| blob | 99ca040e3f900e399b59e2b5233fa9169ac61ccb |
1 ------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3 ------------------------------------------------------------------------------
4 /proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
7 2.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8 move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
9 ------------------------------------------------------------------------------
10 Version 1.3 Kernel version 2.2.12
11 Kernel version 2.4.0-test11-pre4
12 ------------------------------------------------------------------------------
13 fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
15 Table of Contents
16 -----------------
18 0 Preface
19 0.1 Introduction/Credits
20 0.2 Legal Stuff
22 1 Collecting System Information
23 1.1 Process-Specific Subdirectories
24 1.2 Kernel data
25 1.3 IDE devices in /proc/ide
26 1.4 Networking info in /proc/net
27 1.5 SCSI info
28 1.6 Parallel port info in /proc/parport
29 1.7 TTY info in /proc/tty
30 1.8 Miscellaneous kernel statistics in /proc/stat
31 1.9 Ext4 file system parameters
33 2 Modifying System Parameters
35 3 Per-Process Parameters
36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
37 score
38 3.2 /proc/<pid>/oom_score - Display current oom-killer score
39 3.3 /proc/<pid>/io - Display the IO accounting fields
40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
41 3.5 /proc/<pid>/mountinfo - Information about mounts
42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
43 3.7 /proc/<pid>/task/<tid>/children - Information about task children
44 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
45 3.9 /proc/<pid>/map_files - Information about memory mapped files
46 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
47 3.11 /proc/<pid>/patch_state - Livepatch patch operation state
48 3.12 /proc/<pid>/arch_status - Task architecture specific information
50 4 Configuring procfs
51 4.1 Mount options
53 ------------------------------------------------------------------------------
54 Preface
55 ------------------------------------------------------------------------------
57 0.1 Introduction/Credits
58 ------------------------
60 This documentation is part of a soon (or so we hope) to be released book on
61 the SuSE Linux distribution. As there is no complete documentation for the
62 /proc file system and we've used many freely available sources to write these
63 chapters, it seems only fair to give the work back to the Linux community.
64 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
65 afraid it's still far from complete, but we hope it will be useful. As far as
66 we know, it is the first 'all-in-one' document about the /proc file system. It
67 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
68 SPARC, AXP, etc., features, you probably won't find what you are looking for.
69 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
70 additions and patches are welcome and will be added to this document if you
71 mail them to Bodo.
73 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
74 other people for help compiling this documentation. We'd also like to extend a
75 special thank you to Andi Kleen for documentation, which we relied on heavily
76 to create this document, as well as the additional information he provided.
77 Thanks to everybody else who contributed source or docs to the Linux kernel
78 and helped create a great piece of software... :)
80 If you have any comments, corrections or additions, please don't hesitate to
81 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
82 document.
84 The latest version of this document is available online at
85 http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
87 If the above direction does not works for you, you could try the kernel
88 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
89 comandante@zaralinux.com.
91 0.2 Legal Stuff
92 ---------------
94 We don't guarantee the correctness of this document, and if you come to us
95 complaining about how you screwed up your system because of incorrect
96 documentation, we won't feel responsible...
98 ------------------------------------------------------------------------------
99 CHAPTER 1: COLLECTING SYSTEM INFORMATION
100 ------------------------------------------------------------------------------
102 ------------------------------------------------------------------------------
103 In This Chapter
104 ------------------------------------------------------------------------------
105 * Investigating the properties of the pseudo file system /proc and its
106 ability to provide information on the running Linux system
107 * Examining /proc's structure
108 * Uncovering various information about the kernel and the processes running
109 on the system
110 ------------------------------------------------------------------------------
113 The proc file system acts as an interface to internal data structures in the
114 kernel. It can be used to obtain information about the system and to change
115 certain kernel parameters at runtime (sysctl).
117 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
118 show you how you can use /proc/sys to change settings.
120 1.1 Process-Specific Subdirectories
121 -----------------------------------
123 The directory /proc contains (among other things) one subdirectory for each
124 process running on the system, which is named after the process ID (PID).
126 The link self points to the process reading the file system. Each process
127 subdirectory has the entries listed in Table 1-1.
