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 2010-10-30 "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]/numa_maps " (since Linux 2.6.14)"
483 .IR /proc/[pid]/oom_adj " (since Linux 2.6.11)"
484 This file can be used to adjust the score used to select which process
485 should be killed in an out-of-memory (OOM) situation.
486 The kernel uses this value for a bit-shift operation of the process's
489 valid values are in the range \-16 to +15,
490 plus the special value \-17,
491 which disables OOM-killing altogether for this process.
492 A positive score increases the likelihood of this
493 process being killed by the OOM-killer;
494 a negative score decreases the likelihood.
495 The default value for this file is 0;
496 a new process inherits its parent's
499 A process must be privileged
500 .RB ( CAP_SYS_RESOURCE )
503 .IR /proc/[pid]/oom_score " (since Linux 2.6.11)"
504 .\" See mm/oom_kill.c::badness() in the 2.6.25 sources
505 This file displays the current score that the kernel gives to
506 this process for the purpose of selecting a process
508 A higher score means that the process is more likely to be
509 selected by the OOM-killer.
510 The basis for this score is the amount of memory used by the process,
511 with increases (+) or decreases (\-) for factors including:
512 .\" See mm/oom_kill.c::badness() in the 2.6.25 sources
515 whether the process creates a lot of children using
519 whether the process has been running a long time,
520 or has used a lot of CPU time (\-);
522 whether the process has a low nice value (i.e., > 0) (+);
524 whether the process is privileged (\-); and
525 .\" More precisely, if it has CAP_SYS_ADMIN or CAP_SYS_RESOURCE
527 whether the process is making direct hardware access (\-).
528 .\" More precisely, if it has CAP_SYS_RAWIO
533 also reflects the bit-shift adjustment specified by the
535 setting for the process.
536 .\" FIXME Describe /proc/[pid]/pagemap
538 .\" CONFIG_PROC_PAGE_MONITOR
541 UNIX and Linux support the idea of a per-process root of the
542 file system, set by the
545 This file is a symbolic link that points to the process's
546 root directory, and behaves as exe, fd/*, etc. do.
548 .\" The following was still true as at kernel 2.6.13
549 In a multithreaded process, the contents of this symbolic link
550 are not available if the main thread has already terminated
551 (typically by calling
552 .BR pthread_exit (3)).
553 .\" FIXME Describe /proc/[pid]/seccomp
555 .\" FIXME Describe /proc/[pid]/sessionid
556 .\" Added in 2.6.25; read-only; only readable by real UID
557 .\" CONFIG_AUDITSYSCALL
558 .\" FIXME Describe /proc/[pid]/sched
560 .\" CONFIG_SCHED_DEBUG, and additional fields if CONFIG_SCHEDSTATS
561 .\" Displays various scheduling parameters
562 .\" This file can be written, to reset stats
563 .\" FIXME Describe /proc/[pid]/schedstats and
564 .\" /proc/[pid]/task/[tid]/schedstats
566 .\" CONFIG_SCHEDSTATS
568 .IR /proc/[pid]/smaps " (since Linux 2.6.14)"
569 .\" CONFIG_PROC_PAGE_MONITOR
570 This file shows memory consumption for each of the process's mappings.
571 For each of mappings there is a series of lines such as the following:
575 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
585 The first of these lines shows the same information as is displayed
587 .IR /proc/[pid]/maps .
588 The remaining lines show the size of the mapping,
589 the amount of the mapping that is currently resident in RAM,
590 the number of clean and dirty shared pages in the mapping,
591 and the number of clean and dirty private pages in the mapping.
593 This file is only present if the
599 Status information about the process.
603 .IR /usr/src/linux/fs/proc/array.c "."
605 The fields, in order, with their proper
607 format specifiers, are:
614 The filename of the executable, in parentheses.
615 This is visible whether or not the executable is swapped out.
618 One character from the string "RSDZTW" where R is running, S is
619 sleeping in an interruptible wait, D is waiting in uninterruptible
620 disk sleep, Z is zombie, T is traced or stopped (on a signal),
624 The PID of the parent.
627 The process group ID of the process.
630 The session ID of the process.
633 The controlling terminal of the process.
634 (The minor device number is contained in the combination of bits
636 the major device number is in bits 15 to 8.)
639 .\" This field and following, up to and including wchan added 0.99.1
640 The ID of the foreground process group of the controlling
641 terminal of the process.
643 \fIflags\fP %u (%lu before Linux 2.6.22)
644 The kernel flags word of the process.
646 see the PF_* defines in
647 .IR <linux/sched.h> .
648 Details depend on the kernel version.
651 The number of minor faults the process has made which have not
652 required loading a memory page from disk.
656 The number of minor faults that the process's
657 waited-for children have made.
660 The number of major faults the process has made which have
661 required loading a memory page from disk.
664 The number of major faults that the process's
665 waited-for children have made.
668 Amount of time that this process has been scheduled in user mode,
669 measured in clock ticks (divide by
670 .IR sysconf(_SC_CLK_TCK) .
671 This includes guest time, \fIguest_time\fP
672 (time spent running a virtual CPU, see below),
673 so that applications that are not aware of the guest time field
674 do not lose that time from their calculations.
677 Amount of time that this process has been scheduled in kernel mode,
678 measured in clock ticks (divide by
679 .IR sysconf(_SC_CLK_TCK) .
682 Amount of time that this process's
683 waited-for children have been scheduled in user mode,
684 measured in clock ticks (divide by
685 .IR sysconf(_SC_CLK_TCK) .
688 This includes guest time, \fIcguest_time\fP
689 (time spent running a virtual CPU, see below).
692 Amount of time that this process's
693 waited-for children have been scheduled in kernel mode,
694 measured in clock ticks (divide by
695 .IR sysconf(_SC_CLK_TCK) .
698 (Explanation for Linux 2.6)
699 For processes running a real-time scheduling policy
702 .BR sched_setscheduler (2)),
703 this is the negated scheduling priority, minus one;
704 that is, a number in the range \-2 to \-100,
705 corresponding to real-time priorities 1 to 99.
706 For processes running under a non-real-time scheduling policy,
707 this is the raw nice value
708 .RB ( setpriority (2))
709 as represented in the kernel.
710 The kernel stores nice values as numbers
711 in the range 0 (high) to 39 (low),
712 corresponding to the user-visible nice range of \-20 to 19.
714 Before Linux 2.6, this was a scaled value based on
715 the scheduler weighting given to this process.
716 .\" And back in kernel 1.2 days things were different again.
720 .BR setpriority (2)),
721 a value in the range 19 (low priority) to \-20 (high priority).
722 .\" Back in kernel 1.2 days things were different.
725 .\" \fIcounter\fP %ld
726 .\" The current maximum size in jiffies of the process's next timeslice,
727 .\" or what is currently left of its current timeslice, if it is the
728 .\" currently running process.
731 .\" The time in jiffies of the process's next timeout.
732 .\" timeout was removed sometime around 2.1/2.2
733 \fInum_threads\fP %ld
734 Number of threads in this process (since Linux 2.6).
735 Before kernel 2.6, this field was hard coded to 0 as a placeholder
736 for an earlier removed field.
739 \fIitrealvalue\fP %ld
740 The time in jiffies before the next
742 is sent to the process due to an interval timer.
743 Since kernel 2.6.17, this field is no longer maintained,
744 and is hard coded as 0.
746 \fIstarttime\fP %llu (was %lu before Linux 2.6)
747 The time in jiffies the process started after system boot.
750 Virtual memory size in bytes.
753 Resident Set Size: number of pages the process has in real memory.
754 This is just the pages which
755 count toward text, data, or stack space.
