .BR CONFIG_COMPACTION .
6447 .IR compact_isolated " (since Linux 3.8)"
6448 .\" commit 397487db696cae0b026a474a5cd66f4e372995e6
6449 .\" Present only if the kernel was configured with
6450 .\" .BR CONFIG_VM_EVENT_COUNTERS
6452 .\" .BR CONFIG_COMPACTION .
6454 .IR compact_stall " (since Linux 2.6.35)"
6455 .\" commit 56de7263fcf3eb10c8dcdf8d59a9cec831795f3f
6456 See the kernel source file
6457 .IR Documentation/admin\-guide/mm/transhuge.rst .
6458 .\" Present only if the kernel was configured with
6459 .\" .BR CONFIG_VM_EVENT_COUNTERS
6461 .\" .BR CONFIG_COMPACTION .
6463 .IR compact_fail " (since Linux 2.6.35)"
6464 .\" commit 56de7263fcf3eb10c8dcdf8d59a9cec831795f3f
6465 See the kernel source file
6466 .IR Documentation/admin\-guide/mm/transhuge.rst .
6467 .\" Present only if the kernel was configured with
6468 .\" .BR CONFIG_VM_EVENT_COUNTERS
6470 .\" .BR CONFIG_COMPACTION .
6472 .IR compact_success " (since Linux 2.6.35)"
6473 .\" commit 56de7263fcf3eb10c8dcdf8d59a9cec831795f3f
6474 See the kernel source file
6475 .IR Documentation/admin\-guide/mm/transhuge.rst .
6476 .\" Present only if the kernel was configured with
6477 .\" .BR CONFIG_VM_EVENT_COUNTERS
6479 .\" .BR CONFIG_COMPACTION .
6481 .IR htlb_buddy_alloc_success " (since Linux 2.6.26)"
6482 .\" commit 3b1163006332302117b1b2acf226d4014ff46525
6483 .\" Present only if the kernel was configured with
6484 .\" .BR CONFIG_VM_EVENT_COUNTERS
6486 .\" .BR CONFIG_HUGETLB_PAGE .
6488 .IR htlb_buddy_alloc_fail " (since Linux 2.6.26)"
6489 .\" commit 3b1163006332302117b1b2acf226d4014ff46525
6490 .\" Present only if the kernel was configured with
6491 .\" .BR CONFIG_VM_EVENT_COUNTERS
6493 .\" .BR CONFIG_HUGETLB_PAGE .
6495 .IR unevictable_pgs_culled " (since Linux 2.6.28)"
6496 .\" commit bbfd28eee9fbd73e780b19beb3dc562befbb94fa
6497 .\" Present only if the kernel was configured with
6498 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6500 .IR unevictable_pgs_scanned " (since Linux 2.6.28)"
6501 .\" commit bbfd28eee9fbd73e780b19beb3dc562befbb94fa
6502 .\" Present only if the kernel was configured with
6503 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6505 .IR unevictable_pgs_rescued " (since Linux 2.6.28)"
6506 .\" commit bbfd28eee9fbd73e780b19beb3dc562befbb94fa
6507 .\" Present only if the kernel was configured with
6508 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6510 .IR unevictable_pgs_mlocked " (since Linux 2.6.28)"
6511 .\" commit 5344b7e648980cc2ca613ec03a56a8222ff48820
6512 .\" Present only if the kernel was configured with
6513 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6515 .IR unevictable_pgs_munlocked " (since Linux 2.6.28)"
6516 .\" commit 5344b7e648980cc2ca613ec03a56a8222ff48820
6517 .\" Present only if the kernel was configured with
6518 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6520 .IR unevictable_pgs_cleared " (since Linux 2.6.28)"
6521 .\" commit 5344b7e648980cc2ca613ec03a56a8222ff48820
6522 .\" Present only if the kernel was configured with
6523 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6525 .IR unevictable_pgs_stranded " (since Linux 2.6.28)"
6526 .\" commit 5344b7e648980cc2ca613ec03a56a8222ff48820
6527 .\" Present only if the kernel was configured with
6528 .\" .BR CONFIG_VM_EVENT_COUNTERS .
6529 .\" Linux 3.7 removed unevictable_pgs_mlockfreed
6531 .IR thp_fault_alloc " (since Linux 2.6.39)"
6532 .\" commit 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145
6533 See the kernel source file
6534 .IR Documentation/admin\-guide/mm/transhuge.rst .
6535 .\" Present only if the kernel was configured with
6536 .\" .BR CONFIG_VM_EVENT_COUNTERS
6538 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6540 .IR thp_fault_fallback " (since Linux 2.6.39)"
6541 .\" commit 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145
6542 See the kernel source file
6543 .IR Documentation/admin\-guide/mm/transhuge.rst .
6544 .\" Present only if the kernel was configured with
6545 .\" .BR CONFIG_VM_EVENT_COUNTERS
6547 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6549 .IR thp_collapse_alloc " (since Linux 2.6.39)"
6550 .\" commit 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145
6551 See the kernel source file
6552 .IR Documentation/admin\-guide/mm/transhuge.rst .
6553 .\" Present only if the kernel was configured with
6554 .\" .BR CONFIG_VM_EVENT_COUNTERS
6556 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6558 .IR thp_collapse_alloc_failed " (since Linux 2.6.39)"
6559 .\" commit 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145
6560 See the kernel source file
6561 .IR Documentation/admin\-guide/mm/transhuge.rst .
6562 .\" Present only if the kernel was configured with
6563 .\" .BR CONFIG_VM_EVENT_COUNTERS
6565 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6567 .IR thp_split " (since Linux 2.6.39)"
6568 .\" commit 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145
6569 See the kernel source file
6570 .IR Documentation/admin\-guide/mm/transhuge.rst .
