Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1.. SPDX-License-Identifier: GPL-2.0
2
3====================
4The /proc Filesystem
5====================
6
7===================== ======================================= ================
8/proc/sys Terrehon Bowden <terrehon@pacbell.net>, October 7 1999
9 Bodo Bauer <bb@ricochet.net>
102.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
11move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
12fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
13===================== ======================================= ================
14
15
16
17.. Table of Contents
18
19 0 Preface
20 0.1 Introduction/Credits
21 0.2 Legal Stuff
22
23 1 Collecting System Information
24 1.1 Process-Specific Subdirectories
25 1.2 Kernel data
26 1.3 IDE devices in /proc/ide
27 1.4 Networking info in /proc/net
28 1.5 SCSI info
29 1.6 Parallel port info in /proc/parport
30 1.7 TTY info in /proc/tty
31 1.8 Miscellaneous kernel statistics in /proc/stat
32 1.9 Ext4 file system parameters
33
34 2 Modifying System Parameters
35
36 3 Per-Process Parameters
37 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
38 score
39 3.2 /proc/<pid>/oom_score - Display current oom-killer score
40 3.3 /proc/<pid>/io - Display the IO accounting fields
41 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
42 3.5 /proc/<pid>/mountinfo - Information about mounts
43 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
44 3.7 /proc/<pid>/task/<tid>/children - Information about task children
45 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
46 3.9 /proc/<pid>/map_files - Information about memory mapped files
47 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
48 3.11 /proc/<pid>/patch_state - Livepatch patch operation state
49 3.12 /proc/<pid>/arch_status - Task architecture specific information
50 3.13 /proc/<pid>/fd - List of symlinks to open files
51 3.14 /proc/<pid>/ksm_stat - Information about the process's ksm status.
52
53 4 Configuring procfs
54 4.1 Mount options
55
56 5 Filesystem behavior
57
58Preface
59=======
60
610.1 Introduction/Credits
62------------------------
63
64We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
65other people for help compiling this documentation. We'd also like to extend a
66special thank you to Andi Kleen for documentation, which we relied on heavily
67to create this document, as well as the additional information he provided.
68Thanks to everybody else who contributed source or docs to the Linux kernel
69and helped create a great piece of software... :)
70
71The latest version of this document is available online at
72https://www.kernel.org/doc/html/latest/filesystems/proc.html
73
740.2 Legal Stuff
75---------------
76
77We don't guarantee the correctness of this document, and if you come to us
78complaining about how you screwed up your system because of incorrect
79documentation, we won't feel responsible...
80
81Chapter 1: Collecting System Information
82========================================
83
84In This Chapter
85---------------
86* Investigating the properties of the pseudo file system /proc and its
87 ability to provide information on the running Linux system
88* Examining /proc's structure
89* Uncovering various information about the kernel and the processes running
90 on the system
91
92------------------------------------------------------------------------------
93
94The proc file system acts as an interface to internal data structures in the
95kernel. It can be used to obtain information about the system and to change
96certain kernel parameters at runtime (sysctl).
97
98First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
99show you how you can use /proc/sys to change settings.
100
1011.1 Process-Specific Subdirectories
102-----------------------------------
103
104The directory /proc contains (among other things) one subdirectory for each
105process running on the system, which is named after the process ID (PID).
106
107The link 'self' points to the process reading the file system. Each process
108subdirectory has the entries listed in Table 1-1.
109
110A process can read its own information from /proc/PID/* with no extra
111permissions. When reading /proc/PID/* information for other processes, reading
112process is required to have either CAP_SYS_PTRACE capability with
113PTRACE_MODE_READ access permissions, or, alternatively, CAP_PERFMON
114capability. This applies to all read-only information like `maps`, `environ`,
115`pagemap`, etc. The only exception is `mem` file due to its read-write nature,
116which requires CAP_SYS_PTRACE capabilities with more elevated
117PTRACE_MODE_ATTACH permissions; CAP_PERFMON capability does not grant access
118to /proc/PID/mem for other processes.
119
120Note that an open file descriptor to /proc/<pid> or to any of its
121contained files or subdirectories does not prevent <pid> being reused
122for some other process in the event that <pid> exits. Operations on
123open /proc/<pid> file descriptors corresponding to dead processes
124never act on any new process that the kernel may, through chance, have
125also assigned the process ID <pid>. Instead, operations on these FDs
126usually fail with ESRCH.
127
128.. table:: Table 1-1: Process specific entries in /proc
129
130 ============= ===============================================================
131 File Content
132 ============= ===============================================================
133 clear_refs Clears page referenced bits shown in smaps output
134 cmdline Command line arguments
135 cpu Current and last cpu in which it was executed (2.4)(smp)
136 cwd Link to the current working directory
137 environ Values of environment variables
138 exe Link to the executable of this process
139 fd Directory, which contains all file descriptors
140 maps Memory maps to executables and library files (2.4)
141 mem Memory held by this process
142 root Link to the root directory of this process
143 stat Process status
144 statm Process memory status information
145 status Process status in human readable form
146 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function
147 symbol the task is blocked in - or "0" if not blocked.
148 pagemap Page table
149 stack Report full stack trace, enable via CONFIG_STACKTRACE
150 smaps An extension based on maps, showing the memory consumption of
151 each mapping and flags associated with it
152 smaps_rollup Accumulated smaps stats for all mappings of the process. This
153 can be derived from smaps, but is faster and more convenient
154 numa_maps An extension based on maps, showing the memory locality and
155 binding policy as well as mem usage (in pages) of each mapping.
156 ============= ===============================================================
157
158For example, to get the status information of a process, all you have to do is
159read the file /proc/PID/status::
160
161 >cat /proc/self/status
162 Name: cat
163 State: R (running)
164 Tgid: 5452
165 Pid: 5452
166 PPid: 743
167 TracerPid: 0 (2.4)
168 Uid: 501 501 501 501
169 Gid: 100 100 100 100
170 FDSize: 256
171 Groups: 100 14 16
172 Kthread: 0
173 VmPeak: 5004 kB
174 VmSize: 5004 kB
175 VmLck: 0 kB
176 VmHWM: 476 kB
177 VmRSS: 476 kB
178 RssAnon: 352 kB
179 RssFile: 120 kB
180 RssShmem: 4 kB
181 VmData: 156 kB
182 VmStk: 88 kB
183 VmExe: 68 kB
184 VmLib: 1412 kB
185 VmPTE: 20 kb
186 VmSwap: 0 kB
187 HugetlbPages: 0 kB
188 CoreDumping: 0
189 THP_enabled: 1
190 Threads: 1
191 SigQ: 0/28578
192 SigPnd: 0000000000000000
193 ShdPnd: 0000000000000000
194 SigBlk: 0000000000000000
195 SigIgn: 0000000000000000
196 SigCgt: 0000000000000000
197 CapInh: 00000000fffffeff
198 CapPrm: 0000000000000000
199 CapEff: 0000000000000000
200 CapBnd: ffffffffffffffff
201 CapAmb: 0000000000000000
202 NoNewPrivs: 0
203 Seccomp: 0
204 Speculation_Store_Bypass: thread vulnerable
205 SpeculationIndirectBranch: conditional enabled
206 voluntary_ctxt_switches: 0
207 nonvoluntary_ctxt_switches: 1
208
209This shows you nearly the same information you would get if you viewed it with
210the ps command. In fact, ps uses the proc file system to obtain its
211information. But you get a more detailed view of the process by reading the
212file /proc/PID/status. It fields are described in table 1-2.
213
214The statm file contains more detailed information about the process
215memory usage. Its seven fields are explained in Table 1-3. The stat file
216contains detailed information about the process itself. Its fields are
217explained in Table 1-4.
218
219(for SMP CONFIG users)
220
221For making accounting scalable, RSS related information are handled in an
222asynchronous manner and the value may not be very precise. To see a precise
223snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
224It's slow but very precise.
225
226.. table:: Table 1-2: Contents of the status fields (as of 4.19)
227
228 ========================== ===================================================
229 Field Content
230 ========================== ===================================================
231 Name filename of the executable
232 Umask file mode creation mask
233 State state (R is running, S is sleeping, D is sleeping
234 in an uninterruptible wait, Z is zombie,
235 T is traced or stopped)
236 Tgid thread group ID
237 Ngid NUMA group ID (0 if none)
238 Pid process id
239 PPid process id of the parent process
240 TracerPid PID of process tracing this process (0 if not, or
241 the tracer is outside of the current pid namespace)
242 Uid Real, effective, saved set, and file system UIDs
243 Gid Real, effective, saved set, and file system GIDs
244 FDSize number of file descriptor slots currently allocated
245 Groups supplementary group list
246 NStgid descendant namespace thread group ID hierarchy
247 NSpid descendant namespace process ID hierarchy
248 NSpgid descendant namespace process group ID hierarchy
249 NSsid descendant namespace session ID hierarchy
250 Kthread kernel thread flag, 1 is yes, 0 is no
251 VmPeak peak virtual memory size
252 VmSize total program size
253 VmLck locked memory size
254 VmPin pinned memory size
255 VmHWM peak resident set size ("high water mark")
256 VmRSS size of memory portions. It contains the three
257 following parts
258 (VmRSS = RssAnon + RssFile + RssShmem)
259 RssAnon size of resident anonymous memory
260 RssFile size of resident file mappings
261 RssShmem size of resident shmem memory (includes SysV shm,
262 mapping of tmpfs and shared anonymous mappings)
263 VmData size of private data segments
264 VmStk size of stack segments
265 VmExe size of text segment
266 VmLib size of shared library code
267 VmPTE size of page table entries
268 VmSwap amount of swap used by anonymous private data
269 (shmem swap usage is not included)
270 HugetlbPages size of hugetlb memory portions
271 CoreDumping process's memory is currently being dumped
272 (killing the process may lead to a corrupted core)
273 THP_enabled process is allowed to use THP (returns 0 when
274 PR_SET_THP_DISABLE is set on the process to disable
275 THP completely, not just partially)
276 Threads number of threads
277 SigQ number of signals queued/max. number for queue
278 SigPnd bitmap of pending signals for the thread
279 ShdPnd bitmap of shared pending signals for the process
280 SigBlk bitmap of blocked signals
281 SigIgn bitmap of ignored signals
282 SigCgt bitmap of caught signals
283 CapInh bitmap of inheritable capabilities
284 CapPrm bitmap of permitted capabilities
285 CapEff bitmap of effective capabilities
286 CapBnd bitmap of capabilities bounding set
287 CapAmb bitmap of ambient capabilities
288 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...)
