1 .\" Copyright (C) 1994, 1995 by Daniel Quinlan (quinlan@yggdrasil.com)
2 .\" and Copyright (C) 2002-2008 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 .\" This is free documentation; you can redistribute it and/or
10 .\" modify it under the terms of the GNU General Public License as
11 .\" published by the Free Software Foundation; either version 2 of
12 .\" the License, or (at your option) any later version.
14 .\" The GNU General Public License's references to "object code"
15 .\" and "executables" are to be interpreted as the output of any
16 .\" document formatting or typesetting system, including
17 .\" intermediate and printed output.
19 .\" This manual is distributed in the hope that it will be useful,
20 .\" but WITHOUT ANY WARRANTY; without even the implied warranty of
21 .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
22 .\" GNU General Public License for more details.
24 .\" You should have received a copy of the GNU General Public
25 .\" License along with this manual; if not, write to the Free
26 .\" Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111,
29 .\" Modified 1995-05-17 by faith@cs.unc.edu
30 .\" Minor changes by aeb and Marty Leisner (leisner@sdsp.mc.xerox.com).
31 .\" Modified 1996-04-13, 1996-07-22 by aeb@cwi.nl
32 .\" Modified 2001-12-16 by rwhron@earthlink.net
33 .\" Modified 2002-07-13 by jbelton@shaw.ca
34 .\" Modified 2002-07-22, 2003-05-27, 2004-04-06, 2004-05-25
35 .\" by Michael Kerrisk <mtk.manpages@gmail.com>
36 .\" 2004-11-17, mtk -- updated notes on /proc/loadavg
37 .\" 2004-12-01, mtk, rtsig-max and rtsig-nr went away in 2.6.8
38 .\" 2004-12-14, mtk, updated 'statm', and fixed error in order of list
39 .\" 2005-05-12, mtk, updated 'stat'
40 .\" 2005-07-13, mtk, added /proc/sys/fs/mqueue/*
41 .\" 2005-09-16, mtk, Added /proc/sys/fs/suid_dumpable
42 .\" 2005-09-19, mtk, added /proc/zoneinfo
43 .\" 2005-03-01, mtk, moved /proc/sys/fs/mqueue/* material to mq_overview.7.
44 .\" 2008-06-05, mtk, Added /proc/[pid]/oom_score, /proc/[pid]/oom_adj,
45 .\" /proc/[pid]/limits, /proc/[pid]/mountinfo, /proc/[pid]/mountstats,
46 .\" and /proc/[pid]/fdinfo/*.
47 .\" 2008-06-19, mtk, Documented /proc/[pid]/status.
48 .\" 2008-07-15, mtk, added /proc/config.gz
50 .\" FIXME 2.6.13 seems to have /proc/vmcore implemented
51 .\" in the source code, but there is no option available under
52 .\" 'make xconfig'; eventually this should be fixed, and then info
53 .\" from the patch-2.6.13 and change log could be used to write an
54 .\" entry in this man page.
55 .\" Needs CONFIG_VMCORE
57 .\" FIXME cross check against Documentation/filesystems/proc.txt
58 .\" to see what information could be imported from that file
61 .TH PROC 5 2012-02-08 "Linux" "Linux Programmer's Manual"
63 proc \- process information pseudo-file system
67 file system is a pseudo-file system which is used as an interface to
68 kernel data structures.
69 It is commonly mounted at
71 Most of it is read-only, but some files allow kernel variables to be
74 The following outline gives a quick tour through the
80 There is a numerical subdirectory for each running process; the
81 subdirectory is named by the process ID.
82 Each such subdirectory contains the following
83 pseudo-files and directories.
84 .\" FIXME Describe /proc/[pid]/attr and
85 .\" /proc/[pid]/task/[tid]/attr
86 .\" This is a directory
91 .IR /proc/[pid]/auxv " (since 2.6.0-test7)"
92 This contains the contents of the ELF interpreter information passed
93 to the process at exec time.
94 The format is one \fIunsigned long\fP ID
95 plus one \fIunsigned long\fP value for each entry.
96 The last entry contains two zeros.
97 .\" FIXME Describe /proc/[pid]/cgroup and
98 .\" /proc/[pid]/task/[tid]/cgroup
99 .\" Info in Documentation/cgroups.txt
103 .\" FIXME Describe /proc/[pid]/clear_refs
105 .\" "Clears page referenced bits shown in smaps output"
107 .\" CONFIG_PROC_PAGE_MONITOR
109 .I /proc/[pid]/cmdline
110 This holds the complete command line for the process,
111 unless the process is a zombie.
112 .\" In 2.3.26, this also used to be true if the process was swapped out.
113 In the latter case, there is nothing in this file:
114 that is, a read on this file will return 0 characters.
115 The command-line arguments appear in this file as a set of
116 strings separated by null bytes (\(aq\\0\(aq),
117 with a further null byte after the last string.
119 .IR /proc/[pid]/coredump_filter " (since kernel 2.6.23)"
123 .IR /proc/[pid]/cpuset " (since kernel 2.6.12)"
124 .\" and/proc/[pid]/task/[tid]/cpuset
129 This is a symbolic link to the current working directory of the process.
130 To find out the current working directory of process 20,
131 for instance, you can do this:
135 .RB "$" " cd /proc/20/cwd; /bin/pwd"
141 command is often a shell built-in, and might
148 .\" The following was still true as at kernel 2.6.13
149 In a multithreaded process, the contents of this symbolic link
150 are not available if the main thread has already terminated
151 (typically by calling
152 .BR pthread_exit (3)).
154 .I /proc/[pid]/environ
155 This file contains the environment for the process.
156 The entries are separated by null bytes (\(aq\\0\(aq),
157 and there may be a null byte at the end.
158 Thus, to print out the environment of process 1, you would do:
163 .RB "$" " (cat /proc/1/environ; echo) | tr \(aq\e000\(aq \(aq\en\(aq"
169 Under Linux 2.2 and later, this file is a symbolic link
170 containing the actual pathname of the executed command.
171 This symbolic link can be dereferenced normally; attempting to open
172 it will open the executable.
175 to run another copy of the same executable as is being run by
177 .\" The following was still true as at kernel 2.6.13
178 In a multithreaded process, the contents of this symbolic link
179 are not available if the main thread has already terminated
180 (typically by calling
181 .BR pthread_exit (3)).
183 Under Linux 2.0 and earlier
185 is a pointer to the binary which was executed,
186 and appears as a symbolic link.
189 call on this file under Linux 2.0 returns a string in the format:
193 For example, [0301]:1502 would be inode 1502 on device major 03 (IDE,
194 MFM, etc. drives) minor 01 (first partition on the first drive).
199 option can be used to locate the file.
202 This is a subdirectory containing one entry for each file which the
203 process has open, named by its file descriptor, and which is a
204 symbolic link to the actual file.
205 Thus, 0 is standard input, 1 standard output, 2 standard error, etc.
207 .\" The following was still true as at kernel 2.6.13
208 In a multithreaded process, the contents of this directory
209 are not available if the main thread has already terminated
210 (typically by calling
211 .BR pthread_exit (3)).
213 Programs that will take a filename as a command-line argument,
214 but will not take input from standard input if no argument is supplied,
215 or that write to a file named as a command-line argument,
216 but will not send their output to standard output
217 if no argument is supplied, can nevertheless be made to use
218 standard input or standard out using
220 For example, assuming that
222 is the flag designating an input file and
224 is the flag designating an output file:
228 .RB "$" " foobar \-i /proc/self/fd/0 \-o /proc/self/fd/1 ..."
232 and you have a working filter.
233 .\" The following is not true in my tests (MTK):
234 .\" Note that this will not work for
235 .\" programs that seek on their files, as the files in the fd directory
236 .\" are not seekable.
239 is approximately the same as
241 in some UNIX and UNIX-like systems.
242 Most Linux MAKEDEV scripts symbolically link
248 Most systems provide symbolic links
253 which respectively link to the files
260 Thus the example command above could be written as:
264 .RB "$" " foobar \-i /dev/stdin \-o /dev/stdout ..."
267 .\" FIXME Describe /proc/[pid]/loginuid
268 .\" Added in 2.6.11; updating requires CAP_AUDIT_CONTROL
269 .\" CONFIG_AUDITSYSCALL
271 .IR /proc/[pid]/fdinfo/ " (since kernel 2.6.22)"
272 This is a subdirectory containing one entry for each file which the
273 process has open, named by its file descriptor.
274 The contents of each file can be read to obtain information
275 about the corresponding file descriptor, for example:
279 .RB "$" " cat /proc/12015/fdinfo/4"
287 field is a decimal number showing the current file offset.