756 This does not include pages
757 which have not been demand-loaded in, or which are swapped out.
760 Current soft limit in bytes on the rss of the process;
761 see the description of
767 The address above which program text can run.
770 The address below which program text can run.
773 The address of the start (i.e., bottom) of the stack.
776 The current value of ESP (stack pointer), as found in the
777 kernel stack page for the process.
780 The current EIP (instruction pointer).
784 The bitmap of pending signals, displayed as a decimal number.
785 Obsolete, because it does not provide information on real-time signals; use
786 .I /proc/[pid]/status
790 The bitmap of blocked signals, displayed as a decimal number.
791 Obsolete, because it does not provide information on real-time signals; use
792 .I /proc/[pid]/status
796 The bitmap of ignored signals, displayed as a decimal number.
797 Obsolete, because it does not provide information on real-time signals; use
798 .I /proc/[pid]/status
802 The bitmap of caught signals, displayed as a decimal number.
803 Obsolete, because it does not provide information on real-time signals; use
804 .I /proc/[pid]/status
808 This is the "channel" in which the process is waiting.
810 address of a system call, and can be looked up in a namelist if you
812 (If you have an up-to-date
813 .IR /etc/psdatabase ,
815 try \fIps \-l\fP to see the WCHAN field in action.)
818 .\" nswap was added in 2.0
819 Number of pages swapped (not maintained).
822 .\" cnswap was added in 2.0
823 Cumulative \fInswap\fP for child processes (not maintained).
825 \fIexit_signal\fP %d (since Linux 2.1.22)
826 Signal to be sent to parent when we die.
828 \fIprocessor\fP %d (since Linux 2.2.8)
829 CPU number last executed on.
831 \fIrt_priority\fP %u (since Linux 2.5.19; was %lu before Linux 2.6.22)
832 Real-time scheduling priority, a number in the range 1 to 99 for
833 processes scheduled under a real-time policy,
834 or 0, for non-real-time processes (see
835 .BR sched_setscheduler (2)).
838 \fIpolicy\fP %u (since Linux 2.5.19; was %lu before Linux 2.6.22)
839 Scheduling policy (see
840 .BR sched_setscheduler (2)).
841 Decode using the SCHED_* constants in
844 \fIdelayacct_blkio_ticks\fP %llu (since Linux 2.6.18)
845 Aggregated block I/O delays, measured in clock ticks (centiseconds).
847 \fIguest_time\fP %lu (since Linux 2.6.24)
848 Guest time of the process (time spent running a virtual CPU
849 for a guest operating system), measured in clock ticks (divide by
850 .IR sysconf(_SC_CLK_TCK) .
852 \fIcguest_time\fP %ld (since Linux 2.6.24)
853 Guest time of the process's children, measured in clock ticks (divide by
854 .IR sysconf(_SC_CLK_TCK) .
858 Provides information about memory usage, measured in pages.
863 size total program size
864 (same as VmSize in \fI/proc/[pid]/status\fP)
865 resident resident set size
866 (same as VmRSS in \fI/proc/[pid]/status\fP)
867 share shared pages (from shared mappings)
869 .\" (not including libs; broken, includes data segment)
870 lib library (unused in Linux 2.6)
872 .\" (including libs; broken, includes library text)
873 dt dirty pages (unused in Linux 2.6)
877 .I /proc/[pid]/status
878 Provides much of the information in
882 in a format that's easier for humans to parse.
887 .RB "$" " cat /proc/$$/status"
894 Uid: 1000 1000 1000 1000
910 SigPnd: 0000000000000000
911 ShdPnd: 0000000000000000
912 SigBlk: 0000000000010000
913 SigIgn: 0000000000384004
914 SigCgt: 000000004b813efb
915 CapInh: 0000000000000000
916 CapPrm: 0000000000000000
917 CapEff: 0000000000000000
918 CapBnd: ffffffffffffffff
919 Cpus_allowed: 00000001
923 voluntary_ctxt_switches: 150
924 nonvoluntary_ctxt_switches: 545
928 The fields are as follows:
932 Command run by this process.
935 Current state of the process. One of
946 Thread group ID (i.e., Process ID).
953 PID of process tracing this process (0 if not being traced).
956 Real, effective, saved set, and file system UIDs (GIDs).
959 Number of file descriptor slots currently allocated.
962 Supplementary group list.
965 Peak virtual memory size.
971 Locked memory size (see
975 Peak resident set size ("high water mark").
980 .IR VmData ", " VmStk ", " VmExe :
981 Size of data, stack, and text segments.
984 Shared library code size.
987 Page table entries size (since Linux 2.6.10).
990 Number of threads in process containing this thread.
992 .IR SigPnd ", " ShdPnd :
993 Number of signals pending for thread and for process as a whole (see
998 .IR SigBlk ", " SigIgn ", " SigCgt :
999 Masks indicating signals being blocked, ignored, and caught (see
1002 .IR CapInh ", " CapPrm ", " CapEff :
1003 Masks of capabilities enabled in inheritable, permitted, and effective sets
1005 .BR capabilities (7)).
1008 Capability Bounding set
1009 (since kernel 2.6.26, see
1010 .BR capabilities (7)).
1013 Mask of CPUs on which this process may run
1014 (since Linux 2.6.24, see
1017 .IR Cpus_allowed_list :
1018 Same as previous, but in "list format"
1019 (since Linux 2.6.26, see
1023 Mask of memory nodes allowed to this process
1024 (since Linux 2.6.24, see
1027 .IR Mems_allowed_list :
1028 Same as previous, but in "list format"
1029 (since Linux 2.6.26, see
1032 .IR voluntary_context_switches ", " nonvoluntary_context_switches :
1033 Number of voluntary and involuntary context switches (since Linux 2.6.23).
1036 .IR /proc/[pid]/task " (since Linux 2.6.0-test6)"
1037 This is a directory that contains one subdirectory
1038 for each thread in the process.
1039 The name of each subdirectory is the numerical thread ID
1043 Within each of these subdirectories, there is a set of
1044 files with the same names and contents as under the
1047 For attributes that are shared by all threads, the contents for
1048 each of the files under the
1050 subdirectories will be the same as in the corresponding
1054 (e.g., in a multithreaded process, all of the
1056 files will have the same value as the
1058 file in the parent directory, since all of the threads in a process
1059 share a working directory).
1060 For attributes that are distinct for each thread,
1061 the corresponding files under
1063 may have different values (e.g., various fields in each of the
1064 .I task/[tid]/status
1065 files may be different for each thread).
1067 .\" The following was still true as at kernel 2.6.13
1068 In a multithreaded process, the contents of the
1070 directory are not available if the main thread has already terminated
1071 (typically by calling
1072 .BR pthread_exit (3)).
1075 Advanced power management version and battery information when
1077 is defined at kernel compilation time.
1080 Contains subdirectories for installed busses.
1083 Subdirectory for PCMCIA devices when
1085 is set at kernel compilation time.
1087 .I /proc/bus/pccard/drivers
1090 Contains various bus subdirectories and pseudo-files containing
1091 information about PCI busses, installed devices, and device
1093 Some of these files are not ASCII.
1095 .I /proc/bus/pci/devices
1096 Information about PCI devices.
1097 They may be accessed through
1103 Arguments passed to the Linux kernel at boot time.
1104 Often done via a boot manager such as
1109 .IR /proc/config.gz " (since Linux 2.6)"
1110 This file exposes the configuration options that were used
1111 to build the currently running kernel,
1112 in the same format as they would be shown in the
1114 file that resulted when configuring the kernel (using
1115 .IR "make xconfig" ,
1118 The file contents are compressed; view or search them using
1122 As long as no changes have been made to the following file,
1125 are the same as those provided by :
1129 cat /lib/modules/$(uname -r)/build/.config
1134 is only provided if the kernel is configured with
1135 .BR CONFIG_IKCONFIG_PROC .