289 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
290 Speculation_Store_Bypass speculative store bypass mitigation status
291 SpeculationIndirectBranch indirect branch speculation mode
292 Cpus_allowed mask of CPUs on which this process may run
293 Cpus_allowed_list Same as previous, but in "list format"
294 Mems_allowed mask of memory nodes allowed to this process
295 Mems_allowed_list Same as previous, but in "list format"
296 voluntary_ctxt_switches number of voluntary context switches
297 nonvoluntary_ctxt_switches number of non voluntary context switches
298 ========================== ===================================================
299
300
301.. table:: Table 1-3: Contents of the statm fields (as of 2.6.8-rc3)
302
303 ======== =============================== ==============================
304 Field Content
305 ======== =============================== ==============================
306 size total program size (pages) (same as VmSize in status)
307 resident size of memory portions (pages) (same as VmRSS in status)
308 shared number of pages that are shared (i.e. backed by a file, same
309 as RssFile+RssShmem in status)
310 trs number of pages that are 'code' (not including libs; broken,
311 includes data segment)
312 lrs number of pages of library (always 0 on 2.6)
313 drs number of pages of data/stack (including libs; broken,
314 includes library text)
315 dt number of dirty pages (always 0 on 2.6)
316 ======== =============================== ==============================
317
318
319.. table:: Table 1-4: Contents of the stat fields (as of 2.6.30-rc7)
320
321 ============= ===============================================================
322 Field Content
323 ============= ===============================================================
324 pid process id
325 tcomm filename of the executable
326 state state (R is running, S is sleeping, D is sleeping in an
327 uninterruptible wait, Z is zombie, T is traced or stopped)
328 ppid process id of the parent process
329 pgrp pgrp of the process
330 sid session id
331 tty_nr tty the process uses
332 tty_pgrp pgrp of the tty
333 flags task flags
334 min_flt number of minor faults
335 cmin_flt number of minor faults with child's
336 maj_flt number of major faults
337 cmaj_flt number of major faults with child's
338 utime user mode jiffies
339 stime kernel mode jiffies
340 cutime user mode jiffies with child's
341 cstime kernel mode jiffies with child's
342 priority priority level
343 nice nice level
344 num_threads number of threads
345 it_real_value (obsolete, always 0)
346 start_time time the process started after system boot
347 vsize virtual memory size
348 rss resident set memory size
349 rsslim current limit in bytes on the rss
350 start_code address above which program text can run
351 end_code address below which program text can run
352 start_stack address of the start of the main process stack
353 esp current value of ESP
354 eip current value of EIP
355 pending bitmap of pending signals
356 blocked bitmap of blocked signals
357 sigign bitmap of ignored signals
358 sigcatch bitmap of caught signals
359 0 (place holder, used to be the wchan address,
360 use /proc/PID/wchan instead)
361 0 (place holder)
362 0 (place holder)
363 exit_signal signal to send to parent thread on exit
364 task_cpu which CPU the task is scheduled on
365 rt_priority realtime priority
366 policy scheduling policy (man sched_setscheduler)
367 blkio_ticks time spent waiting for block IO
368 gtime guest time of the task in jiffies
369 cgtime guest time of the task children in jiffies
370 start_data address above which program data+bss is placed
371 end_data address below which program data+bss is placed
372 start_brk address above which program heap can be expanded with brk()
373 arg_start address above which program command line is placed
374 arg_end address below which program command line is placed
375 env_start address above which program environment is placed
376 env_end address below which program environment is placed
377 exit_code the thread's exit_code in the form reported by the waitpid
378 system call
379 ============= ===============================================================
380
381The /proc/PID/maps file contains the currently mapped memory regions and
382their access permissions.
383
384The format is::
385
386 address perms offset dev inode pathname
387
388 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
389 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
390 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
391 a7cb1000-a7cb2000 ---p 00000000 00:00 0
392 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
393 a7eb2000-a7eb3000 ---p 00000000 00:00 0
394 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
395 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
396 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
397 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
398 a800b000-a800e000 rw-p 00000000 00:00 0
399 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
400 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
401 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
402 a8024000-a8027000 rw-p 00000000 00:00 0
403 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
404 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
405 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
406 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
407 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
408
409where "address" is the address space in the process that it occupies, "perms"
410is a set of permissions::
411
412 r = read
413 w = write
414 x = execute
415 s = shared
416 p = private (copy on write)
417
418"offset" is the offset into the mapping, "dev" is the device (major:minor), and
419"inode" is the inode on that device. 0 indicates that no inode is associated
420with the memory region, as the case would be with BSS (uninitialized data).
421The "pathname" shows the name associated file for this mapping. If the mapping
422is not associated with a file:
423
424 =================== ===========================================
425 [heap] the heap of the program
426 [stack] the stack of the main process
427 [vdso] the "virtual dynamic shared object",
428 the kernel system call handler
429 [anon:<name>] a private anonymous mapping that has been
430 named by userspace
431 [anon_shmem:<name>] an anonymous shared memory mapping that has
432 been named by userspace
433 =================== ===========================================
434
435 or if empty, the mapping is anonymous.
436
437Starting with 6.11 kernel, /proc/PID/maps provides an alternative
438ioctl()-based API that gives ability to flexibly and efficiently query and
439filter individual VMAs. This interface is binary and is meant for more
440efficient and easy programmatic use. `struct procmap_query`, defined in
441linux/fs.h UAPI header, serves as an input/output argument to the
442`PROCMAP_QUERY` ioctl() command. See comments in linus/fs.h UAPI header for
443details on query semantics, supported flags, data returned, and general API
444usage information.
445
446The /proc/PID/smaps is an extension based on maps, showing the memory
447consumption for each of the process's mappings. For each mapping (aka Virtual
448Memory Area, or VMA) there is a series of lines such as the following::
449
450 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
451
452 Size: 1084 kB
453 KernelPageSize: 4 kB
454 MMUPageSize: 4 kB
455 Rss: 892 kB
456 Pss: 374 kB
457 Pss_Dirty: 0 kB
458 Shared_Clean: 892 kB
459 Shared_Dirty: 0 kB
460 Private_Clean: 0 kB
461 Private_Dirty: 0 kB
462 Referenced: 892 kB
463 Anonymous: 0 kB
464 KSM: 0 kB
465 LazyFree: 0 kB
466 AnonHugePages: 0 kB
467 FilePmdMapped: 0 kB
468 ShmemPmdMapped: 0 kB
469 Shared_Hugetlb: 0 kB
470 Private_Hugetlb: 0 kB
471 Swap: 0 kB
472 SwapPss: 0 kB
473 Locked: 0 kB
474 THPeligible: 0
475 VmFlags: rd ex mr mw me dw
476
477The first of these lines shows the same information as is displayed for
478the mapping in /proc/PID/maps. Following lines show the size of the
479mapping (size); the smallest possible page size allocated when backing a
480VMA (KernelPageSize), which is the granularity in which VMA modifications
481can be performed; the smallest possible page size that could be used by the
482MMU (MMUPageSize) when backing a VMA; the amount of the mapping that is
483currently resident in RAM (RSS); the process's proportional share of this
484mapping (PSS); and the number of clean and dirty shared and private pages
485in the mapping.
486
487"KernelPageSize" always corresponds to "MMUPageSize", except when a larger
488kernel page size is emulated on a system with a smaller page size used by the
489MMU, which is the case for some PPC64 setups with hugetlb. Furthermore,
490"KernelPageSize" and "MMUPageSize" always correspond to the smallest
491possible granularity (fallback) that can be encountered in a VMA throughout
492its lifetime. These values are not affected by Transparent Huge Pages
493being in effect, or any usage of larger MMU page sizes (either through
494architectural huge-page mappings or other explicit/implicit coalescing of
495virtual ranges performed by the MMU). "AnonHugePages", "ShmemPmdMapped" and
496"FilePmdMapped" provide insight into the usage of PMD-level architectural
497huge-page mappings.
498
499The "proportional set size" (PSS) of a process is the count of pages it has
500in memory, where each page is divided by the number of processes sharing it.
501So if a process has 1000 pages all to itself, and 1000 shared with one other
502process, its PSS will be 1500. "Pss_Dirty" is the portion of PSS which
503consists of dirty pages. ("Pss_Clean" is not included, but it can be
504calculated by subtracting "Pss_Dirty" from "Pss".)
505
506Traditionally, a page is accounted as "private" if it is mapped exactly once,
507and a page is accounted as "shared" when mapped multiple times, even when
508mapped in the same process multiple times. Note that this accounting is
509independent of MAP_SHARED.
510
511In some kernel configurations, the semantics of pages part of a larger
512allocation (e.g., THP) can differ: a page is accounted as "private" if all
513pages part of the corresponding large allocation are *certainly* mapped in the
514same process, even if the page is mapped multiple times in that process. A
515page is accounted as "shared" if any page page of the larger allocation
516is *maybe* mapped in a different process. In some cases, a large allocation
517might be treated as "maybe mapped by multiple processes" even though this
518is no longer the case.
519
520Some kernel configurations do not track the precise number of times a page part
521of a larger allocation is mapped. In this case, when calculating the PSS, the
522average number of mappings per page in this larger allocation might be used
523as an approximation for the number of mappings of a page. The PSS calculation
524will be imprecise in this case.
525
526"Referenced" indicates the amount of memory currently marked as referenced or
527accessed.
528
529"Anonymous" shows the amount of memory that does not belong to any file. Even
530a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
531and a page is modified, the file page is replaced by a private anonymous copy.
532
533"KSM" reports how many of the pages are KSM pages. Note that KSM-placed zeropages
534are not included, only actual KSM pages.
535
536"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE).
537The memory isn't freed immediately with madvise(). It's freed in memory
538pressure if the memory is clean. Please note that the printed value might
539be lower than the real value due to optimizations used in the current
540implementation. If this is not desirable please file a bug report.
541
542"AnonHugePages", "ShmemPmdMapped" and "FilePmdMapped" show the amount of
543memory backed by Transparent Huge Pages that are currently mapped by
544architectural huge-page mappings at the PMD level. "AnonHugePages"
545corresponds to memory that does not belong to a file, "ShmemPmdMapped" to
546shared memory (shmem/tmpfs) and "FilePmdMapped" to file-backed memory
547(excluding shmem/tmpfs).
548
549There are no dedicated entries for Transparent Huge Pages (or similar concepts)
550that are not mapped by architectural huge-page mappings at the PMD level.
551
552"Shared_Hugetlb" and "Private_Hugetlb" show the amounts of memory backed by
553hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
554reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
555
556"Swap" shows how much would-be-anonymous memory is also used, but out on swap.
557
558For shmem mappings, "Swap" includes also the size of the mapped (and not
559replaced by copy-on-write) part of the underlying shmem object out on swap.
560"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this
561does not take into account swapped out page of underlying shmem objects.
562"Locked" indicates whether the mapping is locked in memory or not.
563
564"THPeligible" indicates whether the mapping is eligible for allocating
565naturally aligned THP pages of any currently enabled size. 1 if true, 0
566otherwise.
567
568If both the kernel and the CPU support protection keys (pkeys),
569"ProtectionKey" indicates the memory protection key associated with the
570virtual memory area.
571
572"VmFlags" field deserves a separate description. This member represents the
573kernel flags associated with the particular virtual memory area in two letter
574encoded manner. The codes are the following:
575
576 == =============================================================
577 rd readable
578 wr writeable
579 ex executable
580 sh shared
581 mr may read
582 mw may write
583 me may execute
584 ms may share
585 gd stack segment growns down
586 pf pure PFN range
587 lo pages are locked in memory
588 io memory mapped I/O area
589 sr sequential read advise provided
590 rr random read advise provided
591 dc do not copy area on fork
592 de do not expand area on remapping
593 ac area is accountable
594 nr swap space is not reserved for the area
595 ht area uses huge tlb pages
596 sf synchronous page fault
597 ar architecture specific flag
598 wf wipe on fork
599 dd do not include area into core dump
600 sd soft dirty flag
601 mm mixed map area
602 hg huge page advise flag
603 nh no huge page advise flag
604 mg mergeable advise flag
605 bt arm64 BTI guarded page
606 mt arm64 MTE allocation tags are enabled
607 um userfaultfd missing tracking
608 uw userfaultfd wr-protect tracking
609 ui userfaultfd minor fault
610 ss shadow/guarded control stack page
611 sl sealed
612 lf lock on fault pages
613 dp always lazily freeable mapping
614 gu maybe contains guard regions (if not set, definitely doesn't)
615 == =============================================================
616
617Note that there is no guarantee that every flag and associated mnemonic will
618be present in all further kernel releases. Things get changed, the flags may
619be vanished or the reverse -- new added. Interpretation of their meaning
620might change in future as well. So each consumer of these flags has to
621follow each specific kernel version for the exact semantic.