290 field is an octal number that displays the
291 file access mode and file status flags (see
294 The files in this directory are readable only by the owner of the process.
295 .\" FIXME document /proc/[pid]/io
297 .\" .IR /proc/[pid]/io " (since kernel 2.6.20)"
299 .IR /proc/[pid]/limits " (since kernel 2.6.24)"
300 This file displays the soft limit, hard limit, and units of measurement
301 for each of the process's resource limits (see
303 The file is protected to only allow reading by the real UID of the process.
306 A file containing the currently mapped memory regions and their access
313 address perms offset dev inode pathname
314 08048000-08056000 r-xp 00000000 03:0c 64593 /usr/sbin/gpm
315 08056000-08058000 rw-p 0000d000 03:0c 64593 /usr/sbin/gpm
316 08058000-0805b000 rwxp 00000000 00:00 0
317 40000000-40013000 r-xp 00000000 03:0c 4165 /lib/ld-2.2.4.so
318 40013000-40015000 rw-p 00012000 03:0c 4165 /lib/ld-2.2.4.so
319 4001f000-40135000 r-xp 00000000 03:0c 45494 /lib/libc-2.2.4.so
320 40135000-4013e000 rw-p 00115000 03:0c 45494 /lib/libc-2.2.4.so
321 4013e000-40142000 rw-p 00000000 00:00 0
322 bffff000-c0000000 rwxp 00000000 00:00 0
326 where "address" is the address space in the process that it occupies,
327 "perms" is a set of permissions:
335 p = private (copy on write)
339 "offset" is the offset into the file/whatever, "dev" is the device
340 (major:minor), and "inode" is the inode on that device.
341 0 indicates that no inode is associated with the memory region,
342 as the case would be with BSS (uninitialized data).
344 Under Linux 2.0 there is no field giving pathname.
347 This file can be used to access the pages of a process's memory through
353 .IR /proc/[pid]/mountinfo " (since Linux 2.6.26)"
354 .\" This info adapted from Documentation/filesystems/proc.txt
355 This file contains information about mount points.
356 It contains lines of the form:
360 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
361 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
365 The numbers in parentheses are labels for the descriptions below:
369 mount ID: unique identifier of the mount (may be reused after
373 parent ID: ID of parent mount (or of self for the top of the mount tree).
376 major:minor: value of
378 for files on file system (see
382 root: root of the mount within the file system.
385 mount point: mount point relative to the process's root.
388 mount options: per-mount options.
391 optional fields: zero or more fields of the form "tag[:value]".
394 separator: marks the end of the optional fields.
397 file system type: name of file system in the form "type[.subtype]".
400 mount source: file system-specific information or "none".
403 super options: per-super block options.
406 Parsers should ignore all unrecognized optional fields.
407 Currently the possible optional fields are:
411 mount is shared in peer group X
414 mount is slave to peer group X
417 mount is slave and receives propagation from peer group X (*)
423 (*) X is the closest dominant peer group under the process's root.
424 If X is the immediate master of the mount,
425 or if there is no dominant peer group under the same root,
426 then only the "master:X" field is present
427 and not the "propagate_from:X" field.
429 For more information on mount propagation see:
430 .I Documentation/filesystems/sharedsubtree.txt
431 in the kernel source tree.
433 .IR /proc/[pid]/mounts " (since Linux 2.4.19)"
434 This is a list of all the file systems currently mounted in the
435 process's mount namespace.
436 The format of this file is documented in
438 Since kernel version 2.6.15, this file is pollable:
439 after opening the file for reading, a change in this file
440 (i.e., a file system mount or unmount) causes
442 to mark the file descriptor as readable, and
446 mark the file as having an error condition.
448 .IR /proc/[pid]/mountstats " (since Linux 2.6.17)"
449 This file exports information (statistics, configuration information)
450 about the mount points in the process's name space.
451 Lines in this file have the form:
454 device /dev/sda7 mounted on /home with fstype ext3 [statistics]
458 The fields in each line are:
462 The name of the mounted device
463 (or "nodevice" if there is no corresponding device).
466 The mount point within the file system tree.
469 The file system type.
472 Optional statistics and configuration information.
473 Currently (as at Linux 2.6.26), only NFS file systems export
474 information via this field.
477 This file is only readable by the owner of the process.
479 .IR /proc/[pid]/ns/ " (since Linux 3.0)"
480 .\" See commit 6b4e306aa3dc94a0545eb9279475b1ab6209a31f
481 This is a subdirectory containing one entry for each namespace that
482 supports being manipulated by
484 For information about namespaces, see
487 .IR /proc/[pid]/ns/ipc " (since Linux 3.0)"
488 Bind mounting this file (see
490 to somewhere else in the filesystem keeps
491 the IPC namespace of the process specified by
493 alive even if all processes currently in the namespace terminate.
495 Opening this file returns a file handle for the IPC namespace
496 of the process specified by
498 As long as this file descriptor remains open,
499 the IPC namespace will remain alive,
500 even if all processes in the namespace terminate.
501 The file descriptor can be passed to
504 .IR /proc/[pid]/ns/net " (since Linux 3.0)"
505 Bind mounting this file (see
507 to somewhere else in the filesystem keeps
508 the network namespace of the process specified by
510 alive even if all processes in the namespace terminate.
512 Opening this file returns a file handle for the network namespace
513 of the process specified by
515 As long as this file descriptor remains open,
516 the network namespace will remain alive,
517 even if all processes in the namespace terminate.
518 The file descriptor can be passed to
521 .IR /proc/[pid]/ns/uts " (since Linux 3.0)"
522 Bind mounting this file (see
524 to somewhere else in the filesystem keeps
525 the UTS namespace of the process specified by
527 alive even if all processes currently in the namespace terminate.
529 Opening this file returns a file handle for the UTS namespace
530 of the process specified by
532 As long as this file descriptor remains open,
533 the UTS namespace will remain alive,
534 even if all processes in the namespace terminate.
535 The file descriptor can be passed to
538 .IR /proc/[pid]/numa_maps " (since Linux 2.6.14)"
542 .IR /proc/[pid]/oom_adj " (since Linux 2.6.11)"
543 This file can be used to adjust the score used to select which process
544 should be killed in an out-of-memory (OOM) situation.
545 The kernel uses this value for a bit-shift operation of the process's
548 valid values are in the range \-16 to +15,
549 plus the special value \-17,
550 which disables OOM-killing altogether for this process.
551 A positive score increases the likelihood of this
552 process being killed by the OOM-killer;
553 a negative score decreases the likelihood.
554 The default value for this file is 0;
555 a new process inherits its parent's
558 A process must be privileged
559 .RB ( CAP_SYS_RESOURCE )
562 .IR /proc/[pid]/oom_score " (since Linux 2.6.11)"
563 .\" See mm/oom_kill.c::badness() in the 2.6.25 sources
564 This file displays the current score that the kernel gives to
565 this process for the purpose of selecting a process
567 A higher score means that the process is more likely to be
568 selected by the OOM-killer.
569 The basis for this score is the amount of memory used by the process,
570 with increases (+) or decreases (\-) for factors including:
571 .\" See mm/oom_kill.c::badness() in the 2.6.25 sources
574 whether the process creates a lot of children using
578 whether the process has been running a long time,
579 or has used a lot of CPU time (\-);
581 whether the process has a low nice value (i.e., > 0) (+);
583 whether the process is privileged (\-); and
584 .\" More precisely, if it has CAP_SYS_ADMIN or CAP_SYS_RESOURCE
586 whether the process is making direct hardware access (\-).
587 .\" More precisely, if it has CAP_SYS_RAWIO
592 also reflects the bit-shift adjustment specified by the
594 setting for the process.
595 .\" FIXME Describe /proc/[pid]/pagemap
597 .\" CONFIG_PROC_PAGE_MONITOR
600 UNIX and Linux support the idea of a per-process root of the
601 file system, set by the
604 This file is a symbolic link that points to the process's
605 root directory, and behaves as exe, fd/*, etc. do.
607 .\" The following was still true as at kernel 2.6.13
608 In a multithreaded process, the contents of this symbolic link
609 are not available if the main thread has already terminated
610 (typically by calling
611 .BR pthread_exit (3)).
612 .\" FIXME Describe /proc/[pid]/seccomp
614 .\" FIXME Describe /proc/[pid]/sessionid
615 .\" Added in 2.6.25; read-only; only readable by real UID
616 .\" CONFIG_AUDITSYSCALL
617 .\" FIXME Describe /proc/[pid]/sched
619 .\" CONFIG_SCHED_DEBUG, and additional fields if CONFIG_SCHEDSTATS
620 .\" Displays various scheduling parameters
621 .\" This file can be written, to reset stats
622 .\" FIXME Describe /proc/[pid]/schedstats and
623 .\" /proc/[pid]/task/[tid]/schedstats
625 .\" CONFIG_SCHEDSTATS
627 .IR /proc/[pid]/smaps " (since Linux 2.6.14)"
628 .\" CONFIG_PROC_PAGE_MONITOR
629 This file shows memory consumption for each of the process's mappings.
630 For each of mappings there is a series of lines such as the following:
634 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
644 The first of these lines shows the same information as is displayed
646 .IR /proc/[pid]/maps .
647 The remaining lines show the size of the mapping,
648 the amount of the mapping that is currently resident in RAM,
649 the number of clean and dirty shared pages in the mapping,
650 and the number of clean and dirty private pages in the mapping.
652 This file is only present if the
658 Status information about the process.
662 .IR /usr/src/linux/fs/proc/array.c "."
664 The fields, in order, with their proper
666 format specifiers, are:
673 The filename of the executable, in parentheses.
674 This is visible whether or not the executable is swapped out.
677 One character from the string "RSDZTW" where R is running, S is
678 sleeping in an interruptible wait, D is waiting in uninterruptible
679 disk sleep, Z is zombie, T is traced or stopped (on a signal),
683 The PID of the parent.
686 The process group ID of the process.
689 The session ID of the process.
692 The controlling terminal of the process.
693 (The minor device number is contained in the combination of bits
695 the major device number is in bits 15 to 8.)