1138 This is a collection of CPU and system architecture dependent items,
1139 for each supported architecture a different list.
1140 Two common entries are \fIprocessor\fP which gives CPU number and
1141 \fIbogomips\fP; a system constant that is calculated
1142 during kernel initialization.
1143 SMP machines have information for
1147 Text listing of major numbers and device groups.
1148 This can be used by MAKEDEV scripts for consistency with the kernel.
1150 .IR /proc/diskstats " (since Linux 2.5.69)"
1151 This file contains disk I/O statistics for each disk device.
1152 See the kernel source file
1153 .I Documentation/iostats.txt
1154 for further information.
1157 This is a list of the registered \fIISA\fP DMA (direct memory access)
1163 .I /proc/execdomains
1164 List of the execution domains (ABI personalities).
1167 Frame buffer information when
1169 is defined during kernel compilation.
1171 .I /proc/filesystems
1172 A text listing of the file systems which are supported by the kernel,
1173 namely file systems which were compiled into the kernel or whose kernel
1174 modules are currently loaded.
1176 .BR filesystems (5).)
1177 If a file system is marked with "nodev",
1178 this means that it does not require a block device to be mounted
1179 (e.g., virtual file system, network file system).
1181 Incidentally, this file may be used by
1183 when no file system is specified and it didn't manage to determine the
1185 Then file systems contained in this file are tried
1186 (excepted those that are marked with "nodev").
1193 exists on systems with the IDE bus.
1194 There are directories for each IDE channel and attached device.
1199 cache buffer size in KB
1200 capacity number of sectors
1201 driver driver version
1202 geometry physical and logical geometry
1203 identify in hexadecimal
1205 model manufacturer's model number
1206 settings drive settings
1207 smart_thresholds in hexadecimal
1208 smart_values in hexadecimal
1214 utility provides access to this information in a friendly format.
1217 This is used to record the number of interrupts per CPU per IO device.
1219 for the i386 and x86_64 architectures, at least, this also includes
1220 interrupts internal to the system (that is, not associated with a device
1221 as such), such as NMI (nonmaskable interrupt), LOC (local timer interrupt),
1222 and for SMP systems, TLB (TLB flush interrupt), RES (rescheduling
1223 interrupt), CAL (remote function call interrupt), and possibly others.
1224 Very easy to read formatting, done in ASCII.
1227 I/O memory map in Linux 2.4.
1230 This is a list of currently registered Input-Output port regions that
1233 .IR /proc/kallsyms " (since Linux 2.5.71)"
1234 This holds the kernel exported symbol definitions used by the
1236 tools to dynamically link and bind loadable modules.
1237 In Linux 2.5.47 and earlier, a similar file with slightly different syntax
1242 This file represents the physical memory of the system and is stored
1243 in the ELF core file format.
1244 With this pseudo-file, and an unstripped
1246 .RI ( /usr/src/linux/vmlinux )
1247 binary, GDB can be used to
1248 examine the current state of any kernel data structures.
1250 The total length of the file is the size of physical memory (RAM) plus
1254 This file can be used instead of the
1256 system call to read kernel messages.
1257 A process must have superuser
1258 privileges to read this file, and only one process should read this
1260 This file should not be read if a syslog process is running
1263 system call facility to log kernel messages.
1265 Information in this file is retrieved with the
1269 .IR /proc/ksyms " (Linux 1.1.23-2.5.47)"
1271 .IR /proc/kallsyms .
1274 The first three fields in this file are load average figures
1275 giving the number of jobs in the run queue (state R)
1276 or waiting for disk I/O (state D) averaged over 1, 5, and 15 minutes.
1277 They are the same as the load average numbers given by
1280 The fourth field consists of two numbers separated by a slash (/).
1281 The first of these is the number of currently executing kernel
1282 scheduling entities (processes, threads);
1283 this will be less than or equal to the number of CPUs.
1284 The value after the slash is the number of kernel scheduling entities
1285 that currently exist on the system.
1286 The fifth field is the PID of the process that was most
1287 recently created on the system.
1290 This file shows current file locks
1291 .RB ( flock "(2) and " fcntl (2))
1295 .IR /proc/malloc " (only up to and including Linux 2.2)"
1296 .\" It looks like this only ever did something back in 1.0 days
1297 This file is only present if
1298 .B CONFIG_DEBUG_MALLOC
1299 was defined during compilation.
1302 This file reports statistics about memory usage on the system.
1305 to report the amount of free and used memory (both physical and swap)
1306 on the system as well as the shared memory and buffers used by the
1310 A text list of the modules that have been loaded by the system.
1315 Before kernel 2.4.19, this file was a list
1316 of all the file systems currently mounted on the system.
1317 With the introduction of per-process mount namespaces in
1318 Linux 2.4.19, this file became a link to
1319 .IR /proc/self/mounts ,
1320 which lists the mount points of the process's own mount namespace.
1321 The format of this file is documented in
1325 Memory Type Range Registers.
1326 See the kernel source file
1327 .I Documentation/mtrr.txt
1331 various net pseudo-files, all of which give the status of some part of
1332 the networking layer.
1333 These files contain ASCII structures and are,
1334 therefore, readable with
1336 However, the standard
1338 suite provides much cleaner access to these files.
1341 This holds an ASCII readable dump of the kernel ARP table used for
1342 address resolutions.
1343 It will show both dynamically learned and preprogrammed ARP entries.
1349 IP address HW type Flags HW address Mask Device
1350 192.168.0.50 0x1 0x2 00:50:BF:25:68:F3 * eth0
1351 192.168.0.250 0x1 0xc 00:00:00:00:00:00 * eth0
1356 Here "IP address" is the IPv4 address of the machine and the "HW type"
1357 is the hardware type of the address from RFC\ 826.
1358 The flags are the internal
1359 flags of the ARP structure (as defined in
1360 .IR /usr/include/linux/if_arp.h )
1362 the "HW address" is the data link layer mapping for that IP address if
1366 The dev pseudo-file contains network device status information.
1368 the number of received and sent packets, the number of errors and
1370 and other basic statistics.
1371 These are used by the
1373 program to report device status.
1379 Inter-| Receive | Transmit
1380 face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
1381 lo: 2776770 11307 0 0 0 0 0 0 2776770 11307 0 0 0 0 0 0
1382 eth0: 1215645 2751 0 0 0 0 0 0 1782404 4324 0 0 0 427 0 0
1383 ppp0: 1622270 5552 1 0 0 0 0 0 354130 5669 0 0 0 0 0 0
1384 tap0: 7714 81 0 0 0 0 0 0 7714 81 0 0 0 0 0 0
1389 .\" .I /proc/net/ipx
1392 .\" .I /proc/net/ipx_route
1395 .I /proc/net/dev_mcast
1397 .IR /usr/src/linux/net/core/dev_mcast.c :
1400 indx interface_name dmi_u dmi_g dmi_address
1401 2 eth0 1 0 01005e000001
1402 3 eth1 1 0 01005e000001
1403 4 eth2 1 0 01005e000001
1408 Internet Group Management Protocol.
1410 .IR /usr/src/linux/net/core/igmp.c .