6571 .\" Present only if the kernel was configured with
6572 .\" .BR CONFIG_VM_EVENT_COUNTERS
6574 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6576 .IR thp_zero_page_alloc " (since Linux 3.8)"
6577 .\" commit d8a8e1f0da3d29d7268b3300c96a059d63901b76
6578 See the kernel source file
6579 .IR Documentation/admin\-guide/mm/transhuge.rst .
6580 .\" Present only if the kernel was configured with
6581 .\" .BR CONFIG_VM_EVENT_COUNTERS
6583 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6585 .IR thp_zero_page_alloc_failed " (since Linux 3.8)"
6586 .\" commit d8a8e1f0da3d29d7268b3300c96a059d63901b76
6587 See the kernel source file
6588 .IR Documentation/admin\-guide/mm/transhuge.rst .
6589 .\" Present only if the kernel was configured with
6590 .\" .BR CONFIG_VM_EVENT_COUNTERS
6592 .\" .BR CONFIG_TRANSPARENT_HUGEPAGE .
6594 .IR balloon_inflate " (since Linux 3.18)"
6595 .\" commit 09316c09dde33aae14f34489d9e3d243ec0d5938
6596 .\" Present only if the kernel was configured with
6597 .\" .BR CONFIG_VM_EVENT_COUNTERS
6599 .\" .BR CONFIG_MEMORY_BALLOON .
6601 .IR balloon_deflate " (since Linux 3.18)"
6602 .\" commit 09316c09dde33aae14f34489d9e3d243ec0d5938
6603 .\" Present only if the kernel was configured with
6604 .\" .BR CONFIG_VM_EVENT_COUNTERS
6606 .\" .BR CONFIG_MEMORY_BALLOON .
6608 .IR balloon_migrate " (since Linux 3.18)"
6609 .\" commit 09316c09dde33aae14f34489d9e3d243ec0d5938
6610 .\" Present only if the kernel was configured with
6611 .\" .BR CONFIG_VM_EVENT_COUNTERS ,
6612 .\" .BR CONFIG_MEMORY_BALLOON ,
6614 .\" .BR CONFIG_BALLOON_COMPACTION .
6616 .IR nr_tlb_remote_flush " (since Linux 3.12)"
6617 .\" commit 9824cf9753ecbe8f5b47aa9b2f218207defea211
6618 .\" Present only if the kernel was configured with
6619 .\" .BR CONFIG_DEBUG_TLBFLUSH
6621 .\" .BR CONFIG_SMP .
6623 .IR nr_tlb_remote_flush_received " (since Linux 3.12)"
6624 .\" commit 9824cf9753ecbe8f5b47aa9b2f218207defea211
6625 .\" Present only if the kernel was configured with
6626 .\" .BR CONFIG_DEBUG_TLBFLUSH
6628 .\" .BR CONFIG_SMP .
6630 .IR nr_tlb_local_flush_all " (since Linux 3.12)"
6631 .\" commit 9824cf9753ecbe8f5b47aa9b2f218207defea211
6632 .\" Present only if the kernel was configured with
6633 .\" .BR CONFIG_DEBUG_TLBFLUSH .
6635 .IR nr_tlb_local_flush_one " (since Linux 3.12)"
6636 .\" commit 9824cf9753ecbe8f5b47aa9b2f218207defea211
6637 .\" Present only if the kernel was configured with
6638 .\" .BR CONFIG_DEBUG_TLBFLUSH .
6640 .IR vmacache_find_calls " (since Linux 3.16)"
6641 .\" commit 4f115147ff802267d0aa41e361c5aa5bd933d896
6642 .\" Present only if the kernel was configured with
6643 .\" .BR CONFIG_DEBUG_VM_VMACACHE .
6645 .IR vmacache_find_hits " (since Linux 3.16)"
6646 .\" commit 4f115147ff802267d0aa41e361c5aa5bd933d896
6647 .\" Present only if the kernel was configured with
6648 .\" .BR CONFIG_DEBUG_VM_VMACACHE .
6650 .IR vmacache_full_flushes " (since Linux 3.19)"
6651 .\" commit f5f302e21257ebb0c074bbafc37606c26d28cc3d
6652 .\" Present only if the kernel was configured with
6653 .\" .BR CONFIG_DEBUG_VM_VMACACHE .
6656 .IR /proc/zoneinfo " (since Linux 2.6.13)"
6657 This file display information about memory zones.
6658 This is useful for analyzing virtual memory behavior.
6659 .\" FIXME more should be said about /proc/zoneinfo
6661 Many files contain strings (e.g., the environment and command line)
6662 that are in the internal format,
6663 with subfields terminated by null bytes (\(aq\e0\(aq).
6664 When inspecting such files, you may find that the results are more readable
6665 if you use a command of the following form to display them:
6669 .RB "$" " cat \fIfile\fP | tr \(aq\e000\(aq \(aq\en\(aq"
6673 This manual page is incomplete, possibly inaccurate, and is the kind
6674 of thing that needs to be updated very often.
6675 .\" .SH ACKNOWLEDGEMENTS
6676 .\" The material on /proc/sys/fs and /proc/sys/kernel is closely based on
6677 .\" kernel source documentation files written by Rik van Riel.
6709 The Linux kernel source files:
6710 .IR Documentation/filesystems/proc.txt ,
6711 .IR Documentation/sysctl/fs.txt ,
6712 .IR Documentation/sysctl/kernel.txt ,
6713 .IR Documentation/sysctl/net.txt ,
6715 .IR Documentation/sysctl/vm.txt .