698 .\" This field and following, up to and including wchan added 0.99.1
699 The ID of the foreground process group of the controlling
700 terminal of the process.
702 \fIflags\fP %u (%lu before Linux 2.6.22)
703 The kernel flags word of the process.
705 see the PF_* defines in
706 .IR <linux/sched.h> .
707 Details depend on the kernel version.
710 The number of minor faults the process has made which have not
711 required loading a memory page from disk.
715 The number of minor faults that the process's
716 waited-for children have made.
719 The number of major faults the process has made which have
720 required loading a memory page from disk.
723 The number of major faults that the process's
724 waited-for children have made.
727 Amount of time that this process has been scheduled in user mode,
728 measured in clock ticks (divide by
729 .IR sysconf(_SC_CLK_TCK) .
730 This includes guest time, \fIguest_time\fP
731 (time spent running a virtual CPU, see below),
732 so that applications that are not aware of the guest time field
733 do not lose that time from their calculations.
736 Amount of time that this process has been scheduled in kernel mode,
737 measured in clock ticks (divide by
738 .IR sysconf(_SC_CLK_TCK) .
741 Amount of time that this process's
742 waited-for children have been scheduled in user mode,
743 measured in clock ticks (divide by
744 .IR sysconf(_SC_CLK_TCK) .
747 This includes guest time, \fIcguest_time\fP
748 (time spent running a virtual CPU, see below).
751 Amount of time that this process's
752 waited-for children have been scheduled in kernel mode,
753 measured in clock ticks (divide by
754 .IR sysconf(_SC_CLK_TCK) .
757 (Explanation for Linux 2.6)
758 For processes running a real-time scheduling policy
761 .BR sched_setscheduler (2)),
762 this is the negated scheduling priority, minus one;
763 that is, a number in the range \-2 to \-100,
764 corresponding to real-time priorities 1 to 99.
765 For processes running under a non-real-time scheduling policy,
766 this is the raw nice value
767 .RB ( setpriority (2))
768 as represented in the kernel.
769 The kernel stores nice values as numbers
770 in the range 0 (high) to 39 (low),
771 corresponding to the user-visible nice range of \-20 to 19.
773 Before Linux 2.6, this was a scaled value based on
774 the scheduler weighting given to this process.
775 .\" And back in kernel 1.2 days things were different again.
779 .BR setpriority (2)),
780 a value in the range 19 (low priority) to \-20 (high priority).
781 .\" Back in kernel 1.2 days things were different.
784 .\" \fIcounter\fP %ld
785 .\" The current maximum size in jiffies of the process's next timeslice,
786 .\" or what is currently left of its current timeslice, if it is the
787 .\" currently running process.
790 .\" The time in jiffies of the process's next timeout.
791 .\" timeout was removed sometime around 2.1/2.2
792 \fInum_threads\fP %ld
793 Number of threads in this process (since Linux 2.6).
794 Before kernel 2.6, this field was hard coded to 0 as a placeholder
795 for an earlier removed field.
798 \fIitrealvalue\fP %ld
799 The time in jiffies before the next
801 is sent to the process due to an interval timer.
802 Since kernel 2.6.17, this field is no longer maintained,
803 and is hard coded as 0.
805 \fIstarttime\fP %llu (was %lu before Linux 2.6)
806 The time in jiffies the process started after system boot.
809 Virtual memory size in bytes.
812 Resident Set Size: number of pages the process has in real memory.
813 This is just the pages which
814 count toward text, data, or stack space.
815 This does not include pages
816 which have not been demand-loaded in, or which are swapped out.
819 Current soft limit in bytes on the rss of the process;
820 see the description of
826 The address above which program text can run.
829 The address below which program text can run.
832 The address of the start (i.e., bottom) of the stack.
835 The current value of ESP (stack pointer), as found in the
836 kernel stack page for the process.
839 The current EIP (instruction pointer).
843 The bitmap of pending signals, displayed as a decimal number.
844 Obsolete, because it does not provide information on real-time signals; use
845 .I /proc/[pid]/status
849 The bitmap of blocked signals, displayed as a decimal number.
850 Obsolete, because it does not provide information on real-time signals; use
851 .I /proc/[pid]/status
855 The bitmap of ignored signals, displayed as a decimal number.
856 Obsolete, because it does not provide information on real-time signals; use
857 .I /proc/[pid]/status
861 The bitmap of caught signals, displayed as a decimal number.
862 Obsolete, because it does not provide information on real-time signals; use
863 .I /proc/[pid]/status
867 This is the "channel" in which the process is waiting.
869 address of a system call, and can be looked up in a namelist if you
871 (If you have an up-to-date
872 .IR /etc/psdatabase ,
874 try \fIps \-l\fP to see the WCHAN field in action.)
877 .\" nswap was added in 2.0
878 Number of pages swapped (not maintained).
881 .\" cnswap was added in 2.0
882 Cumulative \fInswap\fP for child processes (not maintained).
884 \fIexit_signal\fP %d (since Linux 2.1.22)
885 Signal to be sent to parent when we die.
887 \fIprocessor\fP %d (since Linux 2.2.8)
888 CPU number last executed on.
890 \fIrt_priority\fP %u (since Linux 2.5.19; was %lu before Linux 2.6.22)
891 Real-time scheduling priority, a number in the range 1 to 99 for
892 processes scheduled under a real-time policy,
893 or 0, for non-real-time processes (see
894 .BR sched_setscheduler (2)).
897 \fIpolicy\fP %u (since Linux 2.5.19; was %lu before Linux 2.6.22)
898 Scheduling policy (see
899 .BR sched_setscheduler (2)).
900 Decode using the SCHED_* constants in
903 \fIdelayacct_blkio_ticks\fP %llu (since Linux 2.6.18)
904 Aggregated block I/O delays, measured in clock ticks (centiseconds).
906 \fIguest_time\fP %lu (since Linux 2.6.24)
907 Guest time of the process (time spent running a virtual CPU
908 for a guest operating system), measured in clock ticks (divide by
909 .IR sysconf(_SC_CLK_TCK) .
911 \fIcguest_time\fP %ld (since Linux 2.6.24)
912 Guest time of the process's children, measured in clock ticks (divide by
913 .IR sysconf(_SC_CLK_TCK) .
917 Provides information about memory usage, measured in pages.
922 size total program size
923 (same as VmSize in \fI/proc/[pid]/status\fP)
924 resident resident set size
925 (same as VmRSS in \fI/proc/[pid]/status\fP)
926 share shared pages (from shared mappings)
928 .\" (not including libs; broken, includes data segment)
929 lib library (unused in Linux 2.6)
931 .\" (including libs; broken, includes library text)
932 dt dirty pages (unused in Linux 2.6)
936 .I /proc/[pid]/status
937 Provides much of the information in
941 in a format that's easier for humans to parse.
946 .RB "$" " cat /proc/$$/status"
953 Uid: 1000 1000 1000 1000
969 SigPnd: 0000000000000000
970 ShdPnd: 0000000000000000
971 SigBlk: 0000000000010000
972 SigIgn: 0000000000384004
973 SigCgt: 000000004b813efb
974 CapInh: 0000000000000000
975 CapPrm: 0000000000000000
976 CapEff: 0000000000000000
977 CapBnd: ffffffffffffffff
978 Cpus_allowed: 00000001
982 voluntary_ctxt_switches: 150
983 nonvoluntary_ctxt_switches: 545
987 The fields are as follows:
991 Command run by this process.
994 Current state of the process. One of
1005 Thread group ID (i.e., Process ID).
1012 PID of parent process.
1015 PID of process tracing this process (0 if not being traced).
1018 Real, effective, saved set, and file system UIDs (GIDs).
1021 Number of file descriptor slots currently allocated.
1024 Supplementary group list.
1027 Peak virtual memory size.
1030 Virtual memory size.
1033 Locked memory size (see
1037 Peak resident set size ("high water mark").
1042 .IR VmData ", " VmStk ", " VmExe :
1043 Size of data, stack, and text segments.
1046 Shared library code size.
1049 Page table entries size (since Linux 2.6.10).
1052 Number of threads in process containing this thread.
1055 This field contains two slash-separated numbers that relate to
1056 queued signals for the real user ID of this process.
1057 The first of these is the number of currently queued
1058 signals for this real user ID, and the second is the
1059 resource limit on the number of queued signals for this process
1060 (see the description of
1061 .BR RLIMIT_SIGPENDING
1065 .IR SigPnd ", " ShdPnd :
1066 Number of signals pending for thread and for process as a whole (see
1071 .IR SigBlk ", " SigIgn ", " SigCgt :
1072 Masks indicating signals being blocked, ignored, and caught (see
1075 .IR CapInh ", " CapPrm ", " CapEff :
1076 Masks of capabilities enabled in inheritable, permitted, and effective sets
1078 .BR capabilities (7)).
1081 Capability Bounding set
1082 (since kernel 2.6.26, see
1083 .BR capabilities (7)).
1086 Mask of CPUs on which this process may run
1087 (since Linux 2.6.24, see
1090 .IR Cpus_allowed_list :
1091 Same as previous, but in "list format"
1092 (since Linux 2.6.26, see
1096 Mask of memory nodes allowed to this process
1097 (since Linux 2.6.24, see
1100 .IR Mems_allowed_list :
1101 Same as previous, but in "list format"
1102 (since Linux 2.6.26, see
1105 .IR voluntary_context_switches ", " nonvoluntary_context_switches :
1106 Number of voluntary and involuntary context switches (since Linux 2.6.23).
1109 .IR /proc/[pid]/task " (since Linux 2.6.0-test6)"
1110 This is a directory that contains one subdirectory
1111 for each thread in the process.