129 Note that an open a file descriptor to /proc/<pid> or to any of its
130 contained files or subdirectories does not prevent <pid> being reused
131 for some other process in the event that <pid> exits. Operations on
132 open /proc/<pid> file descriptors corresponding to dead processes
133 never act on any new process that the kernel may, through chance, have
134 also assigned the process ID <pid>. Instead, operations on these FDs
135 usually fail with ESRCH.
137 Table 1-1: Process specific entries in /proc
138 ..............................................................................
139 File Content
140 clear_refs Clears page referenced bits shown in smaps output
141 cmdline Command line arguments
142 cpu Current and last cpu in which it was executed (2.4)(smp)
143 cwd Link to the current working directory
144 environ Values of environment variables
145 exe Link to the executable of this process
146 fd Directory, which contains all file descriptors
147 maps Memory maps to executables and library files (2.4)
148 mem Memory held by this process
149 root Link to the root directory of this process
150 stat Process status
151 statm Process memory status information
152 status Process status in human readable form
153 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function
154 symbol the task is blocked in - or "0" if not blocked.
155 pagemap Page table
156 stack Report full stack trace, enable via CONFIG_STACKTRACE
157 smaps An extension based on maps, showing the memory consumption of
158 each mapping and flags associated with it
159 smaps_rollup Accumulated smaps stats for all mappings of the process. This
160 can be derived from smaps, but is faster and more convenient
161 numa_maps An extension based on maps, showing the memory locality and
162 binding policy as well as mem usage (in pages) of each mapping.
163 ..............................................................................
165 For example, to get the status information of a process, all you have to do is
166 read the file /proc/PID/status:
168 >cat /proc/self/status
169 Name: cat
170 State: R (running)
171 Tgid: 5452
172 Pid: 5452
173 PPid: 743
174 TracerPid: 0 (2.4)
175 Uid: 501 501 501 501
176 Gid: 100 100 100 100
177 FDSize: 256
178 Groups: 100 14 16
179 VmPeak: 5004 kB
180 VmSize: 5004 kB
181 VmLck: 0 kB
182 VmHWM: 476 kB
183 VmRSS: 476 kB
184 RssAnon: 352 kB
185 RssFile: 120 kB
186 RssShmem: 4 kB
187 VmData: 156 kB
188 VmStk: 88 kB
189 VmExe: 68 kB
190 VmLib: 1412 kB
191 VmPTE: 20 kb
192 VmSwap: 0 kB
193 HugetlbPages: 0 kB
194 CoreDumping: 0
195 THP_enabled: 1
196 Threads: 1
197 SigQ: 0/28578
198 SigPnd: 0000000000000000
199 ShdPnd: 0000000000000000
200 SigBlk: 0000000000000000
201 SigIgn: 0000000000000000
202 SigCgt: 0000000000000000
203 CapInh: 00000000fffffeff
204 CapPrm: 0000000000000000
205 CapEff: 0000000000000000
206 CapBnd: ffffffffffffffff
207 CapAmb: 0000000000000000
208 NoNewPrivs: 0
209 Seccomp: 0
210 Speculation_Store_Bypass: thread vulnerable
211 voluntary_ctxt_switches: 0
212 nonvoluntary_ctxt_switches: 1
214 This shows you nearly the same information you would get if you viewed it with
215 the ps command. In fact, ps uses the proc file system to obtain its
216 information. But you get a more detailed view of the process by reading the
217 file /proc/PID/status. It fields are described in table 1-2.
219 The statm file contains more detailed information about the process
220 memory usage. Its seven fields are explained in Table 1-3. The stat file
221 contains details information about the process itself. Its fields are
222 explained in Table 1-4.
224 (for SMP CONFIG users)
225 For making accounting scalable, RSS related information are handled in an
226 asynchronous manner and the value may not be very precise. To see a precise
227 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
228 It's slow but very precise.
230 Table 1-2: Contents of the status files (as of 4.19)
231 ..............................................................................