622
623This file is only present if the CONFIG_MMU kernel configuration option is
624enabled.
625
626Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent
627output can be achieved only in the single read call).
628
629This typically manifests when doing partial reads of these files while the
630memory map is being modified. Despite the races, we do provide the following
631guarantees:
632
6331) The mapped addresses never go backwards, which implies no two
634 regions will ever overlap.
6352) If there is something at a given vaddr during the entirety of the
636 life of the smaps/maps walk, there will be some output for it.
637
638The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps,
639but their values are the sums of the corresponding values for all mappings of
640the process. Additionally, it contains these fields:
641
642- Pss_Anon
643- Pss_File
644- Pss_Shmem
645
646They represent the proportional shares of anonymous, file, and shmem pages, as
647described for smaps above. These fields are omitted in smaps since each
648mapping identifies the type (anon, file, or shmem) of all pages it contains.
649Thus all information in smaps_rollup can be derived from smaps, but at a
650significantly higher cost.
651
652The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
653bits on both physical and virtual pages associated with a process, and the
654soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst
655for details).
656To clear the bits for all the pages associated with the process::
657
658 > echo 1 > /proc/PID/clear_refs
659
660To clear the bits for the anonymous pages associated with the process::
661
662 > echo 2 > /proc/PID/clear_refs
663
664To clear the bits for the file mapped pages associated with the process::
665
666 > echo 3 > /proc/PID/clear_refs
667
668To clear the soft-dirty bit::
669
670 > echo 4 > /proc/PID/clear_refs
671
672To reset the peak resident set size ("high water mark") to the process's
673current value::
674
675 > echo 5 > /proc/PID/clear_refs
676
677Any other value written to /proc/PID/clear_refs will have no effect.
678
679The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
680using /proc/kpageflags and number of times a page is mapped using
681/proc/kpagecount. For detailed explanation, see
682Documentation/admin-guide/mm/pagemap.rst.
683
684The /proc/pid/numa_maps is an extension based on maps, showing the memory
685locality and binding policy, as well as the memory usage (in pages) of
686each mapping. The output follows a general format where mapping details get
687summarized separated by blank spaces, one mapping per each file line::
688
689 address policy mapping details
690
691 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
692 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
693 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
694 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
695 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
696 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
697 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
698 320698b000 default file=/lib64/libc-2.12.so
699 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
700 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
701 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
702 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
703 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
704 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
705 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
706 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
707
708Where:
709
710"address" is the starting address for the mapping;
711
712"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst);
713
714"mapping details" summarizes mapping data such as mapping type, page usage counters,
715node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
716size, in KB, that is backing the mapping up.
717
718Note that some kernel configurations do not track the precise number of times
719a page part of a larger allocation (e.g., THP) is mapped. In these
720configurations, "mapmax" might corresponds to the average number of mappings
721per page in such a larger allocation instead.
722
7231.2 Kernel data
724---------------
725
726Similar to the process entries, the kernel data files give information about
727the running kernel. The files used to obtain this information are contained in
728/proc and are listed in Table 1-5. Not all of these will be present in your
729system. It depends on the kernel configuration and the loaded modules, which
730files are there, and which are missing.
731
732.. table:: Table 1-5: Kernel info in /proc
733
734 ============ ===============================================================
735 File Content
736 ============ ===============================================================
737 allocinfo Memory allocations profiling information
738 apm Advanced power management info
739 bootconfig Kernel command line obtained from boot config,
740 and, if there were kernel parameters from the
741 boot loader, a "# Parameters from bootloader:"
742 line followed by a line containing those
743 parameters prefixed by "# ". (5.5)
744 buddyinfo Kernel memory allocator information (see text) (2.5)
745 bus Directory containing bus specific information
746 cmdline Kernel command line, both from bootloader and embedded
747 in the kernel image
748 cpuinfo Info about the CPU
749 devices Available devices (block and character)
750 dma Used DMA channels
751 filesystems Supported filesystems
752 driver Various drivers grouped here, currently rtc (2.4)
753 execdomains Execdomains, related to security (2.4)
754 fb Frame Buffer devices (2.4)
755 fs File system parameters, currently nfs/exports (2.4)
756 ide Directory containing info about the IDE subsystem
757 interrupts Interrupt usage
758 iomem Memory map (2.4)
759 ioports I/O port usage
760 irq Masks for irq to cpu affinity (2.4)(smp?)
761 isapnp ISA PnP (Plug&Play) Info (2.4)
762 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
763 kmsg Kernel messages
764 ksyms Kernel symbol table
765 loadavg Load average of last 1, 5 & 15 minutes;
766 number of processes currently runnable (running or on ready queue);
767 total number of processes in system;
768 last pid created.
769 All fields are separated by one space except "number of
770 processes currently runnable" and "total number of processes
771 in system", which are separated by a slash ('/'). Example:
772 0.61 0.61 0.55 3/828 22084
773 locks Kernel locks
774 meminfo Memory info
775 misc Miscellaneous
776 modules List of loaded modules
777 mounts Mounted filesystems
778 net Networking info (see text)
779 pagetypeinfo Additional page allocator information (see text) (2.5)
780 partitions Table of partitions known to the system
781 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
782 decoupled by lspci (2.4)
783 rtc Real time clock
784 scsi SCSI info (see text)
785 slabinfo Slab pool info
786 softirqs softirq usage
787 stat Overall statistics
788 swaps Swap space utilization
789 sys See chapter 2
790 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
791 tty Info of tty drivers
792 uptime Wall clock since boot, combined idle time of all cpus
793 version Kernel version
794 video bttv info of video resources (2.4)
795 vmallocinfo Show vmalloced areas
796 ============ ===============================================================
797
798You can, for example, check which interrupts are currently in use and what
799they are used for by looking in the file /proc/interrupts::
800
801 > cat /proc/interrupts
802 CPU0
803 0: 8728810 XT-PIC timer
804 1: 895 XT-PIC keyboard
805 2: 0 XT-PIC cascade
806 3: 531695 XT-PIC aha152x
807 4: 2014133 XT-PIC serial
808 5: 44401 XT-PIC pcnet_cs
809 8: 2 XT-PIC rtc
810 11: 8 XT-PIC i82365
811 12: 182918 XT-PIC PS/2 Mouse
812 13: 1 XT-PIC fpu
813 14: 1232265 XT-PIC ide0
814 15: 7 XT-PIC ide1
815 NMI: 0
816
817In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
818output of a SMP machine)::
819
820 > cat /proc/interrupts
821
822 CPU0 CPU1
823 0: 1243498 1214548 IO-APIC-edge timer
824 1: 8949 8958 IO-APIC-edge keyboard
825 2: 0 0 XT-PIC cascade
826 5: 11286 10161 IO-APIC-edge soundblaster
827 8: 1 0 IO-APIC-edge rtc
828 9: 27422 27407 IO-APIC-edge 3c503
829 12: 113645 113873 IO-APIC-edge PS/2 Mouse
830 13: 0 0 XT-PIC fpu
831 14: 22491 24012 IO-APIC-edge ide0
832 15: 2183 2415 IO-APIC-edge ide1
833 17: 30564 30414 IO-APIC-level eth0
834 18: 177 164 IO-APIC-level bttv
835 NMI: 2457961 2457959
836 LOC: 2457882 2457881
837 ERR: 2155
838
839NMI is incremented in this case because every timer interrupt generates a NMI
840(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
841
842LOC is the local interrupt counter of the internal APIC of every CPU.
843
844ERR is incremented in the case of errors in the IO-APIC bus (the bus that
845connects the CPUs in a SMP system. This means that an error has been detected,
846the IO-APIC automatically retry the transmission, so it should not be a big
847problem, but you should read the SMP-FAQ.
848
849In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
850/proc/interrupts to display every IRQ vector in use by the system, not
851just those considered 'most important'. The new vectors are:
852
853THR
854 interrupt raised when a machine check threshold counter
855 (typically counting ECC corrected errors of memory or cache) exceeds
856 a configurable threshold. Only available on some systems.
857
858TRM
859 a thermal event interrupt occurs when a temperature threshold
860 has been exceeded for the CPU. This interrupt may also be generated
861 when the temperature drops back to normal.
862
863SPU
864 a spurious interrupt is some interrupt that was raised then lowered
865 by some IO device before it could be fully processed by the APIC. Hence
866 the APIC sees the interrupt but does not know what device it came from.
867 For this case the APIC will generate the interrupt with a IRQ vector
868 of 0xff. This might also be generated by chipset bugs.
869
870RES, CAL, TLB
871 rescheduling, call and TLB flush interrupts are
872 sent from one CPU to another per the needs of the OS. Typically,
873 their statistics are used by kernel developers and interested users to
874 determine the occurrence of interrupts of the given type.
875
876The above IRQ vectors are displayed only when relevant. For example,
877the threshold vector does not exist on x86_64 platforms. Others are
878suppressed when the system is a uniprocessor. As of this writing, only
879i386 and x86_64 platforms support the new IRQ vector displays.
880
881Of some interest is the introduction of the /proc/irq directory to 2.4.
882It could be used to set IRQ to CPU affinity. This means that you can "hook" an
883IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
884irq subdir is one subdir for each IRQ, and default_smp_affinity.
885
886For example::
887
888 > ls /proc/irq/
889 0 10 12 14 16 18 2 4 6 8 default_smp_affinity
890 1 11 13 15 17 19 3 5 7 9
891 > ls /proc/irq/0/
892 smp_affinity
893
894smp_affinity is a bitmask, in which you can specify which CPUs can handle the
895IRQ. You can set it by doing::
896
897 > echo 1 > /proc/irq/10/smp_affinity
898
899This means that only the first CPU will handle the IRQ, but you can also echo
9005 which means that only the first and third CPU can handle the IRQ.
901
902The contents of each smp_affinity file is the same by default::
903
904 > cat /proc/irq/0/smp_affinity
905 ffffffff
906
907There is an alternate interface, smp_affinity_list which allows specifying
908a CPU range instead of a bitmask::
909
910 > cat /proc/irq/0/smp_affinity_list
911 1024-1031
912
913The default_smp_affinity mask applies to all non-active IRQs, which are the
914IRQs which have not yet been allocated/activated, and hence which lack a
915/proc/irq/[0-9]* directory.
916
917The node file on an SMP system shows the node to which the device using the IRQ
918reports itself as being attached. This hardware locality information does not
919include information about any possible driver locality preference.
920
921The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
922between all the CPUs which are allowed to handle it. As usual the kernel has
923more info than you and does a better job than you, so the defaults are the
924best choice for almost everyone. [Note this applies only to those IO-APIC's
925that support "Round Robin" interrupt distribution.]
926
927There are three more important subdirectories in /proc: net, scsi, and sys.
928The general rule is that the contents, or even the existence of these
929directories, depend on your kernel configuration. If SCSI is not enabled, the
930directory scsi may not exist. The same is true with the net, which is there
931only when networking support is present in the running kernel.
932
933The slabinfo file gives information about memory usage at the slab level.
934Linux uses slab pools for memory management above page level in version 2.2.