1112 The name of each subdirectory is the numerical thread ID
1116 Within each of these subdirectories, there is a set of
1117 files with the same names and contents as under the
1120 For attributes that are shared by all threads, the contents for
1121 each of the files under the
1123 subdirectories will be the same as in the corresponding
1127 (e.g., in a multithreaded process, all of the
1129 files will have the same value as the
1131 file in the parent directory, since all of the threads in a process
1132 share a working directory).
1133 For attributes that are distinct for each thread,
1134 the corresponding files under
1136 may have different values (e.g., various fields in each of the
1137 .I task/[tid]/status
1138 files may be different for each thread).
1140 .\" The following was still true as at kernel 2.6.13
1141 In a multithreaded process, the contents of the
1143 directory are not available if the main thread has already terminated
1144 (typically by calling
1145 .BR pthread_exit (3)).
1148 Advanced power management version and battery information when
1150 is defined at kernel compilation time.
1153 Contains subdirectories for installed busses.
1156 Subdirectory for PCMCIA devices when
1158 is set at kernel compilation time.
1160 .I /proc/bus/pccard/drivers
1163 Contains various bus subdirectories and pseudo-files containing
1164 information about PCI busses, installed devices, and device
1166 Some of these files are not ASCII.
1168 .I /proc/bus/pci/devices
1169 Information about PCI devices.
1170 They may be accessed through
1176 Arguments passed to the Linux kernel at boot time.
1177 Often done via a boot manager such as
1182 .IR /proc/config.gz " (since Linux 2.6)"
1183 This file exposes the configuration options that were used
1184 to build the currently running kernel,
1185 in the same format as they would be shown in the
1187 file that resulted when configuring the kernel (using
1188 .IR "make xconfig" ,
1191 The file contents are compressed; view or search them using
1195 As long as no changes have been made to the following file,
1198 are the same as those provided by :
1202 cat /lib/modules/$(uname \-r)/build/.config
1207 is only provided if the kernel is configured with
1208 .BR CONFIG_IKCONFIG_PROC .
1211 This is a collection of CPU and system architecture dependent items,
1212 for each supported architecture a different list.
1213 Two common entries are \fIprocessor\fP which gives CPU number and
1214 \fIbogomips\fP; a system constant that is calculated
1215 during kernel initialization.
1216 SMP machines have information for
1220 Text listing of major numbers and device groups.
1221 This can be used by MAKEDEV scripts for consistency with the kernel.
1223 .IR /proc/diskstats " (since Linux 2.5.69)"
1224 This file contains disk I/O statistics for each disk device.
1225 See the kernel source file
1226 .I Documentation/iostats.txt
1227 for further information.
1230 This is a list of the registered \fIISA\fP DMA (direct memory access)
1236 .I /proc/execdomains
1237 List of the execution domains (ABI personalities).
1240 Frame buffer information when
1242 is defined during kernel compilation.
1244 .I /proc/filesystems
1245 A text listing of the file systems which are supported by the kernel,
1246 namely file systems which were compiled into the kernel or whose kernel
1247 modules are currently loaded.
1249 .BR filesystems (5).)
1250 If a file system is marked with "nodev",
1251 this means that it does not require a block device to be mounted
1252 (e.g., virtual file system, network file system).
1254 Incidentally, this file may be used by
1256 when no file system is specified and it didn't manage to determine the
1258 Then file systems contained in this file are tried
1259 (excepted those that are marked with "nodev").
1266 exists on systems with the IDE bus.
1267 There are directories for each IDE channel and attached device.
1272 cache buffer size in KB
1273 capacity number of sectors
1274 driver driver version
1275 geometry physical and logical geometry
1276 identify in hexadecimal
1278 model manufacturer's model number
1279 settings drive settings
1280 smart_thresholds in hexadecimal
1281 smart_values in hexadecimal
1287 utility provides access to this information in a friendly format.
1290 This is used to record the number of interrupts per CPU per IO device.
1292 for the i386 and x86_64 architectures, at least, this also includes
1293 interrupts internal to the system (that is, not associated with a device
1294 as such), such as NMI (nonmaskable interrupt), LOC (local timer interrupt),
1295 and for SMP systems, TLB (TLB flush interrupt), RES (rescheduling
1296 interrupt), CAL (remote function call interrupt), and possibly others.
1297 Very easy to read formatting, done in ASCII.
1300 I/O memory map in Linux 2.4.
1303 This is a list of currently registered Input-Output port regions that
1306 .IR /proc/kallsyms " (since Linux 2.5.71)"
1307 This holds the kernel exported symbol definitions used by the
1309 tools to dynamically link and bind loadable modules.
1310 In Linux 2.5.47 and earlier, a similar file with slightly different syntax
1315 This file represents the physical memory of the system and is stored
1316 in the ELF core file format.
1317 With this pseudo-file, and an unstripped
1319 .RI ( /usr/src/linux/vmlinux )
1320 binary, GDB can be used to
1321 examine the current state of any kernel data structures.
1323 The total length of the file is the size of physical memory (RAM) plus
1327 This file can be used instead of the
1329 system call to read kernel messages.
1330 A process must have superuser
1331 privileges to read this file, and only one process should read this
1333 This file should not be read if a syslog process is running
1336 system call facility to log kernel messages.
1338 Information in this file is retrieved with the
1342 .IR /proc/ksyms " (Linux 1.1.23-2.5.47)"
1344 .IR /proc/kallsyms .
1347 The first three fields in this file are load average figures
1348 giving the number of jobs in the run queue (state R)
1349 or waiting for disk I/O (state D) averaged over 1, 5, and 15 minutes.
1350 They are the same as the load average numbers given by
1353 The fourth field consists of two numbers separated by a slash (/).
1354 The first of these is the number of currently runnable kernel
1355 scheduling entities (processes, threads).
1356 The value after the slash is the number of kernel scheduling entities
1357 that currently exist on the system.
1358 The fifth field is the PID of the process that was most
1359 recently created on the system.
1362 This file shows current file locks
1363 .RB ( flock "(2) and " fcntl (2))
1367 .IR /proc/malloc " (only up to and including Linux 2.2)"
1368 .\" It looks like this only ever did something back in 1.0 days
1369 This file is only present if
1370 .B CONFIG_DEBUG_MALLOC
1371 was defined during compilation.
1374 This file reports statistics about memory usage on the system.
1377 to report the amount of free and used memory (both physical and swap)
1378 on the system as well as the shared memory and buffers used by the
1382 A text list of the modules that have been loaded by the system.
1387 Before kernel 2.4.19, this file was a list
1388 of all the file systems currently mounted on the system.
1389 With the introduction of per-process mount namespaces in
1390 Linux 2.4.19, this file became a link to
1391 .IR /proc/self/mounts ,
1392 which lists the mount points of the process's own mount namespace.
1393 The format of this file is documented in
1397 Memory Type Range Registers.
1398 See the kernel source file
1399 .I Documentation/mtrr.txt
1403 various net pseudo-files, all of which give the status of some part of
1404 the networking layer.
1405 These files contain ASCII structures and are,
1406 therefore, readable with
1408 However, the standard
1410 suite provides much cleaner access to these files.
1413 This holds an ASCII readable dump of the kernel ARP table used for
1414 address resolutions.
1415 It will show both dynamically learned and preprogrammed ARP entries.
1421 IP address HW type Flags HW address Mask Device
1422 192.168.0.50 0x1 0x2 00:50:BF:25:68:F3 * eth0
1423 192.168.0.250 0x1 0xc 00:00:00:00:00:00 * eth0
1428 Here "IP address" is the IPv4 address of the machine and the "HW type"
1429 is the hardware type of the address from RFC\ 826.
1430 The flags are the internal
1431 flags of the ARP structure (as defined in
1432 .IR /usr/include/linux/if_arp.h )
1434 the "HW address" is the data link layer mapping for that IP address if
1438 The dev pseudo-file contains network device status information.
1440 the number of received and sent packets, the number of errors and
1442 and other basic statistics.
1443 These are used by the
1445 program to report device status.
1451 Inter-| Receive | Transmit
1452 face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
1453 lo: 2776770 11307 0 0 0 0 0 0 2776770 11307 0 0 0 0 0 0
1454 eth0: 1215645 2751 0 0 0 0 0 0 1782404 4324 0 0 0 427 0 0
1455 ppp0: 1622270 5552 1 0 0 0 0 0 354130 5669 0 0 0 0 0 0
1456 tap0: 7714 81 0 0 0 0 0 0 7714 81 0 0 0 0 0 0
1461 .\" .I /proc/net/ipx
1464 .\" .I /proc/net/ipx_route
1467 .I /proc/net/dev_mcast
1469 .IR /usr/src/linux/net/core/dev_mcast.c :
1472 indx interface_name dmi_u dmi_g dmi_address
1473 2 eth0 1 0 01005e000001
1474 3 eth1 1 0 01005e000001
1475 4 eth2 1 0 01005e000001
1480 Internet Group Management Protocol.
1482 .IR /usr/src/linux/net/core/igmp.c .
1485 This file uses the same format as the
1487 file and contains the current reverse mapping database used to provide
1489 reverse address lookup services.