232 Field Content
233 Name filename of the executable
234 Umask file mode creation mask
235 State state (R is running, S is sleeping, D is sleeping
236 in an uninterruptible wait, Z is zombie,
237 T is traced or stopped)
238 Tgid thread group ID
239 Ngid NUMA group ID (0 if none)
240 Pid process id
241 PPid process id of the parent process
242 TracerPid PID of process tracing this process (0 if not)
243 Uid Real, effective, saved set, and file system UIDs
244 Gid Real, effective, saved set, and file system GIDs
245 FDSize number of file descriptor slots currently allocated
246 Groups supplementary group list
247 NStgid descendant namespace thread group ID hierarchy
248 NSpid descendant namespace process ID hierarchy
249 NSpgid descendant namespace process group ID hierarchy
250 NSsid descendant namespace session ID hierarchy
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 (VmRSS = RssAnon + RssFile + RssShmem)
258 RssAnon size of resident anonymous memory
259 RssFile size of resident file mappings
260 RssShmem size of resident shmem memory (includes SysV shm,
261 mapping of tmpfs and shared anonymous mappings)
262 VmData size of private data segments
263 VmStk size of stack segments
264 VmExe size of text segment
265 VmLib size of shared library code
266 VmPTE size of page table entries
267 VmSwap amount of swap used by anonymous private data
268 (shmem swap usage is not included)
269 HugetlbPages size of hugetlb memory portions
270 CoreDumping process's memory is currently being dumped
271 (killing the process may lead to a corrupted core)
272 THP_enabled process is allowed to use THP (returns 0 when
273 PR_SET_THP_DISABLE is set on the process
274 Threads number of threads
275 SigQ number of signals queued/max. number for queue
276 SigPnd bitmap of pending signals for the thread
277 ShdPnd bitmap of shared pending signals for the process
278 SigBlk bitmap of blocked signals
279 SigIgn bitmap of ignored signals
280 SigCgt bitmap of caught signals
281 CapInh bitmap of inheritable capabilities
282 CapPrm bitmap of permitted capabilities
283 CapEff bitmap of effective capabilities
284 CapBnd bitmap of capabilities bounding set
285 CapAmb bitmap of ambient capabilities
286 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...)
287 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
288 Speculation_Store_Bypass speculative store bypass mitigation status
289 Cpus_allowed mask of CPUs on which this process may run
290 Cpus_allowed_list Same as previous, but in "list format"
291 Mems_allowed mask of memory nodes allowed to this process
292 Mems_allowed_list Same as previous, but in "list format"
293 voluntary_ctxt_switches number of voluntary context switches
294 nonvoluntary_ctxt_switches number of non voluntary context switches
295 ..............................................................................
297 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
298 ..............................................................................
299 Field Content
300 size total program size (pages) (same as VmSize in status)
301 resident size of memory portions (pages) (same as VmRSS in status)
302 shared number of pages that are shared (i.e. backed by a file, same
303 as RssFile+RssShmem in status)
304 trs number of pages that are 'code' (not including libs; broken,
305 includes data segment)
306 lrs number of pages of library (always 0 on 2.6)
307 drs number of pages of data/stack (including libs; broken,
308 includes library text)
309 dt number of dirty pages (always 0 on 2.6)
310 ..............................................................................
313 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
314 ..............................................................................
315 Field Content
316 pid process id
317 tcomm filename of the executable
318 state state (R is running, S is sleeping, D is sleeping in an
319 uninterruptible wait, Z is zombie, T is traced or stopped)
320 ppid process id of the parent process
321 pgrp pgrp of the process
322 sid session id
323 tty_nr tty the process uses
324 tty_pgrp pgrp of the tty
325 flags task flags
326 min_flt number of minor faults
327 cmin_flt number of minor faults with child's
328 maj_flt number of major faults
329 cmaj_flt number of major faults with child's
330 utime user mode jiffies
331 stime kernel mode jiffies
332 cutime user mode jiffies with child's
333 cstime kernel mode jiffies with child's
334 priority priority level
335 nice nice level
336 num_threads number of threads
337 it_real_value (obsolete, always 0)
338 start_time time the process started after system boot
339 vsize virtual memory size
340 rss resident set memory size
341 rsslim current limit in bytes on the rss
342 start_code address above which program text can run
343 end_code address below which program text can run
344 start_stack address of the start of the main process stack
345 esp current value of ESP
346 eip current value of EIP
347 pending bitmap of pending signals
348 blocked bitmap of blocked signals
349 sigign bitmap of ignored signals
350 sigcatch bitmap of caught signals
351 0 (place holder, used to be the wchan address, use /proc/PID/wchan instead)
352 0 (place holder)
353 0 (place holder)
354 exit_signal signal to send to parent thread on exit
355 task_cpu which CPU the task is scheduled on
356 rt_priority realtime priority
357 policy scheduling policy (man sched_setscheduler)
358 blkio_ticks time spent waiting for block IO
359 gtime guest time of the task in jiffies
360 cgtime guest time of the task children in jiffies
361 start_data address above which program data+bss is placed
362 end_data address below which program data+bss is placed
363 start_brk address above which program heap can be expanded with brk()
364 arg_start address above which program command line is placed
365 arg_end address below which program command line is placed
366 env_start address above which program environment is placed
367 env_end address below which program environment is placed
368 exit_code the thread's exit_code in the form reported by the waitpid system call
369 ..............................................................................