935Commonly used objects have their own slab pool (such as network buffers,
936directory cache, and so on).
937
938::
939
940 > cat /proc/buddyinfo
941
942 Node 0, zone DMA 0 4 5 4 4 3 ...
943 Node 0, zone Normal 1 0 0 1 101 8 ...
944 Node 0, zone HighMem 2 0 0 1 1 0 ...
945
946External fragmentation is a problem under some workloads, and buddyinfo is a
947useful tool for helping diagnose these problems. Buddyinfo will give you a
948clue as to how big an area you can safely allocate, or why a previous
949allocation failed.
950
951Each column represents the number of pages of a certain order which are
952available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
953ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
954available in ZONE_NORMAL, etc...
955
956More information relevant to external fragmentation can be found in
957pagetypeinfo::
958
959 > cat /proc/pagetypeinfo
960 Page block order: 9
961 Pages per block: 512
962
963 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
964 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
965 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
966 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
967 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
968 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
969 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
970 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
971 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
972 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
973 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
974
975 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
976 Node 0, zone DMA 2 0 5 1 0
977 Node 0, zone DMA32 41 6 967 2 0
978
979Fragmentation avoidance in the kernel works by grouping pages of different
980migrate types into the same contiguous regions of memory called page blocks.
981A page block is typically the size of the default hugepage size, e.g. 2MB on
982X86-64. By keeping pages grouped based on their ability to move, the kernel
983can reclaim pages within a page block to satisfy a high-order allocation.
984
985The pagetypinfo begins with information on the size of a page block. It
986then gives the same type of information as buddyinfo except broken down
987by migrate-type and finishes with details on how many page blocks of each
988type exist.
989
990If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
991from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can
992make an estimate of the likely number of huge pages that can be allocated
993at a given point in time. All the "Movable" blocks should be allocatable
994unless memory has been mlock()'d. Some of the Reclaimable blocks should
995also be allocatable although a lot of filesystem metadata may have to be
996reclaimed to achieve this.
997
998
999allocinfo
1000~~~~~~~~~
1001
1002Provides information about memory allocations at all locations in the code
1003base. Each allocation in the code is identified by its source file, line
1004number, module (if originates from a loadable module) and the function calling
1005the allocation. The number of bytes allocated and number of calls at each
1006location are reported. The first line indicates the version of the file, the
1007second line is the header listing fields in the file.
1008If file version is 2.0 or higher then each line may contain additional
1009<key>:<value> pairs representing extra information about the call site.
1010For example if the counters are not accurate, the line will be appended with
1011"accurate:no" pair.
1012
1013Supported markers in v2:
1014accurate:no
1015
1016 Absolute values of the counters in this line are not accurate
1017 because of the failure to allocate memory to track some of the
1018 allocations made at this location. Deltas in these counters are
1019 accurate, therefore counters can be used to track allocation size
1020 and count changes.
1021
1022Example output.
1023
1024::
1025
1026 > tail -n +3 /proc/allocinfo | sort -rn
1027 127664128 31168 mm/page_ext.c:270 func:alloc_page_ext
1028 56373248 4737 mm/slub.c:2259 func:alloc_slab_page
1029 14880768 3633 mm/readahead.c:247 func:page_cache_ra_unbounded
1030 14417920 3520 mm/mm_init.c:2530 func:alloc_large_system_hash
1031 13377536 234 block/blk-mq.c:3421 func:blk_mq_alloc_rqs
1032 11718656 2861 mm/filemap.c:1919 func:__filemap_get_folio
1033 9192960 2800 kernel/fork.c:307 func:alloc_thread_stack_node
1034 4206592 4 net/netfilter/nf_conntrack_core.c:2567 func:nf_ct_alloc_hashtable
1035 4136960 1010 drivers/staging/ctagmod/ctagmod.c:20 [ctagmod] func:ctagmod_start
1036 3940352 962 mm/memory.c:4214 func:alloc_anon_folio
1037 2894464 22613 fs/kernfs/dir.c:615 func:__kernfs_new_node
1038 ...
1039
1040
1041meminfo
1042~~~~~~~
1043
1044Provides information about distribution and utilization of memory. This
1045varies by architecture and compile options. Some of the counters reported
1046here overlap. The memory reported by the non overlapping counters may not
1047add up to the overall memory usage and the difference for some workloads
1048can be substantial. In many cases there are other means to find out
1049additional memory using subsystem specific interfaces, for instance
1050/proc/net/sockstat for TCP memory allocations.
1051
1052Example output. You may not have all of these fields.
1053
1054::
1055
1056 > cat /proc/meminfo
1057
1058 MemTotal: 32858820 kB
1059 MemFree: 21001236 kB
1060 MemAvailable: 27214312 kB
1061 Buffers: 581092 kB
1062 Cached: 5587612 kB
1063 SwapCached: 0 kB
1064 Active: 3237152 kB
1065 Inactive: 7586256 kB
1066 Active(anon): 94064 kB
1067 Inactive(anon): 4570616 kB
1068 Active(file): 3143088 kB
1069 Inactive(file): 3015640 kB
1070 Unevictable: 0 kB
1071 Mlocked: 0 kB
1072 SwapTotal: 0 kB
1073 SwapFree: 0 kB
1074 Zswap: 1904 kB
1075 Zswapped: 7792 kB
1076 Dirty: 12 kB
1077 Writeback: 0 kB
1078 AnonPages: 4654780 kB
1079 Mapped: 266244 kB
1080 Shmem: 9976 kB
1081 KReclaimable: 517708 kB
1082 Slab: 660044 kB
1083 SReclaimable: 517708 kB
1084 SUnreclaim: 142336 kB
1085 KernelStack: 11168 kB
1086 PageTables: 20540 kB
1087 SecPageTables: 0 kB
1088 NFS_Unstable: 0 kB
1089 Bounce: 0 kB
1090 WritebackTmp: 0 kB
1091 CommitLimit: 16429408 kB
1092 Committed_AS: 7715148 kB
1093 VmallocTotal: 34359738367 kB
1094 VmallocUsed: 40444 kB
1095 VmallocChunk: 0 kB
1096 Percpu: 29312 kB
1097 EarlyMemtestBad: 0 kB
1098 HardwareCorrupted: 0 kB
1099 AnonHugePages: 4149248 kB
1100 ShmemHugePages: 0 kB
1101 ShmemPmdMapped: 0 kB
1102 FileHugePages: 0 kB
1103 FilePmdMapped: 0 kB
1104 CmaTotal: 0 kB
1105 CmaFree: 0 kB
1106 Unaccepted: 0 kB
1107 Balloon: 0 kB
1108 GPUActive: 0 kB
1109 GPUReclaim: 0 kB
1110 HugePages_Total: 0
1111 HugePages_Free: 0
1112 HugePages_Rsvd: 0
1113 HugePages_Surp: 0
1114 Hugepagesize: 2048 kB
1115 Hugetlb: 0 kB
1116 DirectMap4k: 401152 kB
1117 DirectMap2M: 10008576 kB
1118 DirectMap1G: 24117248 kB
1119
1120MemTotal
1121 Total usable RAM (i.e. physical RAM minus a few reserved
1122 bits and the kernel binary code)
1123MemFree
1124 Total free RAM. On highmem systems, the sum of LowFree+HighFree
1125MemAvailable
1126 An estimate of how much memory is available for starting new
1127 applications, without swapping. Calculated from MemFree,
1128 SReclaimable, the size of the file LRU lists, and the low
1129 watermarks in each zone.
1130 The estimate takes into account that the system needs some
1131 page cache to function well, and that not all reclaimable
1132 slab will be reclaimable, due to items being in use. The
1133 impact of those factors will vary from system to system.
1134Buffers
1135 Relatively temporary storage for raw disk blocks
1136 shouldn't get tremendously large (20MB or so)
1137Cached
1138 In-memory cache for files read from the disk (the
1139 pagecache) as well as tmpfs & shmem.
1140 Doesn't include SwapCached.
1141SwapCached
1142 Memory that once was swapped out, is swapped back in but
1143 still also is in the swapfile (if memory is needed it
1144 doesn't need to be swapped out AGAIN because it is already
1145 in the swapfile. This saves I/O)
1146Active
1147 Memory that has been used more recently and usually not
1148 reclaimed unless absolutely necessary.
1149Inactive
1150 Memory which has been less recently used. It is more
1151 eligible to be reclaimed for other purposes
1152Unevictable
1153 Memory allocated for userspace which cannot be reclaimed, such
1154 as mlocked pages, ramfs backing pages, secret memfd pages etc.
1155Mlocked
1156 Memory locked with mlock().
1157HighTotal, HighFree
1158 Highmem is all memory above ~860MB of physical memory.
1159 Highmem areas are for use by userspace programs, or
1160 for the pagecache. The kernel must use tricks to access
1161 this memory, making it slower to access than lowmem.
1162LowTotal, LowFree
1163 Lowmem is memory which can be used for everything that
1164 highmem can be used for, but it is also available for the
1165 kernel's use for its own data structures. Among many
1166 other things, it is where everything from the Slab is
1167 allocated. Bad things happen when you're out of lowmem.
1168SwapTotal
1169 total amount of swap space available
1170SwapFree
1171 Memory which has been evicted from RAM, and is temporarily
1172 on the disk
1173Zswap
1174 Memory consumed by the zswap backend (compressed size)
1175Zswapped
1176 Amount of anonymous memory stored in zswap (original size)
1177Dirty
1178 Memory which is waiting to get written back to the disk
1179Writeback
1180 Memory which is actively being written back to the disk
1181AnonPages
1182 Non-file backed pages mapped into userspace page tables. Note that
1183 some kernel configurations might consider all pages part of a
1184 larger allocation (e.g., THP) as "mapped", as soon as a single
1185 page is mapped.
1186Mapped
1187 files which have been mmapped, such as libraries. Note that some
1188 kernel configurations might consider all pages part of a larger
1189 allocation (e.g., THP) as "mapped", as soon as a single page is
1190 mapped.
1191Shmem
1192 Total memory used by shared memory (shmem) and tmpfs
1193KReclaimable
1194 Kernel allocations that the kernel will attempt to reclaim
1195 under memory pressure. Includes SReclaimable (below), and other
1196 direct allocations with a shrinker.
1197Slab
1198 in-kernel data structures cache
1199SReclaimable
1200 Part of Slab, that might be reclaimed, such as caches
1201SUnreclaim
1202 Part of Slab, that cannot be reclaimed on memory pressure
1203KernelStack
1204 Memory consumed by the kernel stacks of all tasks
1205PageTables
1206 Memory consumed by userspace page tables
1207SecPageTables
1208 Memory consumed by secondary page tables, this currently includes
1209 KVM mmu and IOMMU allocations on x86 and arm64.
1210NFS_Unstable
1211 Always zero. Previously counted pages which had been written to
1212 the server, but has not been committed to stable storage.
1213Bounce
1214 Always zero. Previously memory used for block device
1215 "bounce buffers".
1216WritebackTmp
1217 Always zero. Previously memory used by FUSE for temporary
1218 writeback buffers.
1219CommitLimit
1220 Based on the overcommit ratio ('vm.overcommit_ratio'),
1221 this is the total amount of memory currently available to
1222 be allocated on the system. This limit is only adhered to
1223 if strict overcommit accounting is enabled (mode 2 in
1224 'vm.overcommit_memory').