1490 If RARP is not configured into the
1492 this file will not be present.
1495 Holds a dump of the RAW socket table.
1496 Much of the information is not of
1498 apart from debugging.
1499 The "sl" value is the kernel hash slot for the
1501 the "local_address" is the local address and protocol number pair.
1503 the internal status of the socket.
1504 The "tx_queue" and "rx_queue" are the
1505 outgoing and incoming data queue in terms of kernel memory usage.
1506 The "tr", "tm\->when", and "rexmits" fields are not used by RAW.
1508 field holds the effective UID of the creator of the socket.
1510 .\" .I /proc/net/route
1511 .\" No information, but looks similar to
1515 This file holds the ASCII data needed for the IP, ICMP, TCP, and UDP
1517 information bases for an SNMP agent.
1520 Holds a dump of the TCP socket table.
1521 Much of the information is not
1522 of use apart from debugging.
1523 The "sl" value is the kernel hash slot
1524 for the socket, the "local_address" is the local address and port number pair.
1525 The "rem_address" is the remote address and port number pair
1527 \&"St" is the internal status of the socket.
1528 The "tx_queue" and "rx_queue" are the
1529 outgoing and incoming data queue in terms of kernel memory usage.
1530 The "tr", "tm\->when", and "rexmits" fields hold internal information of
1531 the kernel socket state and are only useful for debugging.
1533 field holds the effective UID of the creator of the socket.
1536 Holds a dump of the UDP socket table.
1537 Much of the information is not of
1538 use apart from debugging.
1539 The "sl" value is the kernel hash slot for the
1540 socket, the "local_address" is the local address and port number pair.
1541 The "rem_address" is the remote address and port number pair
1542 (if connected). "St" is the internal status of the socket.
1543 The "tx_queue" and "rx_queue" are the outgoing and incoming data queue
1544 in terms of kernel memory usage.
1545 The "tr", "tm\->when", and "rexmits" fields
1546 are not used by UDP.
1548 field holds the effective UID of the creator of the socket.
1554 sl local_address rem_address st tx_queue rx_queue tr rexmits tm\->when uid
1555 1: 01642C89:0201 0C642C89:03FF 01 00000000:00000001 01:000071BA 00000000 0
1556 1: 00000000:0801 00000000:0000 0A 00000000:00000000 00:00000000 6F000100 0
1557 1: 00000000:0201 00000000:0000 0A 00000000:00000000 00:00000000 00000000 0
1563 Lists the UNIX domain sockets present within the system and their
1569 Num RefCount Protocol Flags Type St Path
1570 0: 00000002 00000000 00000000 0001 03
1571 1: 00000001 00000000 00010000 0001 01 /dev/printer
1576 Here "Num" is the kernel table slot number, "RefCount" is the number
1577 of users of the socket, "Protocol" is currently always 0, "Flags"
1578 represent the internal kernel flags holding the status of the
1580 Currently, type is always "1" (UNIX domain datagram sockets are
1581 not yet supported in the kernel).
1582 \&"St" is the internal state of the
1583 socket and Path is the bound path (if any) of the socket.
1586 Contains major and minor numbers of each partition as well as number
1587 of blocks and partition name.
1590 This is a listing of all PCI devices found during kernel initialization
1591 and their configuration.
1593 This file has been deprecated in favor of a new
1596 .RI ( /proc/bus/pci ).
1597 It became optional in Linux 2.2 (available with
1598 .B CONFIG_PCI_OLD_PROC
1599 set at kernel compilation).
1600 It became once more nonoptionally enabled in Linux 2.4.
1601 Next, it was deprecated in Linux 2.6 (still available with
1602 .B CONFIG_PCI_LEGACY_PROC
1603 set), and finally removed altogether since Linux 2.6.17.
1604 .\" FIXME /proc/sched_debug
1606 .\" .IR /proc/sched_debug " (since Linux 2.6.23)"
1607 .\" See also /proc/[pid]/sched
1610 A directory with the
1612 mid-level pseudo-file and various SCSI low-level
1614 which contain a file for each SCSI host in this system, all of
1615 which give the status of some part of the SCSI IO subsystem.
1616 These files contain ASCII structures and are, therefore, readable with
1619 You can also write to some of the files to reconfigure the subsystem or
1620 switch certain features on or off.
1623 This is a listing of all SCSI devices known to the kernel.
1624 The listing is similar to the one seen during bootup.
1625 scsi currently supports only the \fIadd-single-device\fP command which
1626 allows root to add a hotplugged device to the list of known devices.
1632 echo \(aqscsi add-single-device 1 0 5 0\(aq > /proc/scsi/scsi
1637 host scsi1 to scan on SCSI channel 0 for a device on ID 5 LUN 0.
1639 is already a device known on this address or the address is invalid, an
1640 error will be returned.
1642 .I /proc/scsi/[drivername]
1643 \fI[drivername]\fP can currently be NCR53c7xx, aha152x, aha1542, aha1740,
1644 aic7xxx, buslogic, eata_dma, eata_pio, fdomain, in2000, pas16, qlogic,
1645 scsi_debug, seagate, t128, u15-24f, ultrastore, or wd7000.
1646 These directories show up for all drivers that registered at least one
1648 Every directory contains one file per registered host.
1649 Every host-file is named after the number the host was assigned during
1652 Reading these files will usually show driver and host configuration,
1655 Writing to these files allows different things on different hosts.
1656 For example, with the \fIlatency\fP and \fInolatency\fP commands,
1657 root can switch on and off command latency measurement code in the
1659 With the \fIlockup\fP and \fIunlock\fP commands,
1660 root can control bus lockups simulated by the scsi_debug driver.
1663 This directory refers to the process accessing the
1666 and is identical to the
1668 directory named by the process ID of the same process.
1671 Information about kernel caches.
1672 Since Linux 2.6.16 this file is only present if the
1674 kernel configuration option is enabled.
1696 kernel/system statistics.
1697 Varies with architecture.
1702 \fIcpu 3357 0 4313 1362393\fP
1703 The amount of time, measured in units of
1704 USER_HZ (1/100ths of a second on most architectures, use
1705 .IR sysconf(_SC_CLK_TCK)
1706 to obtain the right value),
1707 .\" 1024 on Alpha and ia64
1708 that the system spent in user mode,
1709 user mode with low priority (nice), system mode, and the
1710 idle task, respectively.
1711 .\" FIXME Actually, the following info about the /proc/stat 'cpu' field
1712 .\" does not seem to be quite right (at least in 2.6.12)
1713 The last value should be USER_HZ times the
1714 second entry in the uptime pseudo-file.
1716 In Linux 2.6 this line includes three additional columns:
1718 \- time waiting for I/O to complete (since 2.5.41);
1720 \- time servicing interrupts (since 2.6.0-test4);
1722 \- time servicing softirqs (since 2.6.0-test4).
1724 Since Linux 2.6.11, there is an eighth column,
1726 \- stolen time, which is the time spent in other operating systems when
1727 running in a virtualized environment
1729 Since Linux 2.6.24, there is a ninth column,
1731 which is the time spent running a virtual CPU for guest
1732 operating systems under the control of the Linux kernel.
1733 .\" See Changelog entry for 5e84cfde51cf303d368fcb48f22059f37b3872de
1735 \fIpage 5741 1808\fP
1736 The number of pages the system paged in and the number that were paged
1740 The number of swap pages that have been brought in and out.
1742 .\" FIXME The following is not the full picture for the 'intr' of
1743 .\" /proc/stat on 2.6:
1745 This line shows counts of interrupts serviced since boot time,
1746 for each of the possible system interrupts.
1747 The first column is the total of all interrupts serviced;
1748 each subsequent column is the total for a particular interrupt.
1750 \fIdisk_io: (2,0):(31,30,5764,1,2) (3,0):\fP...
1751 (major,disk_idx):(noinfo, read_io_ops, blks_read, write_io_ops, blks_written)
1756 The number of context switches that the system underwent.
1758 \fIbtime 769041601\fP
1759 boot time, in seconds since the Epoch, 1970-01-01 00:00:00 +0000 (UTC).
1761 \fIprocesses 86031\fP
1762 Number of forks since boot.
1764 \fIprocs_running 6\fP
1765 Number of processes in runnable state.
1766 (Linux 2.5.45 onward.)
1768 \fIprocs_blocked 2\fP
1769 Number of processes blocked waiting for I/O to complete.
1770 (Linux 2.5.45 onward.)
1779 This directory (present since 1.3.57) contains a number of files
1780 and subdirectories corresponding to kernel variables.
1781 These variables can be read and sometimes modified using
1782 the \fI/proc\fP file system, and the (deprecated)
1786 .IR /proc/sys/abi " (since Linux 2.4.10)"
1787 This directory may contain files with application binary information.
1788 .\" On some systems, it is not present.
1789 See the kernel source file
1790 .I Documentation/sysctl/abi.txt
1791 for more information.
1794 This directory may be empty.
1797 This directory contains device-specific information (e.g.,
1798 .IR dev/cdrom/info ).
1800 some systems, it may be empty.
1803 This directory contains the files and subdirectories for kernel variables
1804 related to file systems.
1806 .I /proc/sys/fs/binfmt_misc
1807 Documentation for files in this directory can be found
1808 in the kernel sources in
1809 .IR Documentation/binfmt_misc.txt .