371 The /proc/PID/maps file contains the currently mapped memory regions and
372 their access permissions.
374 The format is:
376 address perms offset dev inode pathname
378 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
379 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
380 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
381 a7cb1000-a7cb2000 ---p 00000000 00:00 0
382 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
383 a7eb2000-a7eb3000 ---p 00000000 00:00 0
384 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
385 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
386 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
387 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
388 a800b000-a800e000 rw-p 00000000 00:00 0
389 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
390 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
391 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
392 a8024000-a8027000 rw-p 00000000 00:00 0
393 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
394 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
395 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
396 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
397 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
399 where "address" is the address space in the process that it occupies, "perms"
400 is a set of permissions:
402 r = read
403 w = write
404 x = execute
405 s = shared
406 p = private (copy on write)
408 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
409 "inode" is the inode on that device. 0 indicates that no inode is associated
410 with the memory region, as the case would be with BSS (uninitialized data).
411 The "pathname" shows the name associated file for this mapping. If the mapping
412 is not associated with a file:
414 [heap] = the heap of the program
415 [stack] = the stack of the main process
416 [vdso] = the "virtual dynamic shared object",
417 the kernel system call handler
419 or if empty, the mapping is anonymous.
421 The /proc/PID/smaps is an extension based on maps, showing the memory
422 consumption for each of the process's mappings. For each mapping (aka Virtual
423 Memory Area, or VMA) there is a series of lines such as the following:
425 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
427 Size: 1084 kB
428 KernelPageSize: 4 kB
429 MMUPageSize: 4 kB
430 Rss: 892 kB
431 Pss: 374 kB
432 Shared_Clean: 892 kB
433 Shared_Dirty: 0 kB
434 Private_Clean: 0 kB
435 Private_Dirty: 0 kB
436 Referenced: 892 kB
437 Anonymous: 0 kB
438 LazyFree: 0 kB
439 AnonHugePages: 0 kB
440 ShmemPmdMapped: 0 kB
441 Shared_Hugetlb: 0 kB
442 Private_Hugetlb: 0 kB
443 Swap: 0 kB
444 SwapPss: 0 kB
445 KernelPageSize: 4 kB
446 MMUPageSize: 4 kB
447 Locked: 0 kB
448 THPeligible: 0
449 VmFlags: rd ex mr mw me dw
451 The first of these lines shows the same information as is displayed for the
452 mapping in /proc/PID/maps. Following lines show the size of the mapping
453 (size); the size of each page allocated when backing a VMA (KernelPageSize),
454 which is usually the same as the size in the page table entries; the page size
455 used by the MMU when backing a VMA (in most cases, the same as KernelPageSize);
456 the amount of the mapping that is currently resident in RAM (RSS); the
457 process' proportional share of this mapping (PSS); and the number of clean and
458 dirty shared and private pages in the mapping.
460 The "proportional set size" (PSS) of a process is the count of pages it has
461 in memory, where each page is divided by the number of processes sharing it.
462 So if a process has 1000 pages all to itself, and 1000 shared with one other
463 process, its PSS will be 1500.