1225
1226 The CommitLimit is calculated with the following formula::
1227
1228 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
1229 overcommit_ratio / 100 + [total swap pages]
1230
1231 For example, on a system with 1G of physical RAM and 7G
1232 of swap with a `vm.overcommit_ratio` of 30 it would
1233 yield a CommitLimit of 7.3G.
1234
1235 For more details, see the memory overcommit documentation
1236 in mm/overcommit-accounting.
1237Committed_AS
1238 The amount of memory presently allocated on the system.
1239 The committed memory is a sum of all of the memory which
1240 has been allocated by processes, even if it has not been
1241 "used" by them as of yet. A process which malloc()'s 1G
1242 of memory, but only touches 300M of it will show up as
1243 using 1G. This 1G is memory which has been "committed" to
1244 by the VM and can be used at any time by the allocating
1245 application. With strict overcommit enabled on the system
1246 (mode 2 in 'vm.overcommit_memory'), allocations which would
1247 exceed the CommitLimit (detailed above) will not be permitted.
1248 This is useful if one needs to guarantee that processes will
1249 not fail due to lack of memory once that memory has been
1250 successfully allocated.
1251VmallocTotal
1252 total size of vmalloc virtual address space
1253VmallocUsed
1254 amount of vmalloc area which is used
1255VmallocChunk
1256 largest contiguous block of vmalloc area which is free
1257Percpu
1258 Memory allocated to the percpu allocator used to back percpu
1259 allocations. This stat excludes the cost of metadata.
1260EarlyMemtestBad
1261 The amount of RAM/memory in kB, that was identified as corrupted
1262 by early memtest. If memtest was not run, this field will not
1263 be displayed at all. Size is never rounded down to 0 kB.
1264 That means if 0 kB is reported, you can safely assume
1265 there was at least one pass of memtest and none of the passes
1266 found a single faulty byte of RAM.
1267HardwareCorrupted
1268 The amount of RAM/memory in KB, the kernel identifies as
1269 corrupted.
1270AnonHugePages
1271 Non-file backed huge pages mapped into userspace page tables
1272ShmemHugePages
1273 Memory used by shared memory (shmem) and tmpfs allocated
1274 with huge pages
1275ShmemPmdMapped
1276 Shared memory mapped into userspace with huge pages
1277FileHugePages
1278 Memory used for filesystem data (page cache) allocated
1279 with huge pages
1280FilePmdMapped
1281 Page cache mapped into userspace with huge pages
1282CmaTotal
1283 Memory reserved for the Contiguous Memory Allocator (CMA)
1284CmaFree
1285 Free remaining memory in the CMA reserves
1286Unaccepted
1287 Memory that has not been accepted by the guest
1288Balloon
1289 Memory returned to Host by VM Balloon Drivers
1290GPUActive
1291 System memory allocated to active GPU objects
1292GPUReclaim
1293 System memory stored in GPU pools for reuse. This memory is not
1294 counted in GPUActive. It is shrinker reclaimable memory kept in a reuse
1295 pool because it has non-standard page table attributes, like WC or UC.
1296HugePages_Total, HugePages_Free, HugePages_Rsvd, HugePages_Surp, Hugepagesize, Hugetlb
1297 See Documentation/admin-guide/mm/hugetlbpage.rst.
1298DirectMap4k, DirectMap2M, DirectMap1G
1299 Breakdown of page table sizes used in the kernel's
1300 identity mapping of RAM
1301
1302vmallocinfo
1303~~~~~~~~~~~
1304
1305Provides information about vmalloced/vmaped areas. One line per area,
1306containing the virtual address range of the area, size in bytes,
1307caller information of the creator, and optional information depending
1308on the kind of area:
1309
1310 ========== ===================================================
1311 pages=nr number of pages
1312 phys=addr if a physical address was specified
1313 ioremap I/O mapping (ioremap() and friends)
1314 vmalloc vmalloc() area
1315 vmap vmap()ed pages
1316 user VM_USERMAP area
1317 vpages buffer for pages pointers was vmalloced (huge area)
1318 N<node>=nr (Only on NUMA kernels)
1319 Number of pages allocated on memory node <node>
1320 ========== ===================================================
1321
1322::
1323
1324 > cat /proc/vmallocinfo
1325 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
1326 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
1327 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
1328 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
1329 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
1330 phys=7fee8000 ioremap
1331 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
1332 phys=7fee7000 ioremap
1333 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
1334 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
1335 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
1336 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
1337 pages=2 vmalloc N1=2
1338 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
1339 /0x130 [x_tables] pages=4 vmalloc N0=4
1340 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
1341 pages=14 vmalloc N2=14
1342 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
1343 pages=4 vmalloc N1=4
1344 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
1345 pages=2 vmalloc N1=2
1346 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
1347 pages=10 vmalloc N0=10
1348
1349
1350softirqs
1351~~~~~~~~
1352
1353Provides counts of softirq handlers serviced since boot time, for each CPU.
1354
1355::
1356
1357 > cat /proc/softirqs
1358 CPU0 CPU1 CPU2 CPU3
1359 HI: 0 0 0 0
1360 TIMER: 27166 27120 27097 27034
1361 NET_TX: 0 0 0 17
1362 NET_RX: 42 0 0 39
1363 BLOCK: 0 0 107 1121
1364 TASKLET: 0 0 0 290
1365 SCHED: 27035 26983 26971 26746
1366 HRTIMER: 0 0 0 0
1367 RCU: 1678 1769 2178 2250
1368
13691.3 Networking info in /proc/net
1370--------------------------------
1371
1372The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1373additional values you get for IP version 6 if you configure the kernel to
1374support this. Table 1-9 lists the files and their meaning.
1375
1376
1377.. table:: Table 1-8: IPv6 info in /proc/net
1378
1379 ========== =====================================================
1380 File Content
1381 ========== =====================================================
1382 udp6 UDP sockets (IPv6)
1383 tcp6 TCP sockets (IPv6)
1384 raw6 Raw device statistics (IPv6)
1385 igmp6 IP multicast addresses, which this host joined (IPv6)
1386 if_inet6 List of IPv6 interface addresses
1387 ipv6_route Kernel routing table for IPv6
1388 rt6_stats Global IPv6 routing tables statistics
1389 sockstat6 Socket statistics (IPv6)
1390 snmp6 Snmp data (IPv6)
1391 ========== =====================================================
1392
1393.. table:: Table 1-9: Network info in /proc/net
1394
1395 ============= ================================================================
1396 File Content
1397 ============= ================================================================
1398 arp Kernel ARP table
1399 dev network devices with statistics
1400 dev_mcast the Layer2 multicast groups a device is listening too
1401 (interface index, label, number of references, number of bound
1402 addresses).
1403 dev_stat network device status
1404 ip_fwchains Firewall chain linkage
1405 ip_fwnames Firewall chain names
1406 ip_masq Directory containing the masquerading tables
1407 ip_masquerade Major masquerading table
1408 netstat Network statistics
1409 raw raw device statistics
1410 route Kernel routing table
1411 rpc Directory containing rpc info
1412 rt_cache Routing cache
1413 snmp SNMP data
1414 sockstat Socket statistics
1415 softnet_stat Per-CPU incoming packets queues statistics of online CPUs
1416 tcp TCP sockets
1417 udp UDP sockets
1418 unix UNIX domain sockets
1419 wireless Wireless interface data (Wavelan etc)
1420 igmp IP multicast addresses, which this host joined
1421 psched Global packet scheduler parameters.
1422 netlink List of PF_NETLINK sockets
1423 ip_mr_vifs List of multicast virtual interfaces
1424 ip_mr_cache List of multicast routing cache
1425 ============= ================================================================
1426
1427You can use this information to see which network devices are available in
1428your system and how much traffic was routed over those devices::
1429
1430 > cat /proc/net/dev
1431 Inter-|Receive |[...
1432 face |bytes packets errs drop fifo frame compressed multicast|[...
1433 lo: 908188 5596 0 0 0 0 0 0 [...
1434 ppp0:15475140 20721 410 0 0 410 0 0 [...
1435 eth0: 614530 7085 0 0 0 0 0 1 [...
1436
1437 ...] Transmit
1438 ...] bytes packets errs drop fifo colls carrier compressed
1439 ...] 908188 5596 0 0 0 0 0 0
1440 ...] 1375103 17405 0 0 0 0 0 0
1441 ...] 1703981 5535 0 0 0 3 0 0
1442
1443In addition, each Channel Bond interface has its own directory. For
1444example, the bond0 device will have a directory called /proc/net/bond0/.
1445It will contain information that is specific to that bond, such as the
1446current slaves of the bond, the link status of the slaves, and how
1447many times the slaves link has failed.
1448
14491.4 SCSI info
1450-------------
1451
1452If you have a SCSI or ATA host adapter in your system, you'll find a
1453subdirectory named after the driver for this adapter in /proc/scsi.
1454You'll also see a list of all recognized SCSI devices in /proc/scsi::
1455
1456 >cat /proc/scsi/scsi
1457 Attached devices:
1458 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1459 Vendor: IBM Model: DGHS09U Rev: 03E0
1460 Type: Direct-Access ANSI SCSI revision: 03
1461 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1462 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1463 Type: CD-ROM ANSI SCSI revision: 02
1464
1465
1466The directory named after the driver has one file for each adapter found in
1467the system. These files contain information about the controller, including
1468the used IRQ and the IO address range. The amount of information shown is
1469dependent on the adapter you use. The example shows the output for an Adaptec
1470AHA-2940 SCSI adapter::
1471
1472 > cat /proc/scsi/aic7xxx/0
1473
1474 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1475 Compile Options:
1476 TCQ Enabled By Default : Disabled
1477 AIC7XXX_PROC_STATS : Disabled
1478 AIC7XXX_RESET_DELAY : 5
1479 Adapter Configuration:
1480 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1481 Ultra Wide Controller
1482 PCI MMAPed I/O Base: 0xeb001000
1483 Adapter SEEPROM Config: SEEPROM found and used.
1484 Adaptec SCSI BIOS: Enabled
1485 IRQ: 10
1486 SCBs: Active 0, Max Active 2,
1487 Allocated 15, HW 16, Page 255
1488 Interrupts: 160328
1489 BIOS Control Word: 0x18b6
1490 Adapter Control Word: 0x005b
1491 Extended Translation: Enabled
1492 Disconnect Enable Flags: 0xffff
1493 Ultra Enable Flags: 0x0001
1494 Tag Queue Enable Flags: 0x0000
1495 Ordered Queue Tag Flags: 0x0000
1496 Default Tag Queue Depth: 8
1497 Tagged Queue By Device array for aic7xxx host instance 0:
1498 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1499 Actual queue depth per device for aic7xxx host instance 0:
1500 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1501 Statistics:
1502 (scsi0:0:0:0)
1503 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1504 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1505 Total transfers 160151 (74577 reads and 85574 writes)
1506 (scsi0:0:6:0)
1507 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1508 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1509 Total transfers 0 (0 reads and 0 writes)
1510
1511
15121.5 Parallel port info in /proc/parport
1513---------------------------------------
1514
1515The directory /proc/parport contains information about the parallel ports of
1516your system. It has one subdirectory for each port, named after the port
1517number (0,1,2,...).
1518
1519These directories contain the four files shown in Table 1-10.
1520
1521
1522.. table:: Table 1-10: Files in /proc/parport
1523
1524 ========= ====================================================================
1525 File Content
1526 ========= ====================================================================
1527 autoprobe Any IEEE-1284 device ID information that has been acquired.