1811 .IR /proc/sys/fs/dentry-state " (since Linux 2.2)"
1812 This file contains information about the status of the
1813 directory cache (dcache).
1814 The file contains six numbers,
1815 .IR nr_dentry ", " nr_unused ", " age_limit " (age in seconds), "
1817 (pages requested by system) and two dummy values.
1821 is the number of allocated dentries (dcache entries).
1822 This field is unused in Linux 2.2.
1825 is the number of unused dentries.
1828 .\" looks like this is unused in kernels 2.2 to 2.6
1829 is the age in seconds after which dcache entries
1830 can be reclaimed when memory is short.
1833 .\" looks like this is unused in kernels 2.2 to 2.6
1834 is nonzero when the kernel has called shrink_dcache_pages() and the
1835 dcache isn't pruned yet.
1838 .I /proc/sys/fs/dir-notify-enable
1839 This file can be used to disable or enable the
1841 interface described in
1843 on a system-wide basis.
1844 A value of 0 in this file disables the interface,
1845 and a value of 1 enables it.
1847 .I /proc/sys/fs/dquot-max
1848 This file shows the maximum number of cached disk quota entries.
1849 On some (2.4) systems, it is not present.
1850 If the number of free cached disk quota entries is very low and
1851 you have some awesome number of simultaneous system users,
1852 you might want to raise the limit.
1854 .I /proc/sys/fs/dquot-nr
1855 This file shows the number of allocated disk quota
1856 entries and the number of free disk quota entries.
1858 .IR /proc/sys/fs/epoll " (since Linux 2.6.28)"
1859 This directory contains the file
1860 .IR max_user_watches ,
1861 which can be used to limit the amount of kernel memory consumed by the
1864 For further details, see
1867 .I /proc/sys/fs/file-max
1869 a system-wide limit on the number of open files for all processes.
1872 which can be used by a process to set the per-process limit,
1874 on the number of files it may open.)
1876 of error messages about running out of file handles,
1877 try increasing this value:
1883 echo 100000 > /proc/sys/fs/file-max
1889 imposes an upper limit on the value that may be placed in
1893 .IR /proc/sys/fs/file-max ","
1895 .I /proc/sys/fs/inode-max
1896 to 3-4 times the new
1898 .IR /proc/sys/fs/file-max ","
1899 or you will run out of inodes.
1901 .I /proc/sys/fs/file-nr
1902 This (read-only) file gives the number of files presently opened.
1903 It contains three numbers: the number of allocated file handles;
1904 the number of free file handles; and the maximum number of file handles.
1905 The kernel allocates file handles dynamically, but it
1906 doesn't free them again.
1907 If the number of allocated files is close to the
1908 maximum, you should consider increasing the maximum.
1909 When the number of free file handles is
1910 large, you've encountered a peak in your usage of file
1911 handles and you probably don't need to increase the maximum.
1913 .I /proc/sys/fs/inode-max
1914 This file contains the maximum number of in-memory inodes.
1915 On some (2.4) systems, it may not be present.
1916 This value should be 3-4 times larger
1919 since \fIstdin\fP, \fIstdout\fP
1920 and network sockets also need an inode to handle them.
1921 When you regularly run out of inodes, you need to increase this value.
1923 .I /proc/sys/fs/inode-nr
1924 This file contains the first two values from
1927 .I /proc/sys/fs/inode-state
1929 contains seven numbers:
1931 .IR nr_free_inodes ,
1933 and four dummy values.
1935 is the number of inodes the system has allocated.
1936 This can be slightly more than
1938 because Linux allocates them one page full at a time.
1940 represents the number of free inodes.
1946 and the system needs to prune the inode list instead of allocating more.
1948 .IR /proc/sys/fs/inotify " (since Linux 2.6.13)"
1949 This directory contains files
1950 .IR max_queued_events ", " max_user_instances ", and " max_user_watches ,
1951 that can be used to limit the amount of kernel memory consumed by the
1954 For further details, see
1957 .I /proc/sys/fs/lease-break-time
1958 This file specifies the grace period that the kernel grants to a process
1959 holding a file lease
1961 after it has sent a signal to that process notifying it
1962 that another process is waiting to open the file.
1963 If the lease holder does not remove or downgrade the lease within
1964 this grace period, the kernel forcibly breaks the lease.
1966 .I /proc/sys/fs/leases-enable
1967 This file can be used to enable or disable file leases
1969 on a system-wide basis.
1970 If this file contains the value 0, leases are disabled.
1971 A nonzero value enables leases.
1973 .IR /proc/sys/fs/mqueue " (since Linux 2.6.6)"
1974 This directory contains files
1975 .IR msg_max ", " msgsize_max ", and " queues_max ,
1976 controlling the resources used by POSIX message queues.
1981 .IR /proc/sys/fs/overflowgid " and " /proc/sys/fs/overflowuid
1983 allow you to change the value of the fixed UID and GID.
1984 The default is 65534.
1985 Some file systems only support 16-bit UIDs and GIDs, although in Linux
1986 UIDs and GIDs are 32 bits.
1987 When one of these file systems is mounted
1988 with writes enabled, any UID or GID that would exceed 65535 is translated
1989 to the overflow value before being written to disk.
1991 .IR /proc/sys/fs/pipe-max-size " (since Linux 2.6.35)"
1992 The value in this file defines an upper limit for raising the capacity
1997 This limit applies only to unprivileged processes.
1998 The default value for this file is 1,048,576.
1999 The value assigned to this file may be rounded upward,
2000 to reflect the value actually employed for a convenient implementation.
2001 To determine the rounded-up value,
2002 display the contents of this file after assigning a value to it.
2003 The minimum value that can be assigned to this file is the system page size.
2005 .IR /proc/sys/fs/suid_dumpable " (since Linux 2.6.13)"
2006 .\" The following is based on text from Documentation/sysctl/kernel.txt
2007 The value in this file determines whether core dump files are
2008 produced for set-user-ID or otherwise protected/tainted binaries.
2009 Three different integer values can be specified:
2012 This provides the traditional (pre-Linux 2.6.13) behavior.
2013 A core dump will not be produced for a process which has
2014 changed credentials (by calling
2017 or similar, or by executing a set-user-ID or set-group-ID program)
2018 or whose binary does not have read permission enabled.
2021 All processes dump core when possible.
2022 The core dump is owned by the file system user ID of the dumping process
2023 and no security is applied.
2024 This is intended for system debugging situations only.
2025 Ptrace is unchecked.
2027 \fI2\ ("suidsafe")\fP
2028 Any binary which normally would not be dumped (see "0" above)
2029 is dumped readable by root only.
2030 This allows the user to remove the core dump file but not to read it.
2031 For security reasons core dumps in this mode will not overwrite one
2032 another or other files.
2033 This mode is appropriate when administrators are
2034 attempting to debug problems in a normal environment.
2036 .I /proc/sys/fs/super-max
2038 controls the maximum number of superblocks, and
2039 thus the maximum number of mounted file systems the kernel
2041 You only need to increase
2043 if you need to mount more file systems than the current value in
2047 .I /proc/sys/fs/super-nr
2049 contains the number of file systems currently mounted.
2052 This directory contains files controlling a range of kernel parameters,
2055 .I /proc/sys/kernel/acct
2057 contains three numbers:
2062 If BSD-style process accounting is enabled these values control
2064 If free space on file system where the log lives goes below
2066 percent accounting suspends.
2067 If free space gets above
2069 percent accounting resumes.
2072 how often the kernel checks the amount of free space (value is in
2074 Default values are 4, 2 and 30.
2075 That is, suspend accounting if 2% or less space is free; resume it
2076 if 4% or more space is free; consider information about amount of free space
2077 valid for 30 seconds.
2079 .IR /proc/sys/kernel/cap-bound " (from Linux 2.2 to 2.6.24)"
2080 This file holds the value of the kernel
2081 .I "capability bounding set"
2082 (expressed as a signed decimal number).
2083 This set is ANDed against the capabilities permitted to a process
2086 Starting with Linux 2.6.25,
2087 the system-wide capability bounding set disappeared,
2088 and was replaced by a per-thread bounding set; see
2089 .BR capabilities (7).
2091 .I /proc/sys/kernel/core_pattern
2095 .I /proc/sys/kernel/core_uses_pid
2099 .I /proc/sys/kernel/ctrl-alt-del
2101 controls the handling of Ctrl-Alt-Del from the keyboard.
2102 When the value in this file is 0, Ctrl-Alt-Del is trapped and
2105 program to handle a graceful restart.
2106 When the value is greater than zero, Linux's reaction to a Vulcan
2107 Nerve Pinch (tm) will be an immediate reboot, without even
2108 syncing its dirty buffers.
2109 Note: when a program (like dosemu) has the keyboard in "raw"
2110 mode, the ctrl-alt-del is intercepted by the program before it
2111 ever reaches the kernel tty layer, and it's up to the program
2112 to decide what to do with it.
2114 .I /proc/sys/kernel/hotplug
2116 contains the path for the hotplug policy agent.