1413 This file uses the same format as the
1415 file and contains the current reverse mapping database used to provide
1417 reverse address lookup services.
1418 If RARP is not configured into the
1420 this file will not be present.
1423 Holds a dump of the RAW socket table.
1424 Much of the information is not of
1426 apart from debugging.
1427 The "sl" value is the kernel hash slot for the
1429 the "local_address" is the local address and protocol number pair.
1431 the internal status of the socket.
1432 The "tx_queue" and "rx_queue" are the
1433 outgoing and incoming data queue in terms of kernel memory usage.
1434 The "tr", "tm\->when", and "rexmits" fields are not used by RAW.
1436 field holds the effective UID of the creator of the socket.
1438 .\" .I /proc/net/route
1439 .\" No information, but looks similar to
1443 This file holds the ASCII data needed for the IP, ICMP, TCP, and UDP
1445 information bases for an SNMP agent.
1448 Holds a dump of the TCP socket table.
1449 Much of the information is not
1450 of use apart from debugging.
1451 The "sl" value is the kernel hash slot
1452 for the socket, the "local_address" is the local address and port number pair.
1453 The "rem_address" is the remote address and port number pair
1455 \&"St" is the internal status of the socket.
1456 The "tx_queue" and "rx_queue" are the
1457 outgoing and incoming data queue in terms of kernel memory usage.
1458 The "tr", "tm\->when", and "rexmits" fields hold internal information of
1459 the kernel socket state and are only useful for debugging.
1461 field holds the effective UID of the creator of the socket.
1464 Holds a dump of the UDP socket table.
1465 Much of the information is not of
1466 use apart from debugging.
1467 The "sl" value is the kernel hash slot for the
1468 socket, the "local_address" is the local address and port number pair.
1469 The "rem_address" is the remote address and port number pair
1470 (if connected). "St" is the internal status of the socket.
1471 The "tx_queue" and "rx_queue" are the outgoing and incoming data queue
1472 in terms of kernel memory usage.
1473 The "tr", "tm\->when", and "rexmits" fields
1474 are not used by UDP.
1476 field holds the effective UID of the creator of the socket.
1482 sl local_address rem_address st tx_queue rx_queue tr rexmits tm\->when uid
1483 1: 01642C89:0201 0C642C89:03FF 01 00000000:00000001 01:000071BA 00000000 0
1484 1: 00000000:0801 00000000:0000 0A 00000000:00000000 00:00000000 6F000100 0
1485 1: 00000000:0201 00000000:0000 0A 00000000:00000000 00:00000000 00000000 0
1491 Lists the UNIX domain sockets present within the system and their
1497 Num RefCount Protocol Flags Type St Path
1498 0: 00000002 00000000 00000000 0001 03
1499 1: 00000001 00000000 00010000 0001 01 /dev/printer
1504 Here "Num" is the kernel table slot number, "RefCount" is the number
1505 of users of the socket, "Protocol" is currently always 0, "Flags"
1506 represent the internal kernel flags holding the status of the
1508 Currently, type is always "1" (UNIX domain datagram sockets are
1509 not yet supported in the kernel).
1510 \&"St" is the internal state of the
1511 socket and Path is the bound path (if any) of the socket.
1514 Contains major and minor numbers of each partition as well as number
1515 of blocks and partition name.
1518 This is a listing of all PCI devices found during kernel initialization
1519 and their configuration.
1521 This file has been deprecated in favor of a new
1524 .RI ( /proc/bus/pci ).
1525 It became optional in Linux 2.2 (available with
1526 .B CONFIG_PCI_OLD_PROC
1527 set at kernel compilation).
1528 It became once more nonoptionally enabled in Linux 2.4.
1529 Next, it was deprecated in Linux 2.6 (still available with
1530 .B CONFIG_PCI_LEGACY_PROC
1531 set), and finally removed altogether since Linux 2.6.17.
1532 .\" FIXME /proc/sched_debug
1534 .\" .IR /proc/sched_debug " (since Linux 2.6.23)"
1535 .\" See also /proc/[pid]/sched
1538 A directory with the
1540 mid-level pseudo-file and various SCSI low-level
1542 which contain a file for each SCSI host in this system, all of
1543 which give the status of some part of the SCSI IO subsystem.
1544 These files contain ASCII structures and are, therefore, readable with
1547 You can also write to some of the files to reconfigure the subsystem or
1548 switch certain features on or off.
1551 This is a listing of all SCSI devices known to the kernel.
1552 The listing is similar to the one seen during bootup.
1553 scsi currently supports only the \fIadd-single-device\fP command which
1554 allows root to add a hotplugged device to the list of known devices.
1560 echo \(aqscsi add-single-device 1 0 5 0\(aq > /proc/scsi/scsi
1565 host scsi1 to scan on SCSI channel 0 for a device on ID 5 LUN 0.
1567 is already a device known on this address or the address is invalid, an
1568 error will be returned.
1570 .I /proc/scsi/[drivername]
1571 \fI[drivername]\fP can currently be NCR53c7xx, aha152x, aha1542, aha1740,
1572 aic7xxx, buslogic, eata_dma, eata_pio, fdomain, in2000, pas16, qlogic,
1573 scsi_debug, seagate, t128, u15-24f, ultrastore, or wd7000.
1574 These directories show up for all drivers that registered at least one
1576 Every directory contains one file per registered host.
1577 Every host-file is named after the number the host was assigned during
1580 Reading these files will usually show driver and host configuration,
1583 Writing to these files allows different things on different hosts.
1584 For example, with the \fIlatency\fP and \fInolatency\fP commands,
1585 root can switch on and off command latency measurement code in the
1587 With the \fIlockup\fP and \fIunlock\fP commands,
1588 root can control bus lockups simulated by the scsi_debug driver.
1591 This directory refers to the process accessing the
1594 and is identical to the
1596 directory named by the process ID of the same process.
1599 Information about kernel caches.
1600 Since Linux 2.6.16 this file is only present if the
1602 kernel configuration option is enabled.
1624 kernel/system statistics.
1625 Varies with architecture.
1630 \fIcpu 3357 0 4313 1362393\fP
1631 The amount of time, measured in units of
1632 USER_HZ (1/100ths of a second on most architectures, use
1633 .IR sysconf(_SC_CLK_TCK)
1634 to obtain the right value),
1635 .\" 1024 on Alpha and ia64
1636 that the system spent in user mode,
1637 user mode with low priority (nice), system mode, and the
1638 idle task, respectively.
1639 .\" FIXME Actually, the following info about the /proc/stat 'cpu' field
1640 .\" does not seem to be quite right (at least in 2.6.12)
1641 The last value should be USER_HZ times the
1642 second entry in the uptime pseudo-file.
1644 In Linux 2.6 this line includes three additional columns:
1646 \- time waiting for I/O to complete (since 2.5.41);
1648 \- time servicing interrupts (since 2.6.0-test4);
1650 \- time servicing softirqs (since 2.6.0-test4).
1652 Since Linux 2.6.11, there is an eighth column,
1654 \- stolen time, which is the time spent in other operating systems when
1655 running in a virtualized environment
1657 Since Linux 2.6.24, there is a ninth column,
1659 which is the time spent running a virtual CPU for guest
1660 operating systems under the control of the Linux kernel.
1661 .\" See Changelog entry for 5e84cfde51cf303d368fcb48f22059f37b3872de
1663 \fIpage 5741 1808\fP
1664 The number of pages the system paged in and the number that were paged
1668 The number of swap pages that have been brought in and out.
1670 .\" FIXME The following is not the full picture for the 'intr' of
1671 .\" /proc/stat on 2.6:
1673 This line shows counts of interrupts serviced since boot time,
1674 for each of the possible system interrupts.