1528 devices list of the device drivers using that port. A + will appear by the
1529 name of the device currently using the port (it might not appear
1530 against any).
1531 hardware Parallel port's base address, IRQ line and DMA channel.
1532 irq IRQ that parport is using for that port. This is in a separate
1533 file to allow you to alter it by writing a new value in (IRQ
1534 number or none).
1535 ========= ====================================================================
1536
15371.6 TTY info in /proc/tty
1538-------------------------
1539
1540Information about the available and actually used tty's can be found in the
1541directory /proc/tty. You'll find entries for drivers and line disciplines in
1542this directory, as shown in Table 1-11.
1543
1544
1545.. table:: Table 1-11: Files in /proc/tty
1546
1547 ============= ==============================================
1548 File Content
1549 ============= ==============================================
1550 drivers list of drivers and their usage
1551 ldiscs registered line disciplines
1552 driver/serial usage statistic and status of single tty lines
1553 ============= ==============================================
1554
1555To see which tty's are currently in use, you can simply look into the file
1556/proc/tty/drivers::
1557
1558 > cat /proc/tty/drivers
1559 pty_slave /dev/pts 136 0-255 pty:slave
1560 pty_master /dev/ptm 128 0-255 pty:master
1561 pty_slave /dev/ttyp 3 0-255 pty:slave
1562 pty_master /dev/pty 2 0-255 pty:master
1563 serial /dev/cua 5 64-67 serial:callout
1564 serial /dev/ttyS 4 64-67 serial
1565 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1566 /dev/ptmx /dev/ptmx 5 2 system
1567 /dev/console /dev/console 5 1 system:console
1568 /dev/tty /dev/tty 5 0 system:/dev/tty
1569 unknown /dev/tty 4 1-63 console
1570
1571
15721.7 Miscellaneous kernel statistics in /proc/stat
1573-------------------------------------------------
1574
1575Various pieces of information about kernel activity are available in the
1576/proc/stat file. All of the numbers reported in this file are aggregates
1577since the system first booted. For a quick look, simply cat the file::
1578
1579 > cat /proc/stat
1580 cpu 237902850 368826709 106375398 1873517540 1135548 0 14507935 0 0 0
1581 cpu0 60045249 91891769 26331539 468411416 495718 0 5739640 0 0 0
1582 cpu1 59746288 91759249 26609887 468860630 312281 0 4384817 0 0 0
1583 cpu2 59489247 92985423 26904446 467808813 171668 0 2268998 0 0 0
1584 cpu3 58622065 92190267 26529524 468436680 155879 0 2114478 0 0 0
1585 intr 8688370575 8 3373 0 0 0 0 0 0 1 40791 0 0 353317 0 0 0 0 224789828 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 190974333 41958554 123983334 43 0 224593 0 0 0 <more 0's deleted>
1586 ctxt 22848221062
1587 btime 1605316999
1588 processes 746787147
1589 procs_running 2
1590 procs_blocked 0
1591 softirq 12121874454 100099120 3938138295 127375644 2795979 187870761 0 173808342 3072582055 52608 224184354
1592
1593The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1594lines. These numbers identify the amount of time the CPU has spent performing
1595different kinds of work. Time units are in USER_HZ (typically hundredths of a
1596second). The meanings of the columns are as follows, from left to right:
1597
1598- user: normal processes executing in user mode
1599- nice: niced processes executing in user mode
1600- system: processes executing in kernel mode
1601- idle: twiddling thumbs
1602- iowait: In a word, iowait stands for waiting for I/O to complete. But there
1603 are several problems:
1604
1605 1. CPU will not wait for I/O to complete, iowait is the time that a task is
1606 waiting for I/O to complete. When CPU goes into idle state for
1607 outstanding task I/O, another task will be scheduled on this CPU.
1608 2. In a multi-core CPU, the task waiting for I/O to complete is not running
1609 on any CPU, so the iowait of each CPU is difficult to calculate.
1610 3. The value of iowait field in /proc/stat will decrease in certain
1611 conditions.
1612
1613 So, the iowait is not reliable by reading from /proc/stat.
1614- irq: servicing interrupts
1615- softirq: servicing softirqs
1616- steal: involuntary wait
1617- guest: running a normal guest
1618- guest_nice: running a niced guest
1619
1620The "intr" line gives counts of interrupts serviced since boot time, for each
1621of the possible system interrupts. The first column is the total of all
1622interrupts serviced including unnumbered architecture specific interrupts;
1623each subsequent column is the total for that particular numbered interrupt.
1624Unnumbered interrupts are not shown, only summed into the total.
1625
1626The "ctxt" line gives the total number of context switches across all CPUs.
1627
1628The "btime" line gives the time at which the system booted, in seconds since
1629the Unix epoch.
1630
1631The "processes" line gives the number of processes and threads created, which
1632includes (but is not limited to) those created by calls to the fork() and
1633clone() system calls.
1634
1635The "procs_running" line gives the total number of threads that are
1636running or ready to run (i.e., the total number of runnable threads).
1637
1638The "procs_blocked" line gives the number of processes currently blocked,
1639waiting for I/O to complete.
1640
1641The "softirq" line gives counts of softirqs serviced since boot time, for each
1642of the possible system softirqs. The first column is the total of all
1643softirqs serviced; each subsequent column is the total for that particular
1644softirq.
1645
1646
16471.8 Ext4 file system parameters
1648-------------------------------
1649
1650Information about mounted ext4 file systems can be found in
1651/proc/fs/ext4. Each mounted filesystem will have a directory in
1652/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1653/proc/fs/ext4/sda9 or /proc/fs/ext4/dm-0). The files in each per-device
1654directory are shown in Table 1-12, below.
1655
1656.. table:: Table 1-12: Files in /proc/fs/ext4/<devname>
1657
1658 ============== ==========================================================
1659 File Content
1660 mb_groups details of multiblock allocator buddy cache of free blocks
1661 ============== ==========================================================
1662
16631.9 /proc/consoles
1664-------------------
1665Shows registered system console lines.
1666
1667To see which character device lines are currently used for the system console
1668/dev/console, you may simply look into the file /proc/consoles::
1669
1670 > cat /proc/consoles
1671 tty0 -WU (ECp) 4:7
1672 ttyS0 -W- (Ep) 4:64
1673
1674The columns are:
1675
1676+--------------------+-------------------------------------------------------+
1677| device | name of the device |
1678+====================+=======================================================+
1679| operations | * R = can do read operations |
1680| | * W = can do write operations |
1681| | * U = can do unblank |
1682+--------------------+-------------------------------------------------------+
1683| flags | * E = it is enabled |
1684| | * C = it is preferred console |
1685| | * B = it is primary boot console |
1686| | * p = it is used for printk buffer |
1687| | * b = it is not a TTY but a Braille device |
1688| | * a = it is safe to use when cpu is offline |
1689+--------------------+-------------------------------------------------------+
1690| major:minor | major and minor number of the device separated by a |
1691| | colon |
1692+--------------------+-------------------------------------------------------+
1693
1694Summary
1695-------
1696
1697The /proc file system serves information about the running system. It not only
1698allows access to process data but also allows you to request the kernel status
1699by reading files in the hierarchy.
1700
1701The directory structure of /proc reflects the types of information and makes
1702it easy, if not obvious, where to look for specific data.
1703
1704Chapter 2: Modifying System Parameters
1705======================================
1706
1707In This Chapter
1708---------------
1709
1710* Modifying kernel parameters by writing into files found in /proc/sys
1711* Exploring the files which modify certain parameters
1712* Review of the /proc/sys file tree
1713
1714------------------------------------------------------------------------------
1715
1716A very interesting part of /proc is the directory /proc/sys. This is not only
1717a source of information, it also allows you to change parameters within the
1718kernel. Be very careful when attempting this. You can optimize your system,
1719but you can also cause it to crash. Never alter kernel parameters on a
1720production system. Set up a development machine and test to make sure that
1721everything works the way you want it to. You may have no alternative but to
1722reboot the machine once an error has been made.
1723
1724To change a value, simply echo the new value into the file.
1725You need to be root to do this. You can create your own boot script
1726to perform this every time your system boots.
1727
1728The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1729general things in the operation of the Linux kernel. Since some of the files
1730can inadvertently disrupt your system, it is advisable to read both
1731documentation and source before actually making adjustments. In any case, be
1732very careful when writing to any of these files. The entries in /proc may
1733change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1734review the kernel documentation in the directory linux/Documentation.
1735This chapter is heavily based on the documentation included in the pre 2.2
1736kernels, and became part of it in version 2.2.1 of the Linux kernel.
1737
1738Please see: Documentation/admin-guide/sysctl/ directory for descriptions of
1739these entries.
1740
1741Summary
1742-------
1743
1744Certain aspects of kernel behavior can be modified at runtime, without the
1745need to recompile the kernel, or even to reboot the system. The files in the
1746/proc/sys tree can not only be read, but also modified. You can use the echo
1747command to write value into these files, thereby changing the default settings
1748of the kernel.
1749
1750
1751Chapter 3: Per-process Parameters
1752=================================
1753
17543.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1755--------------------------------------------------------------------------------
1756
1757These files can be used to adjust the badness heuristic used to select which
1758process gets killed in out of memory (oom) conditions.
1759
1760The badness heuristic assigns a value to each candidate task ranging from 0
1761(never kill) to 1000 (always kill) to determine which process is targeted. The
1762units are roughly a proportion along that range of allowed memory the process
1763may allocate from based on an estimation of its current memory and swap use.
1764For example, if a task is using all allowed memory, its badness score will be
17651000. If it is using half of its allowed memory, its score will be 500.
1766
1767The amount of "allowed" memory depends on the context in which the oom killer
1768was called. If it is due to the memory assigned to the allocating task's cpuset
1769being exhausted, the allowed memory represents the set of mems assigned to that
1770cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1771memory represents the set of mempolicy nodes. If it is due to a memory
1772limit (or swap limit) being reached, the allowed memory is that configured
1773limit. Finally, if it is due to the entire system being out of memory, the
1774allowed memory represents all allocatable resources.
1775
1776The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1777is used to determine which task to kill. Acceptable values range from -1000
1778(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1779polarize the preference for oom killing either by always preferring a certain
1780task or completely disabling it. The lowest possible value, -1000, is
1781equivalent to disabling oom killing entirely for that task since it will always
1782report a badness score of 0.
1783
1784Consequently, it is very simple for userspace to define the amount of memory to
1785consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1786example, is roughly equivalent to allowing the remainder of tasks sharing the
1787same system, cpuset, mempolicy, or memory controller resources to use at least
178850% more memory. A value of -500, on the other hand, would be roughly
1789equivalent to discounting 50% of the task's allowed memory from being considered
1790as scoring against the task.
1791
1792For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1793be used to tune the badness score. Its acceptable values range from -16
1794(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1795(OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1796scaled linearly with /proc/<pid>/oom_score_adj.
1797
1798The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1799value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1800requires CAP_SYS_RESOURCE.
1801
1802
18033.2 /proc/<pid>/oom_score - Display current oom-killer score
1804-------------------------------------------------------------
1805
1806This file can be used to check the current score used by the oom-killer for
1807any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1808process should be killed in an out-of-memory situation.
1809
1810Please note that the exported value includes oom_score_adj so it is
1811effectively in range [0,2000].
1812
1813
18143.3 /proc/<pid>/io - Display the IO accounting fields
1815-------------------------------------------------------
1816
1817This file contains IO statistics for each running process.