2117 The default value in this file is
2120 .IR /proc/sys/kernel/domainname " and " /proc/sys/kernel/hostname
2121 can be used to set the NIS/YP domainname and the
2122 hostname of your box in exactly the same way as the commands
2130 .RB "#" " echo \(aqdarkstar\(aq > /proc/sys/kernel/hostname"
2131 .RB "#" " echo \(aqmydomain\(aq > /proc/sys/kernel/domainname"
2135 has the same effect as
2139 .RB "#" " hostname \(aqdarkstar\(aq"
2140 .RB "#" " domainname \(aqmydomain\(aq"
2144 Note, however, that the classic darkstar.frop.org has the
2145 hostname "darkstar" and DNS (Internet Domain Name Server)
2146 domainname "frop.org", not to be confused with the NIS (Network
2147 Information Service) or YP (Yellow Pages) domainname.
2149 domain names are in general different.
2150 For a detailed discussion
2155 .I /proc/sys/kernel/htab-reclaim
2156 (PowerPC only) If this file is set to a nonzero value,
2159 .IR Documentation/powerpc/ppc_htab.txt )
2161 each time the system hits the idle loop.
2163 .I /proc/sys/kernel/l2cr
2164 (PowerPC only) This file
2165 contains a flag that controls the L2 cache of G3 processor
2167 If 0, the cache is disabled.
2170 .I /proc/sys/kernel/modprobe
2171 This file contains the path for the kernel module loader.
2172 The default value is
2173 .IR /sbin/modprobe .
2174 The file is only present if the kernel is built with the
2177 It is described by the kernel source file
2178 .I Documentation/kmod.txt
2179 (only present in kernel 2.4 and earlier).
2181 .I /proc/sys/kernel/msgmax
2183 a system-wide limit specifying the maximum number of bytes in
2184 a single message written on a System V message queue.
2186 .I /proc/sys/kernel/msgmni
2187 This file defines the system-wide limit on the number of
2188 message queue identifiers.
2189 (This file is only present in Linux 2.4 onward.)
2191 .I /proc/sys/kernel/msgmnb
2192 This file defines a system-wide parameter used to initialize the
2194 setting for subsequently created message queues.
2197 setting specifies the maximum number of bytes that may be written to the
2200 .IR /proc/sys/kernel/ostype " and " /proc/sys/kernel/osrelease
2205 .IR /proc/sys/kernel/overflowgid " and " /proc/sys/kernel/overflowuid
2206 These files duplicate the files
2207 .I /proc/sys/fs/overflowgid
2209 .IR /proc/sys/fs/overflowuid .
2211 .I /proc/sys/kernel/panic
2212 This file gives read/write access to the kernel variable
2214 If this is zero, the kernel will loop on a panic; if nonzero
2215 it indicates that the kernel should autoreboot after this number
2218 software watchdog device driver, the recommended setting is 60.
2220 .IR /proc/sys/kernel/panic_on_oops " (since Linux 2.5.68)"
2221 This file controls the kernel's behavior when an oops
2222 or BUG is encountered.
2223 If this file contains 0, then the system
2224 tries to continue operation.
2225 If it contains 1, then the system
2226 delays a few seconds (to give klogd time to record the oops output)
2229 .I /proc/sys/kernel/panic
2230 file is also nonzero then the machine will be rebooted.
2232 .IR /proc/sys/kernel/pid_max " (since Linux 2.5.34)"
2233 This file specifies the value at which PIDs wrap around
2234 (i.e., the value in this file is one greater than the maximum PID).
2235 The default value for this file, 32768,
2236 results in the same range of PIDs as on earlier kernels.
2237 On 32-bit platforms, 32768 is the maximum value for
2241 can be set to any value up to 2^22
2242 .RB ( PID_MAX_LIMIT ,
2243 approximately 4 million).
2244 .\" Prior to 2.6.10, pid_max could also be raised above 32768 on 32-bit
2245 .\" platforms, but this broke /proc/[pid]
2246 .\" See http://marc.theaimsgroup.com/?l=linux-kernel&m=109513010926152&w=2
2248 .IR /proc/sys/kernel/powersave-nap " (PowerPC only)"
2249 This file contains a flag.
2250 If set, Linux-PPC will use the "nap" mode of
2252 otherwise the "doze" mode will be used.
2254 .I /proc/sys/kernel/printk
2255 The four values in this file are
2256 .IR console_loglevel ,
2257 .IR default_message_loglevel ,
2258 .IR minimum_console_level ,
2260 .IR default_console_loglevel .
2261 These values influence
2263 behavior when printing or logging error messages.
2266 for more info on the different loglevels.
2267 Messages with a higher priority than
2269 will be printed to the console.
2270 Messages without an explicit priority will be printed with priority
2271 .IR default_message_level .
2272 .I minimum_console_loglevel
2273 is the minimum (highest) value to which
2276 .I default_console_loglevel
2277 is the default value for
2278 .IR console_loglevel .
2280 .IR /proc/sys/kernel/pty " (since Linux 2.6.4)"
2281 This directory contains two files relating to the number of UNIX 98
2282 pseudoterminals (see
2286 .I /proc/sys/kernel/pty/max
2287 This file defines the maximum number of pseudoterminals.
2289 .I /proc/sys/kernel/pty/nr
2291 indicates how many pseudoterminals are currently in use.
2293 .I /proc/sys/kernel/random
2295 contains various parameters controlling the operation of the file
2299 for further information.
2301 .I /proc/sys/kernel/real-root-dev
2302 This file is documented in the kernel source file
2303 .IR Documentation/initrd.txt .
2305 .IR /proc/sys/kernel/reboot-cmd " (Sparc only) "
2306 This file seems to be a way to give an argument to the SPARC
2307 ROM/Flash boot loader.
2308 Maybe to tell it what to do after
2311 .I /proc/sys/kernel/rtsig-max
2312 (Only in kernels up to and including 2.6.7; see
2314 This file can be used to tune the maximum number
2315 of POSIX real-time (queued) signals that can be outstanding
2318 .I /proc/sys/kernel/rtsig-nr
2319 (Only in kernels up to and including 2.6.7.)
2320 This file shows the number POSIX real-time signals currently queued.
2322 .IR /proc/sys/kernel/sem " (since Linux 2.4)"
2323 This file contains 4 numbers defining limits for System V IPC semaphores.
2324 These fields are, in order:
2327 The maximum semaphores per semaphore set.
2329 A system-wide limit on the number of semaphores in all semaphore sets.
2331 The maximum number of operations that may be specified in a
2335 A system-wide limit on the maximum number of semaphore identifiers.
2338 .I /proc/sys/kernel/sg-big-buff
2340 shows the size of the generic SCSI device (sg) buffer.
2341 You can't tune it just yet, but you could change it at
2342 compile time by editing
2343 .I include/scsi/sg.h
2347 However, there shouldn't be any reason to change this value.
2349 .I /proc/sys/kernel/shmall
2351 contains the system-wide limit on the total number of pages of
2352 System V shared memory.
2354 .I /proc/sys/kernel/shmmax
2356 can be used to query and set the run-time limit
2357 on the maximum (System V IPC) shared memory segment size that can be
2359 Shared memory segments up to 1GB are now supported in the
2361 This value defaults to
2364 .I /proc/sys/kernel/shmmni
2365 (available in Linux 2.4 and onward)
2367 specifies the system-wide maximum number of System V shared memory
2368 segments that can be created.
2370 .I /proc/sys/kernel/sysrq
2371 This file controls the functions allowed to be invoked by the SysRq key.
2373 the file contains 1 meaning that every possible SysRq request is allowed
2374 (in older kernel versions, SysRq was disabled by default,
2375 and you were required to specifically enable it at run-time,
2376 but this is not the case any more).
2377 Possible values in this file are:
2379 0 - disable sysrq completely
2380 1 - enable all functions of sysrq
2381 >1 - bitmask of allowed sysrq functions, as follows:
2382 2 - enable control of console logging level
2383 4 - enable control of keyboard (SAK, unraw)
2384 8 - enable debugging dumps of processes etc.
2385 16 - enable sync command
2386 32 - enable remount read-only
2387 64 - enable signalling of processes (term, kill, oom-kill)
2388 128 - allow reboot/poweroff
2389 256 - allow nicing of all real-time tasks
2391 This file is only present if the
2392 .B CONFIG_MAGIC_SYSRQ
2393 kernel configuration option is enabled.
2394 For further details see the kernel source file
2395 .IR Documentation/sysrq.txt .
2397 .I /proc/sys/kernel/version
2398 This file contains a string like:
2400 #5 Wed Feb 25 21:49:24 MET 1998
2403 this is the fifth kernel built from this source base and the
2404 date behind it indicates the time the kernel was built.
2406 .IR /proc/sys/kernel/threads-max " (since Linux 2.3.11)"
2407 This file specifies the system-wide limit on the number of
2408 threads (tasks) that can be created on the system.
2410 .IR /proc/sys/kernel/zero-paged " (PowerPC only) "
2413 When enabled (nonzero), Linux-PPC will pre-zero pages in
2414 the idle loop, possibly speeding up get_free_pages.
2417 This directory contains networking stuff.
2418 Explanations for some of the files under this directory can be found in
2423 .I /proc/sys/net/core/somaxconn
2424 This file defines a ceiling value for the
2430 manual page for details.
2433 This directory may be empty.
2436 This directory supports Sun remote procedure call for network file system
2438 On some systems, it is not present.
2441 This directory contains files for memory management tuning, buffer and
2444 .IR /proc/sys/vm/drop_caches " (since Linux 2.6.16)"
2445 Writing to this file causes the kernel to drop clean caches, dentries and
2446 inodes from memory, causing that memory to become free.