1675 The first column is the total of all interrupts serviced;
1676 each subsequent column is the total for a particular interrupt.
1678 \fIdisk_io: (2,0):(31,30,5764,1,2) (3,0):\fP...
1679 (major,disk_idx):(noinfo, read_io_ops, blks_read, write_io_ops, blks_written)
1684 The number of context switches that the system underwent.
1686 \fIbtime 769041601\fP
1687 boot time, in seconds since the Epoch, 1970-01-01 00:00:00 +0000 (UTC).
1689 \fIprocesses 86031\fP
1690 Number of forks since boot.
1692 \fIprocs_running 6\fP
1693 Number of processes in runnable state.
1694 (Linux 2.5.45 onward.)
1696 \fIprocs_blocked 2\fP
1697 Number of processes blocked waiting for I/O to complete.
1698 (Linux 2.5.45 onward.)
1707 This directory (present since 1.3.57) contains a number of files
1708 and subdirectories corresponding to kernel variables.
1709 These variables can be read and sometimes modified using
1710 the \fI/proc\fP file system, and the (deprecated)
1714 .IR /proc/sys/abi " (since Linux 2.4.10)"
1715 This directory may contain files with application binary information.
1716 .\" On some systems, it is not present.
1717 See the kernel source file
1718 .I Documentation/sysctl/abi.txt
1719 for more information.
1722 This directory may be empty.
1725 This directory contains device-specific information (e.g.,
1726 .IR dev/cdrom/info ).
1728 some systems, it may be empty.
1731 This directory contains the files and subdirectories for kernel variables
1732 related to file systems.
1734 .I /proc/sys/fs/binfmt_misc
1735 Documentation for files in this directory can be found
1736 in the kernel sources in
1737 .IR Documentation/binfmt_misc.txt .
1739 .IR /proc/sys/fs/dentry-state " (since Linux 2.2)"
1740 This file contains information about the status of the
1741 directory cache (dcache).
1742 The file contains six numbers,
1743 .IR nr_dentry ", " nr_unused ", " age_limit " (age in seconds), "
1745 (pages requested by system) and two dummy values.
1749 is the number of allocated dentries (dcache entries).
1750 This field is unused in Linux 2.2.
1753 is the number of unused dentries.
1756 .\" looks like this is unused in kernels 2.2 to 2.6
1757 is the age in seconds after which dcache entries
1758 can be reclaimed when memory is short.
1761 .\" looks like this is unused in kernels 2.2 to 2.6
1762 is nonzero when the kernel has called shrink_dcache_pages() and the
1763 dcache isn't pruned yet.
1766 .I /proc/sys/fs/dir-notify-enable
1767 This file can be used to disable or enable the
1769 interface described in
1771 on a system-wide basis.
1772 A value of 0 in this file disables the interface,
1773 and a value of 1 enables it.
1775 .I /proc/sys/fs/dquot-max
1776 This file shows the maximum number of cached disk quota entries.
1777 On some (2.4) systems, it is not present.
1778 If the number of free cached disk quota entries is very low and
1779 you have some awesome number of simultaneous system users,
1780 you might want to raise the limit.
1782 .I /proc/sys/fs/dquot-nr
1783 This file shows the number of allocated disk quota
1784 entries and the number of free disk quota entries.
1786 .IR /proc/sys/fs/epoll " (since Linux 2.6.28)"
1787 This directory contains the file
1788 .IR max_user_watches ,
1789 which can be used to limit the amount of kernel memory consumed by the
1792 For further details, see
1795 .I /proc/sys/fs/file-max
1797 a system-wide limit on the number of open files for all processes.
1800 which can be used by a process to set the per-process limit,
1802 on the number of files it may open.)
1804 of error messages about running out of file handles,
1805 try increasing this value:
1811 echo 100000 > /proc/sys/fs/file-max
1817 imposes an upper limit on the value that may be placed in
1821 .IR /proc/sys/fs/file-max ","
1823 .I /proc/sys/fs/inode-max
1824 to 3-4 times the new
1826 .IR /proc/sys/fs/file-max ","
1827 or you will run out of inodes.
1829 .I /proc/sys/fs/file-nr
1830 This (read-only) file gives the number of files presently opened.
1831 It contains three numbers: the number of allocated file handles;
1832 the number of free file handles; and the maximum number of file handles.
1833 The kernel allocates file handles dynamically, but it
1834 doesn't free them again.
1835 If the number of allocated files is close to the
1836 maximum, you should consider increasing the maximum.
1837 When the number of free file handles is
1838 large, you've encountered a peak in your usage of file
1839 handles and you probably don't need to increase the maximum.
1841 .I /proc/sys/fs/inode-max
1842 This file contains the maximum number of in-memory inodes.
1843 On some (2.4) systems, it may not be present.
1844 This value should be 3-4 times larger
1847 since \fIstdin\fP, \fIstdout\fP
1848 and network sockets also need an inode to handle them.
1849 When you regularly run out of inodes, you need to increase this value.
1851 .I /proc/sys/fs/inode-nr
1852 This file contains the first two values from
1855 .I /proc/sys/fs/inode-state
1857 contains seven numbers:
1859 .IR nr_free_inodes ,
1861 and four dummy values.
1863 is the number of inodes the system has allocated.
1864 This can be slightly more than
1866 because Linux allocates them one page full at a time.
1868 represents the number of free inodes.
1874 and the system needs to prune the inode list instead of allocating more.
1876 .IR /proc/sys/fs/inotify " (since Linux 2.6.13)"
1877 This directory contains files
1878 .IR max_queued_events ", " max_user_instances ", and " max_user_watches ,
1879 that can be used to limit the amount of kernel memory consumed by the
1882 For further details, see
1885 .I /proc/sys/fs/lease-break-time
1886 This file specifies the grace period that the kernel grants to a process
1887 holding a file lease
1889 after it has sent a signal to that process notifying it
1890 that another process is waiting to open the file.
1891 If the lease holder does not remove or downgrade the lease within
1892 this grace period, the kernel forcibly breaks the lease.
1894 .I /proc/sys/fs/leases-enable
1895 This file can be used to enable or disable file leases
1897 on a system-wide basis.
1898 If this file contains the value 0, leases are disabled.
1899 A nonzero value enables leases.
1901 .IR /proc/sys/fs/mqueue " (since Linux 2.6.6)"
1902 This directory contains files
1903 .IR msg_max ", " msgsize_max ", and " queues_max ,
1904 controlling the resources used by POSIX message queues.
1909 .IR /proc/sys/fs/overflowgid " and " /proc/sys/fs/overflowuid
1911 allow you to change the value of the fixed UID and GID.
1912 The default is 65534.
1913 Some file systems only support 16-bit UIDs and GIDs, although in Linux
1914 UIDs and GIDs are 32 bits.
1915 When one of these file systems is mounted
1916 with writes enabled, any UID or GID that would exceed 65535 is translated
1917 to the overflow value before being written to disk.
1919 .IR /proc/sys/fs/pipe-max-size " (since Linux 2.6.35)"
1920 The value in this file defines an upper limit for raising the capacity
1925 This limit applies only to unprivileged processes.
1926 The default value for this file is 1,048,576.
1927 The value assigned to this file may be rounded upward,
1928 to reflect the value actually employed for a convenient implementation.
1929 To determine the rounded-up value,
1930 display the contents of this file after assigning a value to it.