1818
1819Example
1820~~~~~~~
1821
1822::
1823
1824 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1825 [1] 3828
1826
1827 test:/tmp # cat /proc/3828/io
1828 rchar: 323934931
1829 wchar: 323929600
1830 syscr: 632687
1831 syscw: 632675
1832 read_bytes: 0
1833 write_bytes: 323932160
1834 cancelled_write_bytes: 0
1835
1836
1837Description
1838~~~~~~~~~~~
1839
1840rchar
1841^^^^^
1842
1843I/O counter: chars read
1844The number of bytes which this task has caused to be read from storage. This
1845is simply the sum of bytes which this process passed to read() and pread().
1846It includes things like tty IO and it is unaffected by whether or not actual
1847physical disk IO was required (the read might have been satisfied from
1848pagecache).
1849
1850
1851wchar
1852^^^^^
1853
1854I/O counter: chars written
1855The number of bytes which this task has caused, or shall cause to be written
1856to disk. Similar caveats apply here as with rchar.
1857
1858
1859syscr
1860^^^^^
1861
1862I/O counter: read syscalls
1863Attempt to count the number of read I/O operations, i.e. syscalls like read()
1864and pread().
1865
1866
1867syscw
1868^^^^^
1869
1870I/O counter: write syscalls
1871Attempt to count the number of write I/O operations, i.e. syscalls like
1872write() and pwrite().
1873
1874
1875read_bytes
1876^^^^^^^^^^
1877
1878I/O counter: bytes read
1879Attempt to count the number of bytes which this process really did cause to
1880be fetched from the storage layer. Done at the submit_bio() level, so it is
1881accurate for block-backed filesystems. <please add status regarding NFS and
1882CIFS at a later time>
1883
1884
1885write_bytes
1886^^^^^^^^^^^
1887
1888I/O counter: bytes written
1889Attempt to count the number of bytes which this process caused to be sent to
1890the storage layer. This is done at page-dirtying time.
1891
1892
1893cancelled_write_bytes
1894^^^^^^^^^^^^^^^^^^^^^
1895
1896The big inaccuracy here is truncate. If a process writes 1MB to a file and
1897then deletes the file, it will in fact perform no writeout. But it will have
1898been accounted as having caused 1MB of write.
1899In other words: The number of bytes which this process caused to not happen,
1900by truncating pagecache. A task can cause "negative" IO too. If this task
1901truncates some dirty pagecache, some IO which another task has been accounted
1902for (in its write_bytes) will not be happening. We _could_ just subtract that
1903from the truncating task's write_bytes, but there is information loss in doing
1904that.
1905
1906
1907.. Note::
1908
1909 At its current implementation state, this is a bit racy on 32-bit machines:
1910 if process A reads process B's /proc/pid/io while process B is updating one
1911 of those 64-bit counters, process A could see an intermediate result.
1912
1913
1914More information about this can be found within the taskstats documentation in
1915Documentation/accounting.
1916
19173.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1918---------------------------------------------------------------
1919When a process is dumped, all anonymous memory is written to a core file as
1920long as the size of the core file isn't limited. But sometimes we don't want
1921to dump some memory segments, for example, huge shared memory or DAX.
1922Conversely, sometimes we want to save file-backed memory segments into a core
1923file, not only the individual files.
1924
1925/proc/<pid>/coredump_filter allows you to customize which memory segments
1926will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1927of memory types. If a bit of the bitmask is set, memory segments of the
1928corresponding memory type are dumped, otherwise they are not dumped.
1929
1930The following 9 memory types are supported:
1931
1932 - (bit 0) anonymous private memory
1933 - (bit 1) anonymous shared memory
1934 - (bit 2) file-backed private memory
1935 - (bit 3) file-backed shared memory
1936 - (bit 4) ELF header pages in file-backed private memory areas (it is
1937 effective only if the bit 2 is cleared)
1938 - (bit 5) hugetlb private memory
1939 - (bit 6) hugetlb shared memory
1940 - (bit 7) DAX private memory
1941 - (bit 8) DAX shared memory
1942
1943 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1944 are always dumped regardless of the bitmask status.
1945
1946 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is
1947 only affected by bit 5-6, and DAX is only affected by bits 7-8.
1948
1949The default value of coredump_filter is 0x33; this means all anonymous memory
1950segments, ELF header pages and hugetlb private memory are dumped.
1951
1952If you don't want to dump all shared memory segments attached to pid 1234,
1953write 0x31 to the process's proc file::
1954
1955 $ echo 0x31 > /proc/1234/coredump_filter
1956
1957When a new process is created, the process inherits the bitmask status from its
1958parent. It is useful to set up coredump_filter before the program runs.
1959For example::
1960
1961 $ echo 0x7 > /proc/self/coredump_filter
1962 $ ./some_program
1963
19643.5 /proc/<pid>/mountinfo - Information about mounts
1965--------------------------------------------------------
1966
1967This file contains lines of the form::
1968
1969 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1970 (1)(2)(3) (4) (5) (6) (n…m) (m+1)(m+2) (m+3) (m+4)
1971
1972 (1) mount ID: unique identifier of the mount (may be reused after umount)
1973 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1974 (3) major:minor: value of st_dev for files on filesystem
1975 (4) root: root of the mount within the filesystem
1976 (5) mount point: mount point relative to the process's root
1977 (6) mount options: per mount options
1978 (n…m) optional fields: zero or more fields of the form "tag[:value]"
1979 (m+1) separator: marks the end of the optional fields
1980 (m+2) filesystem type: name of filesystem of the form "type[.subtype]"
1981 (m+3) mount source: filesystem specific information or "none"
1982 (m+4) super options: per super block options
1983
1984Parsers should ignore all unrecognised optional fields. Currently the
1985possible optional fields are:
1986
1987================ ==============================================================
1988shared:X mount is shared in peer group X
1989master:X mount is slave to peer group X
1990propagate_from:X mount is slave and receives propagation from peer group X [#]_
1991unbindable mount is unbindable
1992================ ==============================================================
1993
1994.. [#] X is the closest dominant peer group under the process's root. If
1995 X is the immediate master of the mount, or if there's no dominant peer
1996 group under the same root, then only the "master:X" field is present
1997 and not the "propagate_from:X" field.
1998
1999For more information on mount propagation see:
2000
2001 Documentation/filesystems/sharedsubtree.rst
2002
2003
20043.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
2005--------------------------------------------------------
2006These files provide a method to access a task's comm value. It also allows for
2007a task to set its own or one of its thread siblings comm value. The comm value
2008is limited in size compared to the cmdline value, so writing anything longer
2009then the kernel's TASK_COMM_LEN (currently 16 chars, including the NUL
2010terminator) will result in a truncated comm value.
2011
2012
20133.7 /proc/<pid>/task/<tid>/children - Information about task children
2014-------------------------------------------------------------------------
2015This file provides a fast way to retrieve first level children pids
2016of a task pointed by <pid>/<tid> pair. The format is a space separated
2017stream of pids.
2018
2019Note the "first level" here -- if a child has its own children they will
2020not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children
2021to obtain the descendants.
2022
2023Since this interface is intended to be fast and cheap it doesn't
2024guarantee to provide precise results and some children might be
2025skipped, especially if they've exited right after we printed their
2026pids, so one needs to either stop or freeze processes being inspected
2027if precise results are needed.
2028
2029
20303.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
2031---------------------------------------------------------------
2032This file provides information associated with an opened file. The regular
2033files have at least four fields -- 'pos', 'flags', 'mnt_id' and 'ino'.
2034The 'pos' represents the current offset of the opened file in decimal
2035form [see lseek(2) for details], 'flags' denotes the octal O_xxx mask the
2036file has been created with [see open(2) for details] and 'mnt_id' represents
2037mount ID of the file system containing the opened file [see 3.5
2038/proc/<pid>/mountinfo for details]. 'ino' represents the inode number of
2039the file.
2040
2041A typical output is::
2042
2043 pos: 0
2044 flags: 0100002
2045 mnt_id: 19
2046 ino: 63107
2047
2048All locks associated with a file descriptor are shown in its fdinfo too::
2049
2050 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
2051
2052The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
2053pair provide additional information particular to the objects they represent.
2054
2055Eventfd files
2056~~~~~~~~~~~~~
2057
2058::
2059
2060 pos: 0
2061 flags: 04002
2062 mnt_id: 9
2063 ino: 63107
2064 eventfd-count: 5a
2065
2066where 'eventfd-count' is hex value of a counter.
2067
2068Signalfd files
2069~~~~~~~~~~~~~~
2070
2071::
2072
2073 pos: 0
2074 flags: 04002
2075 mnt_id: 9
2076 ino: 63107
2077 sigmask: 0000000000000200
2078
2079where 'sigmask' is hex value of the signal mask associated
2080with a file.
2081
2082Epoll files
2083~~~~~~~~~~~
2084
2085::
2086
2087 pos: 0
2088 flags: 02
2089 mnt_id: 9
2090 ino: 63107
2091 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7
2092
2093where 'tfd' is a target file descriptor number in decimal form,
2094'events' is events mask being watched and the 'data' is data
2095associated with a target [see epoll(7) for more details].
2096
2097The 'pos' is current offset of the target file in decimal form
2098[see lseek(2)], 'ino' and 'sdev' are inode and device numbers
2099where target file resides, all in hex format.
2100
2101Fsnotify files
2102~~~~~~~~~~~~~~
2103For inotify files the format is the following::
2104
2105 pos: 0
2106 flags: 02000000
2107 mnt_id: 9
2108 ino: 63107
2109 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
2110
2111where 'wd' is a watch descriptor in decimal form, i.e. a target file
2112descriptor number, 'ino' and 'sdev' are inode and device where the
2113target file resides and the 'mask' is the mask of events, all in hex
2114form [see inotify(7) for more details].
2115
2116If the kernel was built with exportfs support, the path to the target
2117file is encoded as a file handle. The file handle is provided by three
2118fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
2119format.
2120
2121If the kernel is built without exportfs support the file handle won't be
2122printed out.
2123
2124If there is no inotify mark attached yet the 'inotify' line will be omitted.
2125
2126For fanotify files the format is::
2127
2128 pos: 0
2129 flags: 02
2130 mnt_id: 9
2131 ino: 63107
2132 fanotify flags:10 event-flags:0
2133 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
2134 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
2135
2136where fanotify 'flags' and 'event-flags' are values used in fanotify_init
2137call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
2138flags associated with mark which are tracked separately from events
2139mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events
2140mask and 'ignored_mask' is the mask of events which are to be ignored.
2141All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
2142provide information about flags and mask used in fanotify_mark
2143call [see fsnotify manpage for details].
2144
2145While the first three lines are mandatory and always printed, the rest is
2146optional and may be omitted if no marks created yet.
2147
2148Timerfd files
2149~~~~~~~~~~~~~
2150
2151::
2152
2153 pos: 0
2154 flags: 02
2155 mnt_id: 9
2156 ino: 63107
2157 clockid: 0
2158 ticks: 0
2159 settime flags: 01
2160 it_value: (0, 49406829)
2161 it_interval: (1, 0)
2162
2163where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
2164that have occurred [see timerfd_create(2) for details]. 'settime flags' are
2165flags in octal form been used to setup the timer [see timerfd_settime(2) for
2166details]. 'it_value' is remaining time until the timer expiration.
2167'it_interval' is the interval for the timer. Note the timer might be set up
2168with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
2169still exhibits timer's remaining time.