464 Note that even a page which is part of a MAP_SHARED mapping, but has only
465 a single pte mapped, i.e. is currently used by only one process, is accounted
466 as private and not as shared.
467 "Referenced" indicates the amount of memory currently marked as referenced or
468 accessed.
469 "Anonymous" shows the amount of memory that does not belong to any file. Even
470 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
471 and a page is modified, the file page is replaced by a private anonymous copy.
472 "LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE).
473 The memory isn't freed immediately with madvise(). It's freed in memory
474 pressure if the memory is clean. Please note that the printed value might
475 be lower than the real value due to optimizations used in the current
476 implementation. If this is not desirable please file a bug report.
477 "AnonHugePages" shows the ammount of memory backed by transparent hugepage.
478 "ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by
479 huge pages.
480 "Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by
481 hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
482 reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
483 "Swap" shows how much would-be-anonymous memory is also used, but out on swap.
484 For shmem mappings, "Swap" includes also the size of the mapped (and not
485 replaced by copy-on-write) part of the underlying shmem object out on swap.
486 "SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this
487 does not take into account swapped out page of underlying shmem objects.
488 "Locked" indicates whether the mapping is locked in memory or not.
489 "THPeligible" indicates whether the mapping is eligible for allocating THP
490 pages - 1 if true, 0 otherwise. It just shows the current status.
492 "VmFlags" field deserves a separate description. This member represents the kernel
493 flags associated with the particular virtual memory area in two letter encoded
494 manner. The codes are the following:
495 rd - readable
496 wr - writeable
497 ex - executable
498 sh - shared
499 mr - may read
500 mw - may write
501 me - may execute
502 ms - may share
503 gd - stack segment growns down
504 pf - pure PFN range
505 dw - disabled write to the mapped file
506 lo - pages are locked in memory
507 io - memory mapped I/O area
508 sr - sequential read advise provided
509 rr - random read advise provided
510 dc - do not copy area on fork
511 de - do not expand area on remapping
512 ac - area is accountable
513 nr - swap space is not reserved for the area
514 ht - area uses huge tlb pages
515 ar - architecture specific flag
516 dd - do not include area into core dump
517 sd - soft-dirty flag
518 mm - mixed map area
519 hg - huge page advise flag
520 nh - no-huge page advise flag
521 mg - mergable advise flag
523 Note that there is no guarantee that every flag and associated mnemonic will
524 be present in all further kernel releases. Things get changed, the flags may
525 be vanished or the reverse -- new added. Interpretation of their meaning
526 might change in future as well. So each consumer of these flags has to
527 follow each specific kernel version for the exact semantic.
529 This file is only present if the CONFIG_MMU kernel configuration option is
530 enabled.
532 Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent
533 output can be achieved only in the single read call).
534 This typically manifests when doing partial reads of these files while the
535 memory map is being modified. Despite the races, we do provide the following
536 guarantees:
538 1) The mapped addresses never go backwards, which implies no two
539 regions will ever overlap.
540 2) If there is something at a given vaddr during the entirety of the
541 life of the smaps/maps walk, there will be some output for it.
543 The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps,
544 but their values are the sums of the corresponding values for all mappings of
545 the process. Additionally, it contains these fields:
547 Pss_Anon
548 Pss_File
549 Pss_Shmem
551 They represent the proportional shares of anonymous, file, and shmem pages, as
552 described for smaps above. These fields are omitted in smaps since each
553 mapping identifies the type (anon, file, or shmem) of all pages it contains.
554 Thus all information in smaps_rollup can be derived from smaps, but at a
555 significantly higher cost.
557 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
558 bits on both physical and virtual pages associated with a process, and the
559 soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst
560 for details).