1931 The minimum value that can be assigned to this file is the system page size.
1933 .IR /proc/sys/fs/suid_dumpable " (since Linux 2.6.13)"
1934 .\" The following is based on text from Documentation/sysctl/kernel.txt
1935 The value in this file determines whether core dump files are
1936 produced for set-user-ID or otherwise protected/tainted binaries.
1937 Three different integer values can be specified:
1940 This provides the traditional (pre-Linux 2.6.13) behavior.
1941 A core dump will not be produced for a process which has
1942 changed credentials (by calling
1945 or similar, or by executing a set-user-ID or set-group-ID program)
1946 or whose binary does not have read permission enabled.
1949 All processes dump core when possible.
1950 The core dump is owned by the file system user ID of the dumping process
1951 and no security is applied.
1952 This is intended for system debugging situations only.
1953 Ptrace is unchecked.
1955 \fI2\ ("suidsafe")\fP
1956 Any binary which normally would not be dumped (see "0" above)
1957 is dumped readable by root only.
1958 This allows the user to remove the core dump file but not to read it.
1959 For security reasons core dumps in this mode will not overwrite one
1960 another or other files.
1961 This mode is appropriate when administrators are
1962 attempting to debug problems in a normal environment.
1964 .I /proc/sys/fs/super-max
1966 controls the maximum number of superblocks, and
1967 thus the maximum number of mounted file systems the kernel
1969 You only need to increase
1971 if you need to mount more file systems than the current value in
1975 .I /proc/sys/fs/super-nr
1977 contains the number of file systems currently mounted.
1980 This directory contains files controlling a range of kernel parameters,
1983 .I /proc/sys/kernel/acct
1985 contains three numbers:
1990 If BSD-style process accounting is enabled these values control
1992 If free space on file system where the log lives goes below
1994 percent accounting suspends.
1995 If free space gets above
1997 percent accounting resumes.
2000 how often the kernel checks the amount of free space (value is in
2002 Default values are 4, 2 and 30.
2003 That is, suspend accounting if 2% or less space is free; resume it
2004 if 4% or more space is free; consider information about amount of free space
2005 valid for 30 seconds.
2007 .IR /proc/sys/kernel/cap-bound " (from Linux 2.2 to 2.6.24)"
2008 This file holds the value of the kernel
2009 .I "capability bounding set"
2010 (expressed as a signed decimal number).
2011 This set is ANDed against the capabilities permitted to a process
2014 Starting with Linux 2.6.25,
2015 the system-wide capability bounding set disappeared,
2016 and was replaced by a per-thread bounding set; see
2017 .BR capabilities (7).
2019 .I /proc/sys/kernel/core_pattern
2023 .I /proc/sys/kernel/core_uses_pid
2027 .I /proc/sys/kernel/ctrl-alt-del
2029 controls the handling of Ctrl-Alt-Del from the keyboard.
2030 When the value in this file is 0, Ctrl-Alt-Del is trapped and
2033 program to handle a graceful restart.
2034 When the value is greater than zero, Linux's reaction to a Vulcan
2035 Nerve Pinch (tm) will be an immediate reboot, without even
2036 syncing its dirty buffers.
2037 Note: when a program (like dosemu) has the keyboard in "raw"
2038 mode, the ctrl-alt-del is intercepted by the program before it
2039 ever reaches the kernel tty layer, and it's up to the program
2040 to decide what to do with it.
2042 .I /proc/sys/kernel/hotplug
2044 contains the path for the hotplug policy agent.
2045 The default value in this file is
2048 .IR /proc/sys/kernel/domainname " and " /proc/sys/kernel/hostname
2049 can be used to set the NIS/YP domainname and the
2050 hostname of your box in exactly the same way as the commands
2058 .RB "#" " echo \(aqdarkstar\(aq > /proc/sys/kernel/hostname"
2059 .RB "#" " echo \(aqmydomain\(aq > /proc/sys/kernel/domainname"
2063 has the same effect as
2067 .RB "#" " hostname \(aqdarkstar\(aq"
2068 .RB "#" " domainname \(aqmydomain\(aq"
2072 Note, however, that the classic darkstar.frop.org has the
2073 hostname "darkstar" and DNS (Internet Domain Name Server)
2074 domainname "frop.org", not to be confused with the NIS (Network
2075 Information Service) or YP (Yellow Pages) domainname.
2077 domain names are in general different.
2078 For a detailed discussion
2083 .I /proc/sys/kernel/htab-reclaim
2084 (PowerPC only) If this file is set to a nonzero value,
2087 .IR Documentation/powerpc/ppc_htab.txt )
2089 each time the system hits the idle loop.
2091 .I /proc/sys/kernel/l2cr
2092 (PowerPC only) This file
2093 contains a flag that controls the L2 cache of G3 processor
2095 If 0, the cache is disabled.
2098 .I /proc/sys/kernel/modprobe
2099 This file contains the path for the kernel module loader.
2100 The default value is
2101 .IR /sbin/modprobe .
2102 The file is only present if the kernel is built with the
2105 It is described by the kernel source file
2106 .I Documentation/kmod.txt
2107 (only present in kernel 2.4 and earlier).
2109 .I /proc/sys/kernel/msgmax
2111 a system-wide limit specifying the maximum number of bytes in
2112 a single message written on a System V message queue.
2114 .I /proc/sys/kernel/msgmni
2115 This file defines the system-wide limit on the number of
2116 message queue identifiers.
2117 (This file is only present in Linux 2.4 onward.)
2119 .I /proc/sys/kernel/msgmnb
2120 This file defines a system-wide parameter used to initialize the
2122 setting for subsequently created message queues.
2125 setting specifies the maximum number of bytes that may be written to the
2128 .IR /proc/sys/kernel/ostype " and " /proc/sys/kernel/osrelease
2133 .IR /proc/sys/kernel/overflowgid " and " /proc/sys/kernel/overflowuid
2134 These files duplicate the files
2135 .I /proc/sys/fs/overflowgid
2137 .IR /proc/sys/fs/overflowuid .
2139 .I /proc/sys/kernel/panic
2140 This file gives read/write access to the kernel variable
2142 If this is zero, the kernel will loop on a panic; if nonzero
2143 it indicates that the kernel should autoreboot after this number
2146 software watchdog device driver, the recommended setting is 60.
2148 .IR /proc/sys/kernel/panic_on_oops " (since Linux 2.5.68)"
2149 This file controls the kernel's behavior when an oops
2150 or BUG is encountered.
2151 If this file contains 0, then the system
2152 tries to continue operation.
2153 If it contains 1, then the system
2154 delays a few seconds (to give klogd time to record the oops output)
2157 .I /proc/sys/kernel/panic
2158 file is also nonzero then the machine will be rebooted.
2160 .IR /proc/sys/kernel/pid_max " (since Linux 2.5.34)"
2161 This file specifies the value at which PIDs wrap around
2162 (i.e., the value in this file is one greater than the maximum PID).
2163 The default value for this file, 32768,
2164 results in the same range of PIDs as on earlier kernels.
2165 On 32-bit platforms, 32768 is the maximum value for
2169 can be set to any value up to 2^22
2170 .RB ( PID_MAX_LIMIT ,
2171 approximately 4 million).
2172 .\" Prior to 2.6.10, pid_max could also be raised above 32768 on 32-bit
2173 .\" platforms, but this broke /proc/[pid]
2174 .\" See http://marc.theaimsgroup.com/?l=linux-kernel&m=109513010926152&w=2
2176 .IR /proc/sys/kernel/powersave-nap " (PowerPC only)"
2177 This file contains a flag.
2178 If set, Linux-PPC will use the "nap" mode of
2180 otherwise the "doze" mode will be used.