2448 To free pagecache, use
2449 .IR "echo 1 > /proc/sys/vm/drop_caches" ;
2450 to free dentries and inodes, use
2451 .IR "echo 2 > /proc/sys/vm/drop_caches" ;
2452 to free pagecache, dentries and inodes, use
2453 .IR "echo 3 > /proc/sys/vm/drop_caches" .
2455 Because this is a nondestructive operation and dirty objects
2456 are not freeable, the
2461 .IR /proc/sys/vm/legacy_va_layout " (since Linux 2.6.9)"
2462 .\" The following is from Documentation/filesystems/proc.txt
2463 If nonzero, this disables the new 32-bit memory-mapping layout;
2464 the kernel will use the legacy (2.4) layout for all processes.
2466 .IR /proc/sys/vm/memory_failure_early_kill " (since Linux 2.6.32)"
2467 .\" The following is based on the text in Documentation/sysctl/vm.txt
2468 Control how to kill processes when an uncorrected memory error
2469 (typically a 2-bit error in a memory module)
2470 that cannot be handled by the kernel
2471 is detected in the background by hardware.
2472 In some cases (like the page still having a valid copy on disk),
2473 the kernel will handle the failure
2474 transparently without affecting any applications.
2475 But if there is no other up-to-date copy of the data,
2476 it will kill processes to prevent any data corruptions from propagating.
2478 The file has one of the following values:
2481 Kill all processes that have the corrupted-and-not-reloadable page mapped
2482 as soon as the corruption is detected.
2483 Note this is not supported for a few types of pages, like kernel internally
2484 allocated data or the swap cache, but works for the majority of user pages.
2486 Only unmap the corrupted page from all processes and only kill a process
2487 who tries to access it.
2490 The kill is performed using a
2496 Processes can handle this if they want to; see
2500 This feature is only active on architectures/platforms with advanced machine
2501 check handling and depends on the hardware capabilities.
2503 Applications can override the
2504 .I memory_failure_early_kill
2505 setting individually with the
2510 Only present if the kernel was configured with
2511 .BR CONFIG_MEMORY_FAILURE .
2513 .IR /proc/sys/vm/memory_failure_recovery " (since Linux 2.6.32)"
2514 .\" The following is based on the text in Documentation/sysctl/vm.txt
2515 Enable memory failure recovery (when supported by the platform)
2520 Always panic on a memory failure.
2523 Only present if the kernel was configured with
2524 .BR CONFIG_MEMORY_FAILURE .
2526 .IR /proc/sys/vm/oom_dump_tasks " (since Linux 2.6.25)"
2527 .\" The following is from Documentation/sysctl/vm.txt
2528 Enables a system-wide task dump (excluding kernel threads) to be
2529 produced when the kernel performs an OOM-killing.
2530 The dump includes the following information
2531 for each task (thread, process):
2532 thread ID, real user ID, thread group ID (process ID),
2533 virtual memory size, resident set size,
2534 the CPU that the task is scheduled on,
2535 oom_adj score (see the description of
2536 .IR /proc/[pid]/oom_adj ),
2538 This is helpful to determine why the OOM-killer was invoked
2539 and to identify the rogue task that caused it.
2541 If this contains the value zero, this information is suppressed.
2542 On very large systems with thousands of tasks,
2543 it may not be feasible to dump the memory state information for each one.
2544 Such systems should not be forced to incur a performance penalty in
2545 OOM situations when the information may not be desired.
2547 If this is set to nonzero, this information is shown whenever the
2548 OOM-killer actually kills a memory-hogging task.
2550 The default value is 0.
2552 .IR /proc/sys/vm/oom_kill_allocating_task " (since Linux 2.6.24)"
2553 .\" The following is from Documentation/sysctl/vm.txt
2554 This enables or disables killing the OOM-triggering task in
2555 out-of-memory situations.
2557 If this is set to zero, the OOM-killer will scan through the entire
2558 tasklist and select a task based on heuristics to kill.
2559 This normally selects a rogue memory-hogging task that
2560 frees up a large amount of memory when killed.
2562 If this is set to nonzero, the OOM-killer simply kills the task that
2563 triggered the out-of-memory condition.
2564 This avoids a possibly expensive tasklist scan.
2567 .I /proc/sys/vm/panic_on_oom
2568 is nonzero, it takes precedence over whatever value is used in
2569 .IR /proc/sys/vm/oom_kill_allocating_task .
2571 The default value is 0.
2573 .I /proc/sys/vm/overcommit_memory
2574 This file contains the kernel virtual memory accounting mode.
2578 0: heuristic overcommit (this is the default)
2580 1: always overcommit, never check
2582 2: always check, never overcommit
2589 are not checked, and the default check is very weak,
2590 leading to the risk of getting a process "OOM-killed".
2591 Under Linux 2.4 any nonzero value implies mode 1.
2592 In mode 2 (available since Linux 2.6), the total virtual address space
2593 on the system is limited to (SS + RAM*(r/100)),
2594 where SS is the size of the swap space, and RAM
2595 is the size of the physical memory, and r is the contents of the file
2596 .IR /proc/sys/vm/overcommit_ratio .
2598 .I /proc/sys/vm/overcommit_ratio
2599 See the description of
2600 .IR /proc/sys/vm/overcommit_memory .
2602 .IR /proc/sys/vm/panic_on_oom " (since Linux 2.6.18)"
2603 .\" The following is adapted from Documentation/sysctl/vm.txt
2604 This enables or disables a kernel panic in
2605 an out-of-memory situation.
2607 If this file is set to the value 0,
2608 the kernel's OOM-killer will kill some rogue process.
2609 Usually, the OOM-killer is able to kill a rogue process and the
2610 system will survive.
2612 If this file is set to the value 1,
2613 then the kernel normally panics when out-of-memory happens.
2614 However, if a process limits allocations to certain nodes
2615 using memory policies
2620 and those nodes reach memory exhaustion status,
2621 one process may be killed by the OOM-killer.
2622 No panic occurs in this case:
2623 because other nodes' memory may be free,
2624 this means the system as a whole may not have reached
2625 an out-of-memory situation yet.
2627 If this file is set to the value 2,
2628 the kernel always panics when an out-of-memory condition occurs.
2630 The default value is 0.
2631 1 and 2 are for failover of clustering.
2632 Select either according to your policy of failover.
2634 .IR /proc/sys/vm/swappiness
2635 .\" The following is from Documentation/sysctl/vm.txt
2636 The value in this file controls how aggressively the kernel will swap
2638 Higher values increase aggressiveness, lower values
2639 decrease aggressiveness.
2640 The default value is 60.
2642 .IR /proc/sysrq-trigger " (since Linux 2.4.21)"
2643 Writing a character to this file triggers the same SysRq function as
2644 typing ALT-SysRq-<character> (see the description of
2645 .IR /proc/sys/kernel/sysrq ).
2646 This file is normally only writable by
2648 For further details see the kernel source file
2649 .IR Documentation/sysrq.txt .
2652 Subdirectory containing the pseudo-files
2653 .IR msg ", " sem " and " shm "."
2654 These files list the System V Interprocess Communication (IPC) objects
2655 (respectively: message queues, semaphores, and shared memory)
2656 that currently exist on the system,
2657 providing similar information to that available via
2659 These files have headers and are formatted (one IPC object per line)
2660 for easy understanding.
2662 provides further background on the information shown by these files.
2665 Subdirectory containing the pseudo-files and subdirectories for
2666 tty drivers and line disciplines.
2669 This file contains two numbers: the uptime of the system (seconds),
2670 and the amount of time spent in idle process (seconds).
2673 This string identifies the kernel version that is currently running.
2674 It includes the contents of
2675 .IR /proc/sys/kernel/ostype ,
2676 .I /proc/sys/kernel/osrelease
2678 .IR /proc/sys/kernel/version .
2683 Linux version 1.0.9 (quinlan@phaze) #1 Sat May 14 01:51:54 EDT 1994
2687 .\" FIXME Document /proc/timer_list
2689 .\" .IR /proc/timer_list " (since Linux 2.6.21)"
2690 .\" See the 2.6.21 Change log
2691 .\" FIXME Document /proc/timer_stats
2693 .\" .IR /proc/timer_stats " (since Linux 2.6.21)"
2694 .\" See the 2.6.21 Change log
2696 .IR /proc/vmstat " (since Linux 2.6)"
2697 This file displays various virtual memory statistics.
2699 .IR /proc/zoneinfo " (since Linux 2.6.13)"
2700 This file display information about memory zones.
2701 This is useful for analyzing virtual memory behavior.
2702 .\" FIXME more should be said about /proc/zoneinfo
2704 Many strings (i.e., the environment and command line) are in
2705 the internal format, with subfields terminated by null bytes (\(aq\\0\(aq),
2707 may find that things are more readable if you use \fIod \-c\fP or \fItr
2708 "\\000" "\\n"\fP to read them.
2709 Alternatively, \fIecho \`cat <file>\`\fP works well.
2711 This manual page is incomplete, possibly inaccurate, and is the kind
2712 of thing that needs to be updated very often.
2713 .\" .SH ACKNOWLEDGEMENTS
2714 .\" The material on /proc/sys/fs and /proc/sys/kernel is closely based on
2715 .\" kernel source documentation files written by Rik van Riel.
2742 The kernel source files:
2743 .IR Documentation/filesystems/proc.txt ,
2744 .IR Documentation/sysctl/vm.txt