561 To clear the bits for all the pages associated with the process
562 > echo 1 > /proc/PID/clear_refs
564 To clear the bits for the anonymous pages associated with the process
565 > echo 2 > /proc/PID/clear_refs
567 To clear the bits for the file mapped pages associated with the process
568 > echo 3 > /proc/PID/clear_refs
570 To clear the soft-dirty bit
571 > echo 4 > /proc/PID/clear_refs
573 To reset the peak resident set size ("high water mark") to the process's
574 current value:
575 > echo 5 > /proc/PID/clear_refs
577 Any other value written to /proc/PID/clear_refs will have no effect.
579 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
580 using /proc/kpageflags and number of times a page is mapped using
581 /proc/kpagecount. For detailed explanation, see
582 Documentation/admin-guide/mm/pagemap.rst.
584 The /proc/pid/numa_maps is an extension based on maps, showing the memory
585 locality and binding policy, as well as the memory usage (in pages) of
586 each mapping. The output follows a general format where mapping details get
587 summarized separated by blank spaces, one mapping per each file line:
589 address policy mapping details
591 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
592 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
593 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
594 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
595 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
596 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
597 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
598 320698b000 default file=/lib64/libc-2.12.so
599 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
600 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
601 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
602 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
603 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
604 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
605 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
606 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
608 Where:
609 "address" is the starting address for the mapping;
610 "policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst);
611 "mapping details" summarizes mapping data such as mapping type, page usage counters,
612 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
613 size, in KB, that is backing the mapping up.
615 1.2 Kernel data
616 ---------------
618 Similar to the process entries, the kernel data files give information about
619 the running kernel. The files used to obtain this information are contained in
620 /proc and are listed in Table 1-5. Not all of these will be present in your
621 system. It depends on the kernel configuration and the loaded modules, which
622 files are there, and which are missing.
624 Table 1-5: Kernel info in /proc
625 ..............................................................................
626 File Content
627 apm Advanced power management info
628 buddyinfo Kernel memory allocator information (see text) (2.5)
629 bus Directory containing bus specific information
630 cmdline Kernel command line
631 cpuinfo Info about the CPU
632 devices Available devices (block and character)
633 dma Used DMS channels
634 filesystems Supported filesystems
635 driver Various drivers grouped here, currently rtc (2.4)
636 execdomains Execdomains, related to security (2.4)
637 fb Frame Buffer devices (2.4)
638 fs File system parameters, currently nfs/exports (2.4)
639 ide Directory containing info about the IDE subsystem
640 interrupts Interrupt usage
641 iomem Memory map (2.4)
642 ioports I/O port usage
643 irq Masks for irq to cpu affinity (2.4)(smp?)
644 isapnp ISA PnP (Plug&Play) Info (2.4)
645 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
646 kmsg Kernel messages
647 ksyms Kernel symbol table
648 loadavg Load average of last 1, 5 & 15 minutes
649 locks Kernel locks
650 meminfo Memory info
651 misc Miscellaneous
652 modules List of loaded modules
653 mounts Mounted filesystems
654 net Networking info (see text)
655 pagetypeinfo Additional page allocator information (see text) (2.5)
656 partitions Table of partitions known to the system
657 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
658 decoupled by lspci (2.4)
659 rtc Real time clock
660 scsi SCSI info (see text)
661 slabinfo Slab pool info
662 softirqs softirq usage
663 stat Overall statistics
664 swaps Swap space utilization
665 sys See chapter 2
666 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
667 tty Info of tty drivers
668 uptime Wall clock since boot, combined idle time of all cpus
669 version Kernel version
670 video bttv info of video resources (2.4)
671 vmallocinfo Show vmalloced areas
672 ..............................................................................
674 You can, for example, check which interrupts are currently in use and what
675 they are used for by looking in the file /proc/interrupts:
677 > cat /proc/interrupts
678 CPU0
679 0: 8728810 XT-PIC timer
680 1: 895 XT-PIC keyboard
681 2: 0 XT-PIC cascade
682 3: 531695 XT-PIC aha152x
683 4: 2014133 XT-PIC serial
684 5: 44401 XT-PIC pcnet_cs
685 8: 2 XT-PIC rtc
686 11: 8 XT-PIC i82365
687 12: 182918 XT-PIC PS/2 Mouse
688 13: 1 XT-PIC fpu
689 14: 1232265 XT-PIC ide0
690 15: 7 XT-PIC ide1
691 NMI: 0
693 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
694 output of a SMP machine):