2182 .I /proc/sys/kernel/printk
2183 The four values in this file are
2184 .IR console_loglevel ,
2185 .IR default_message_loglevel ,
2186 .IR minimum_console_level ,
2188 .IR default_console_loglevel .
2189 These values influence
2191 behavior when printing or logging error messages.
2194 for more info on the different loglevels.
2195 Messages with a higher priority than
2197 will be printed to the console.
2198 Messages without an explicit priority will be printed with priority
2199 .IR default_message_level .
2200 .I minimum_console_loglevel
2201 is the minimum (highest) value to which
2204 .I default_console_loglevel
2205 is the default value for
2206 .IR console_loglevel .
2208 .IR /proc/sys/kernel/pty " (since Linux 2.6.4)"
2209 This directory contains two files relating to the number of UNIX 98
2210 pseudoterminals (see
2214 .I /proc/sys/kernel/pty/max
2215 This file defines the maximum number of pseudoterminals.
2217 .I /proc/sys/kernel/pty/nr
2219 indicates how many pseudoterminals are currently in use.
2221 .I /proc/sys/kernel/random
2223 contains various parameters controlling the operation of the file
2227 for further information.
2229 .I /proc/sys/kernel/real-root-dev
2230 This file is documented in the kernel source file
2231 .IR Documentation/initrd.txt .
2233 .IR /proc/sys/kernel/reboot-cmd " (Sparc only) "
2234 This file seems to be a way to give an argument to the SPARC
2235 ROM/Flash boot loader.
2236 Maybe to tell it what to do after
2239 .I /proc/sys/kernel/rtsig-max
2240 (Only in kernels up to and including 2.6.7; see
2242 This file can be used to tune the maximum number
2243 of POSIX real-time (queued) signals that can be outstanding
2246 .I /proc/sys/kernel/rtsig-nr
2247 (Only in kernels up to and including 2.6.7.)
2248 This file shows the number POSIX real-time signals currently queued.
2250 .IR /proc/sys/kernel/sem " (since Linux 2.4)"
2251 This file contains 4 numbers defining limits for System V IPC semaphores.
2252 These fields are, in order:
2255 The maximum semaphores per semaphore set.
2257 A system-wide limit on the number of semaphores in all semaphore sets.
2259 The maximum number of operations that may be specified in a
2263 A system-wide limit on the maximum number of semaphore identifiers.
2266 .I /proc/sys/kernel/sg-big-buff
2268 shows the size of the generic SCSI device (sg) buffer.
2269 You can't tune it just yet, but you could change it at
2270 compile time by editing
2271 .I include/scsi/sg.h
2275 However, there shouldn't be any reason to change this value.
2277 .I /proc/sys/kernel/shmall
2279 contains the system-wide limit on the total number of pages of
2280 System V shared memory.
2282 .I /proc/sys/kernel/shmmax
2284 can be used to query and set the run-time limit
2285 on the maximum (System V IPC) shared memory segment size that can be
2287 Shared memory segments up to 1GB are now supported in the
2289 This value defaults to
2292 .I /proc/sys/kernel/shmmni
2293 (available in Linux 2.4 and onward)
2295 specifies the system-wide maximum number of System V shared memory
2296 segments that can be created.
2298 .I /proc/sys/kernel/sysrq
2299 This file controls the functions allowed to be invoked by the SysRq key.
2301 the file contains 1 meaning that every possible SysRq request is allowed
2302 (in older kernel versions, SysRq was disabled by default,
2303 and you were required to specifically enable it at run-time,
2304 but this is not the case any more).
2305 Possible values in this file are:
2307 0 - disable sysrq completely
2308 1 - enable all functions of sysrq
2309 >1 - bitmask of allowed sysrq functions, as follows:
2310 2 - enable control of console logging level
2311 4 - enable control of keyboard (SAK, unraw)
2312 8 - enable debugging dumps of processes etc.
2313 16 - enable sync command
2314 32 - enable remount read-only
2315 64 - enable signalling of processes (term, kill, oom-kill)
2316 128 - allow reboot/poweroff
2317 256 - allow nicing of all real-time tasks
2319 This file is only present if the
2320 .B CONFIG_MAGIC_SYSRQ
2321 kernel configuration option is enabled.
2322 For further details see the kernel source file
2323 .IR Documentation/sysrq.txt .
2325 .I /proc/sys/kernel/version
2326 This file contains a string like:
2328 #5 Wed Feb 25 21:49:24 MET 1998
2331 this is the fifth kernel built from this source base and the
2332 date behind it indicates the time the kernel was built.
2334 .IR /proc/sys/kernel/threads-max " (since Linux 2.3.11)"
2335 This file specifies the system-wide limit on the number of
2336 threads (tasks) that can be created on the system.
2338 .IR /proc/sys/kernel/zero-paged " (PowerPC only) "
2341 When enabled (nonzero), Linux-PPC will pre-zero pages in
2342 the idle loop, possibly speeding up get_free_pages.
2345 This directory contains networking stuff.
2346 Explanations for some of the files under this directory can be found in
2351 .I /proc/sys/net/core/somaxconn
2352 This file defines a ceiling value for the
2358 manual page for details.
2361 This directory may be empty.
2364 This directory supports Sun remote procedure call for network file system
2366 On some systems, it is not present.
2369 This directory contains files for memory management tuning, buffer and
2372 .IR /proc/sys/vm/drop_caches " (since Linux 2.6.16)"
2373 Writing to this file causes the kernel to drop clean caches, dentries and
2374 inodes from memory, causing that memory to become free.
2376 To free pagecache, use
2377 .IR "echo 1 > /proc/sys/vm/drop_caches" ;
2378 to free dentries and inodes, use
2379 .IR "echo 2 > /proc/sys/vm/drop_caches" ;
2380 to free pagecache, dentries and inodes, use
2381 .IR "echo 3 > /proc/sys/vm/drop_caches" .
2383 Because this is a nondestructive operation and dirty objects
2384 are not freeable, the
2389 .IR /proc/sys/vm/legacy_va_layout " (since Linux 2.6.9)"
2390 .\" The following is from Documentation/filesystems/proc.txt
2391 If nonzero, this disables the new 32-bit memory-mapping layout;
2392 the kernel will use the legacy (2.4) layout for all processes.
2394 .IR /proc/sys/vm/memory_failure_early_kill " (since Linux 2.6.32)"
2395 .\" The following is based on the text in Documentation/sysctl/vm.txt
2396 Control how to kill processes when an uncorrected memory error
2397 (typically a 2-bit error in a memory module)
2398 that cannot be handled by the kernel
2399 is detected in the background by hardware.
2400 In some cases (like the page still having a valid copy on disk),
2401 the kernel will handle the failure
2402 transparently without affecting any applications.
2403 But if there is no other up-to-date copy of the data,
2404 it will kill processes to prevent any data corruptions from propagating.
2406 The file has one of the following values:
2409 Kill all processes that have the corrupted-and-not-reloadable page mapped
2410 as soon as the corruption is detected.
2411 Note this is not supported for a few types of pages, like kernel internally
2412 allocated data or the swap cache, but works for the majority of user pages.
2414 Only unmap the corrupted page from all processes and only kill a process
2415 who tries to access it.
2418 The kill is performed using a
2424 Processes can handle this if they want to; see
2428 This feature is only active on architectures/platforms with advanced machine
2429 check handling and depends on the hardware capabilities.