2170
2171DMA Buffer files
2172~~~~~~~~~~~~~~~~
2173
2174::
2175
2176 pos: 0
2177 flags: 04002
2178 mnt_id: 9
2179 ino: 63107
2180 size: 32768
2181 count: 2
2182 exp_name: system-heap
2183
2184where 'size' is the size of the DMA buffer in bytes. 'count' is the file count of
2185the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter.
2186
2187VFIO Device files
2188~~~~~~~~~~~~~~~~~
2189
2190::
2191
2192 pos: 0
2193 flags: 02000002
2194 mnt_id: 17
2195 ino: 5122
2196 vfio-device-syspath: /sys/devices/pci0000:e0/0000:e0:01.1/0000:e1:00.0/0000:e2:05.0/0000:e8:00.0
2197
2198where 'vfio-device-syspath' is the sysfs path corresponding to the VFIO device
2199file.
2200
22013.9 /proc/<pid>/map_files - Information about memory mapped files
2202---------------------------------------------------------------------
2203This directory contains symbolic links which represent memory mapped files
2204the process is maintaining. Example output::
2205
2206 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
2207 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
2208 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
2209 | ...
2210 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
2211 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
2212
2213The name of a link represents the virtual memory bounds of a mapping, i.e.
2214vm_area_struct::vm_start-vm_area_struct::vm_end.
2215
2216The main purpose of the map_files is to retrieve a set of memory mapped
2217files in a fast way instead of parsing /proc/<pid>/maps or
2218/proc/<pid>/smaps, both of which contain many more records. At the same
2219time one can open(2) mappings from the listings of two processes and
2220comparing their inode numbers to figure out which anonymous memory areas
2221are actually shared.
2222
22233.10 /proc/<pid>/timerslack_ns - Task timerslack value
2224---------------------------------------------------------
2225This file provides the value of the task's timerslack value in nanoseconds.
2226This value specifies an amount of time that normal timers may be deferred
2227in order to coalesce timers and avoid unnecessary wakeups.
2228
2229This allows a task's interactivity vs power consumption tradeoff to be
2230adjusted.
2231
2232Writing 0 to the file will set the task's timerslack to the default value.
2233
2234Valid values are from 0 - ULLONG_MAX
2235
2236An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level
2237permissions on the task specified to change its timerslack_ns value.
2238
22393.11 /proc/<pid>/patch_state - Livepatch patch operation state
2240-----------------------------------------------------------------
2241When CONFIG_LIVEPATCH is enabled, this file displays the value of the
2242patch state for the task.
2243
2244A value of '-1' indicates that no patch is in transition.
2245
2246A value of '0' indicates that a patch is in transition and the task is
2247unpatched. If the patch is being enabled, then the task hasn't been
2248patched yet. If the patch is being disabled, then the task has already
2249been unpatched.
2250
2251A value of '1' indicates that a patch is in transition and the task is
2252patched. If the patch is being enabled, then the task has already been
2253patched. If the patch is being disabled, then the task hasn't been
2254unpatched yet.
2255
22563.12 /proc/<pid>/arch_status - task architecture specific status
2257-------------------------------------------------------------------
2258When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the
2259architecture specific status of the task.
2260
2261Example
2262~~~~~~~
2263
2264::
2265
2266 $ cat /proc/6753/arch_status
2267 AVX512_elapsed_ms: 8
2268
2269Description
2270~~~~~~~~~~~
2271
2272x86 specific entries
2273~~~~~~~~~~~~~~~~~~~~~
2274
2275AVX512_elapsed_ms
2276^^^^^^^^^^^^^^^^^^
2277
2278 If AVX512 is supported on the machine, this entry shows the milliseconds
2279 elapsed since the last time AVX512 usage was recorded. The recording
2280 happens on a best effort basis when a task is scheduled out. This means
2281 that the value depends on two factors:
2282
2283 1) The time which the task spent on the CPU without being scheduled
2284 out. With CPU isolation and a single runnable task this can take
2285 several seconds.
2286
2287 2) The time since the task was scheduled out last. Depending on the
2288 reason for being scheduled out (time slice exhausted, syscall ...)
2289 this can be arbitrary long time.
2290
2291 As a consequence the value cannot be considered precise and authoritative
2292 information. The application which uses this information has to be aware
2293 of the overall scenario on the system in order to determine whether a
2294 task is a real AVX512 user or not. Precise information can be obtained
2295 with performance counters.
2296
2297 A special value of '-1' indicates that no AVX512 usage was recorded, thus
2298 the task is unlikely an AVX512 user, but depends on the workload and the
2299 scheduling scenario, it also could be a false negative mentioned above.
2300
23013.13 /proc/<pid>/fd - List of symlinks to open files
2302-------------------------------------------------------
2303This directory contains symbolic links which represent open files
2304the process is maintaining. Example output::
2305
2306 lr-x------ 1 root root 64 Sep 20 17:53 0 -> /dev/null
2307 l-wx------ 1 root root 64 Sep 20 17:53 1 -> /dev/null
2308 lrwx------ 1 root root 64 Sep 20 17:53 10 -> 'socket:[12539]'
2309 lrwx------ 1 root root 64 Sep 20 17:53 11 -> 'socket:[12540]'
2310 lrwx------ 1 root root 64 Sep 20 17:53 12 -> 'socket:[12542]'
2311
2312The number of open files for the process is stored in 'size' member
2313of stat() output for /proc/<pid>/fd for fast access.
2314-------------------------------------------------------
2315
23163.14 /proc/<pid>/ksm_stat - Information about the process's ksm status
2317----------------------------------------------------------------------
2318When CONFIG_KSM is enabled, each process has this file which displays
2319the information of ksm merging status.
2320
2321Example
2322~~~~~~~
2323
2324::
2325
2326 / # cat /proc/self/ksm_stat
2327 ksm_rmap_items 0
2328 ksm_zero_pages 0
2329 ksm_merging_pages 0
2330 ksm_process_profit 0
2331 ksm_merge_any: no
2332 ksm_mergeable: no
2333
2334Description
2335~~~~~~~~~~~
2336
2337ksm_rmap_items
2338^^^^^^^^^^^^^^
2339
2340The number of ksm_rmap_item structures in use. The structure
2341ksm_rmap_item stores the reverse mapping information for virtual
2342addresses. KSM will generate a ksm_rmap_item for each ksm-scanned page of
2343the process.
2344
2345ksm_zero_pages
2346^^^^^^^^^^^^^^
2347
2348When /sys/kernel/mm/ksm/use_zero_pages is enabled, it represent how many
2349empty pages are merged with kernel zero pages by KSM.
2350
2351ksm_merging_pages
2352^^^^^^^^^^^^^^^^^
2353
2354It represents how many pages of this process are involved in KSM merging
2355(not including ksm_zero_pages). It is the same with what
2356/proc/<pid>/ksm_merging_pages shows.
2357
2358ksm_process_profit
2359^^^^^^^^^^^^^^^^^^
2360
2361The profit that KSM brings (Saved bytes). KSM can save memory by merging
2362identical pages, but also can consume additional memory, because it needs
2363to generate a number of rmap_items to save each scanned page's brief rmap
2364information. Some of these pages may be merged, but some may not be abled
2365to be merged after being checked several times, which are unprofitable
2366memory consumed.
2367
2368ksm_merge_any
2369^^^^^^^^^^^^^
2370
2371It specifies whether the process's 'mm is added by prctl() into the
2372candidate list of KSM or not, and if KSM scanning is fully enabled at
2373process level.
2374
2375ksm_mergeable
2376^^^^^^^^^^^^^
2377
2378It specifies whether any VMAs of the process''s mms are currently
2379applicable to KSM.
2380
2381More information about KSM can be found in
2382Documentation/admin-guide/mm/ksm.rst.
2383
2384
2385Chapter 4: Configuring procfs
2386=============================
2387
23884.1 Mount options
2389---------------------
2390
2391The following mount options are supported:
2392
2393 ========= ========================================================
2394 hidepid= Set /proc/<pid>/ access mode.
2395 gid= Set the group authorized to learn processes information.
2396 subset= Show only the specified subset of procfs.
2397 pidns= Specify a the namespace used by this procfs.
2398 ========= ========================================================
2399
2400hidepid=off or hidepid=0 means classic mode - everybody may access all
2401/proc/<pid>/ directories (default).
2402
2403hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/
2404directories but their own. Sensitive files like cmdline, sched*, status are now
2405protected against other users. This makes it impossible to learn whether any
2406user runs specific program (given the program doesn't reveal itself by its
2407behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for
2408other users, poorly written programs passing sensitive information via program
2409arguments are now protected against local eavesdroppers.
2410
2411hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be
2412fully invisible to other users. It doesn't mean that it hides a fact whether a
2413process with a specific pid value exists (it can be learned by other means, e.g.
2414by "kill -0 $PID"), but it hides process's uid and gid, which may be learned by
2415stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of
2416gathering information about running processes, whether some daemon runs with
2417elevated privileges, whether other user runs some sensitive program, whether
2418other users run any program at all, etc.
2419
2420hidepid=ptraceable or hidepid=4 means that procfs should only contain
2421/proc/<pid>/ directories that the caller can ptrace.
2422
2423gid= defines a group authorized to learn processes information otherwise
2424prohibited by hidepid=. If you use some daemon like identd which needs to learn
2425information about processes information, just add identd to this group.
2426
2427subset=pid hides all top level files and directories in the procfs that
2428are not related to tasks.
2429
2430pidns= specifies a pid namespace (either as a string path to something like
2431`/proc/$pid/ns/pid`, or a file descriptor when using `FSCONFIG_SET_FD`) that
2432will be used by the procfs instance when translating pids. By default, procfs
2433will use the calling process's active pid namespace. Note that the pid
2434namespace of an existing procfs instance cannot be modified (attempting to do
2435so will give an `-EBUSY` error).
2436
2437Chapter 5: Filesystem behavior
2438==============================
2439
2440Originally, before the advent of pid namespace, procfs was a global file
2441system. It means that there was only one procfs instance in the system.
2442
2443When pid namespace was added, a separate procfs instance was mounted in
2444each pid namespace. So, procfs mount options are global among all
2445mountpoints within the same namespace::
2446
2447 # grep ^proc /proc/mounts
2448 proc /proc proc rw,relatime,hidepid=2 0 0
2449
2450 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc
2451 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 0
2452 +++ exited with 0 +++
2453
2454 # grep ^proc /proc/mounts
2455 proc /proc proc rw,relatime,hidepid=2 0 0
2456 proc /tmp/proc proc rw,relatime,hidepid=2 0 0
2457
2458and only after remounting procfs mount options will change at all
2459mountpoints::
2460
2461 # mount -o remount,hidepid=1 -t proc proc /tmp/proc
2462
2463 # grep ^proc /proc/mounts
2464 proc /proc proc rw,relatime,hidepid=1 0 0
2465 proc /tmp/proc proc rw,relatime,hidepid=1 0 0
2466
2467This behavior is different from the behavior of other filesystems.
2468
2469The new procfs behavior is more like other filesystems. Each procfs mount
2470creates a new procfs instance. Mount options affect own procfs instance.
2471It means that it became possible to have several procfs instances
2472displaying tasks with different filtering options in one pid namespace::
2473
2474 # mount -o hidepid=invisible -t proc proc /proc
2475 # mount -o hidepid=noaccess -t proc proc /tmp/proc
2476 # grep ^proc /proc/mounts
2477 proc /proc proc rw,relatime,hidepid=invisible 0 0
2478 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 0