696 > cat /proc/interrupts
698 CPU0 CPU1
699 0: 1243498 1214548 IO-APIC-edge timer
700 1: 8949 8958 IO-APIC-edge keyboard
701 2: 0 0 XT-PIC cascade
702 5: 11286 10161 IO-APIC-edge soundblaster
703 8: 1 0 IO-APIC-edge rtc
704 9: 27422 27407 IO-APIC-edge 3c503
705 12: 113645 113873 IO-APIC-edge PS/2 Mouse
706 13: 0 0 XT-PIC fpu
707 14: 22491 24012 IO-APIC-edge ide0
708 15: 2183 2415 IO-APIC-edge ide1
709 17: 30564 30414 IO-APIC-level eth0
710 18: 177 164 IO-APIC-level bttv
711 NMI: 2457961 2457959
712 LOC: 2457882 2457881
713 ERR: 2155
715 NMI is incremented in this case because every timer interrupt generates a NMI
716 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
718 LOC is the local interrupt counter of the internal APIC of every CPU.
720 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
721 connects the CPUs in a SMP system. This means that an error has been detected,
722 the IO-APIC automatically retry the transmission, so it should not be a big
723 problem, but you should read the SMP-FAQ.
725 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
726 /proc/interrupts to display every IRQ vector in use by the system, not
727 just those considered 'most important'. The new vectors are:
729 THR -- interrupt raised when a machine check threshold counter
730 (typically counting ECC corrected errors of memory or cache) exceeds
731 a configurable threshold. Only available on some systems.
733 TRM -- a thermal event interrupt occurs when a temperature threshold
734 has been exceeded for the CPU. This interrupt may also be generated
735 when the temperature drops back to normal.
737 SPU -- a spurious interrupt is some interrupt that was raised then lowered
738 by some IO device before it could be fully processed by the APIC. Hence
739 the APIC sees the interrupt but does not know what device it came from.
740 For this case the APIC will generate the interrupt with a IRQ vector
741 of 0xff. This might also be generated by chipset bugs.
743 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
744 sent from one CPU to another per the needs of the OS. Typically,
745 their statistics are used by kernel developers and interested users to
746 determine the occurrence of interrupts of the given type.
748 The above IRQ vectors are displayed only when relevant. For example,
749 the threshold vector does not exist on x86_64 platforms. Others are
750 suppressed when the system is a uniprocessor. As of this writing, only
751 i386 and x86_64 platforms support the new IRQ vector displays.
753 Of some interest is the introduction of the /proc/irq directory to 2.4.
754 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
755 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
756 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
757 prof_cpu_mask.
759 For example
760 > ls /proc/irq/
761 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
762 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
763 > ls /proc/irq/0/
764 smp_affinity
766 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
767 IRQ, you can set it by doing:
769 > echo 1 > /proc/irq/10/smp_affinity
771 This means that only the first CPU will handle the IRQ, but you can also echo
772 5 which means that only the first and third CPU can handle the IRQ.
774 The contents of each smp_affinity file is the same by default:
776 > cat /proc/irq/0/smp_affinity
777 ffffffff
779 There is an alternate interface, smp_affinity_list which allows specifying
780 a cpu range instead of a bitmask:
782 > cat /proc/irq/0/smp_affinity_list
783 1024-1031
785 The default_smp_affinity mask applies to all non-active IRQs, which are the
786 IRQs which have not yet been allocated/activated, and hence which lack a
787 /proc/irq/[0-9]* directory.
789 The node file on an SMP system shows the node to which the device using the IRQ
790 reports itself as being attached. This hardware locality information does not
791 include information about any possible driver locality preference.
793 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
794 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
796 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
797 between all the CPUs which are allowed to handle it. As usual the kernel has
798 more info than you and does a better job than you, so the defaults are the
799 best choice for almost everyone. [Note this applies only to those IO-APIC's
800 that support "Round Robin" interrupt distribution.]
802 There are three more important subdirectories in /proc: net, scsi, and sys.
803 The general rule is that the contents, or even the existence of these
804 directories, depend on your kernel configuration. If SCSI is not enabled, the
805 directory scsi may not exist. The same is true with the net, which is there
806 only when networking support is present in the running kernel.
808 The slabinfo file gives information about memory usage at the slab level.
809 Linux uses slab pools for memory management above page level in version 2.2.
810 Commonly used objects have their own slab pool (such as network buffers,
811 directory cache, and so on).