2431 Applications can override the
2432 .I memory_failure_early_kill
2433 setting individually with the
2438 Only present if the kernel was configured with
2439 .BR CONFIG_MEMORY_FAILURE .
2441 .IR /proc/sys/vm/memory_failure_recovery " (since Linux 2.6.32)"
2442 .\" The following is based on the text in Documentation/sysctl/vm.txt
2443 Enable memory failure recovery (when supported by the platform)
2448 Always panic on a memory failure.
2451 Only present if the kernel was configured with
2452 .BR CONFIG_MEMORY_FAILURE .
2454 .IR /proc/sys/vm/oom_dump_tasks " (since Linux 2.6.25)"
2455 .\" The following is from Documentation/sysctl/vm.txt
2456 Enables a system-wide task dump (excluding kernel threads) to be
2457 produced when the kernel performs an OOM-killing.
2458 The dump includes the following information
2459 for each task (thread, process):
2460 thread ID, real user ID, thread group ID (process ID),
2461 virtual memory size, resident set size,
2462 the CPU that the task is scheduled on,
2463 oom_adj score (see the description of
2464 .IR /proc/[pid]/oom_adj ),
2466 This is helpful to determine why the OOM-killer was invoked
2467 and to identify the rogue task that caused it.
2469 If this contains the value zero, this information is suppressed.
2470 On very large systems with thousands of tasks,
2471 it may not be feasible to dump the memory state information for each one.
2472 Such systems should not be forced to incur a performance penalty in
2473 OOM situations when the information may not be desired.
2475 If this is set to nonzero, this information is shown whenever the
2476 OOM-killer actually kills a memory-hogging task.
2478 The default value is 0.
2480 .IR /proc/sys/vm/oom_kill_allocating_task " (since Linux 2.6.24)"
2481 .\" The following is from Documentation/sysctl/vm.txt
2482 This enables or disables killing the OOM-triggering task in
2483 out-of-memory situations.
2485 If this is set to zero, the OOM-killer will scan through the entire
2486 tasklist and select a task based on heuristics to kill.
2487 This normally selects a rogue memory-hogging task that
2488 frees up a large amount of memory when killed.
2490 If this is set to nonzero, the OOM-killer simply kills the task that
2491 triggered the out-of-memory condition.
2492 This avoids a possibly expensive tasklist scan.
2495 .I /proc/sys/vm/panic_on_oom
2496 is nonzero, it takes precedence over whatever value is used in
2497 .IR /proc/sys/vm/oom_kill_allocating_task .
2499 The default value is 0.
2501 .I /proc/sys/vm/overcommit_memory
2502 This file contains the kernel virtual memory accounting mode.
2506 0: heuristic overcommit (this is the default)
2508 1: always overcommit, never check
2510 2: always check, never overcommit
2517 are not checked, and the default check is very weak,
2518 leading to the risk of getting a process "OOM-killed".
2519 Under Linux 2.4 any nonzero value implies mode 1.
2520 In mode 2 (available since Linux 2.6), the total virtual address space
2521 on the system is limited to (SS + RAM*(r/100)),
2522 where SS is the size of the swap space, and RAM
2523 is the size of the physical memory, and r is the contents of the file
2524 .IR /proc/sys/vm/overcommit_ratio .
2526 .I /proc/sys/vm/overcommit_ratio
2527 See the description of
2528 .IR /proc/sys/vm/overcommit_memory .
2530 .IR /proc/sys/vm/panic_on_oom " (since Linux 2.6.18)"
2531 .\" The following is adapted from Documentation/sysctl/vm.txt
2532 This enables or disables a kernel panic in
2533 an out-of-memory situation.
2535 If this file is set to the value 0,
2536 the kernel's OOM-killer will kill some rogue process.
2537 Usually, the OOM-killer is able to kill a rogue process and the
2538 system will survive.
2540 If this file is set to the value 1,
2541 then the kernel normally panics when out-of-memory happens.
2542 However, if a process limits allocations to certain nodes
2543 using memory policies
2548 and those nodes reach memory exhaustion status,
2549 one process may be killed by the OOM-killer.
2550 No panic occurs in this case:
2551 because other nodes' memory may be free,
2552 this means the system as a whole may not have reached
2553 an out-of-memory situation yet.
2555 If this file is set to the value 2,
2556 the kernel always panics when an out-of-memory condition occurs.
2558 The default value is 0.
2559 1 and 2 are for failover of clustering.
2560 Select either according to your policy of failover.
2562 .IR /proc/sys/vm/swappiness
2563 .\" The following is from Documentation/sysctl/vm.txt
2564 The value in this file controls how aggressively the kernel will swap
2566 Higher values increase aggressiveness, lower values
2567 decrease aggressiveness.
2568 The default value is 60.
2570 .IR /proc/sysrq-trigger " (since Linux 2.4.21)"
2571 Writing a character to this file triggers the same SysRq function as
2572 typing ALT-SysRq-<character> (see the description of
2573 .IR /proc/sys/kernel/sysrq ).
2574 This file is normally only writable by
2576 For further details see the kernel source file
2577 .IR Documentation/sysrq.txt .
2580 Subdirectory containing the pseudo-files
2581 .IR msg ", " sem " and " shm "."
2582 These files list the System V Interprocess Communication (IPC) objects
2583 (respectively: message queues, semaphores, and shared memory)
2584 that currently exist on the system,
2585 providing similar information to that available via
2587 These files have headers and are formatted (one IPC object per line)
2588 for easy understanding.
2590 provides further background on the information shown by these files.
2593 Subdirectory containing the pseudo-files and subdirectories for
2594 tty drivers and line disciplines.
2597 This file contains two numbers: the uptime of the system (seconds),
2598 and the amount of time spent in idle process (seconds).
2601 This string identifies the kernel version that is currently running.
2602 It includes the contents of
2603 .IR /proc/sys/kernel/ostype ,
2604 .I /proc/sys/kernel/osrelease
2606 .IR /proc/sys/kernel/version .
2611 Linux version 1.0.9 (quinlan@phaze) #1 Sat May 14 01:51:54 EDT 1994
2615 .\" FIXME Document /proc/timer_list
2617 .\" .IR /proc/timer_list " (since Linux 2.6.21)"
2618 .\" See the 2.6.21 Change log
2619 .\" FIXME Document /proc/timer_stats
2621 .\" .IR /proc/timer_stats " (since Linux 2.6.21)"
2622 .\" See the 2.6.21 Change log
2624 .IR /proc/vmstat " (since Linux 2.6)"
2625 This file displays various virtual memory statistics.
2627 .IR /proc/zoneinfo " (since Linux 2.6.13)"
2628 This file display information about memory zones.
2629 This is useful for analyzing virtual memory behavior.
2630 .\" FIXME more should be said about /proc/zoneinfo
2632 Many strings (i.e., the environment and command line) are in
2633 the internal format, with subfields terminated by null bytes (\(aq\\0\(aq),
2635 may find that things are more readable if you use \fIod \-c\fP or \fItr
2636 "\\000" "\\n"\fP to read them.
2637 Alternatively, \fIecho \`cat <file>\`\fP works well.
2639 This manual page is incomplete, possibly inaccurate, and is the kind
2640 of thing that needs to be updated very often.
2641 .\" .SH ACKNOWLEDGEMENTS
2642 .\" The material on /proc/sys/fs and /proc/sys/kernel is closely based on
2643 .\" kernel source documentation files written by Rik van Riel.
2670 The kernel source files:
2671 .IR Documentation/filesystems/proc.txt ,
2672 .IR Documentation/sysctl/vm.txt