813 ..............................................................................
815 > cat /proc/buddyinfo
817 Node 0, zone DMA 0 4 5 4 4 3 ...
818 Node 0, zone Normal 1 0 0 1 101 8 ...
819 Node 0, zone HighMem 2 0 0 1 1 0 ...
821 External fragmentation is a problem under some workloads, and buddyinfo is a
822 useful tool for helping diagnose these problems. Buddyinfo will give you a
823 clue as to how big an area you can safely allocate, or why a previous
824 allocation failed.
826 Each column represents the number of pages of a certain order which are
827 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
828 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
829 available in ZONE_NORMAL, etc...
831 More information relevant to external fragmentation can be found in
832 pagetypeinfo.
834 > cat /proc/pagetypeinfo
835 Page block order: 9
836 Pages per block: 512
838 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
839 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
840 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
841 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
842 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
843 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
844 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
845 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
846 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
847 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
848 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
850 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
851 Node 0, zone DMA 2 0 5 1 0
852 Node 0, zone DMA32 41 6 967 2 0
854 Fragmentation avoidance in the kernel works by grouping pages of different
855 migrate types into the same contiguous regions of memory called page blocks.
856 A page block is typically the size of the default hugepage size e.g. 2MB on
857 X86-64. By keeping pages grouped based on their ability to move, the kernel
858 can reclaim pages within a page block to satisfy a high-order allocation.
860 The pagetypinfo begins with information on the size of a page block. It
861 then gives the same type of information as buddyinfo except broken down
862 by migrate-type and finishes with details on how many page blocks of each
863 type exist.
865 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
866 from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can
867 make an estimate of the likely number of huge pages that can be allocated
868 at a given point in time. All the "Movable" blocks should be allocatable
869 unless memory has been mlock()'d. Some of the Reclaimable blocks should
870 also be allocatable although a lot of filesystem metadata may have to be
871 reclaimed to achieve this.
873 ..............................................................................
875 meminfo:
877 Provides information about distribution and utilization of memory. This
878 varies by architecture and compile options. The following is from a
879 16GB PIII, which has highmem enabled. You may not have all of these fields.
881 > cat /proc/meminfo
883 MemTotal: 16344972 kB
884 MemFree: 13634064 kB
885 MemAvailable: 14836172 kB
886 Buffers: 3656 kB
887 Cached: 1195708 kB
888 SwapCached: 0 kB
889 Active: 891636 kB
890 Inactive: 1077224 kB
891 HighTotal: 15597528 kB
892 HighFree: 13629632 kB
893 LowTotal: 747444 kB
894 LowFree: 4432 kB
895 SwapTotal: 0 kB
896 SwapFree: 0 kB
897 Dirty: 968 kB
898 Writeback: 0 kB
899 AnonPages: 861800 kB
900 Mapped: 280372 kB
901 Shmem: 644 kB
902 KReclaimable: 168048 kB
903 Slab: 284364 kB
904 SReclaimable: 159856 kB
905 SUnreclaim: 124508 kB
906 PageTables: 24448 kB
907 NFS_Unstable: 0 kB
908 Bounce: 0 kB
909 WritebackTmp: 0 kB
910 CommitLimit: 7669796 kB
911 Committed_AS: 100056 kB
912 VmallocTotal: 112216 kB
913 VmallocUsed: 428 kB
914 VmallocChunk: 111088 kB
915 Percpu: 62080 kB
916 HardwareCorrupted: 0 kB
917 AnonHugePages: 49152 kB
918 ShmemHugePages: 0 kB
919 ShmemPmdMapped: 0 kB
922 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
923 bits and the kernel binary code)
924 MemFree: The sum of LowFree+HighFree
925 MemAvailable: An estimate of how much memory is available for starting new
926 applications, without swapping. Calculated from MemFree,
927 SReclaimable, the size of the file LRU lists, and the low
928 watermarks in each zone.
929 The estimate takes into account that the system needs some
930 page cache to function well, and that not all reclaimable
931 slab will be reclaimable, due to items being in use. The
932 impact of those factors will vary from system to system.
933 Buffers: Relatively temporary storage for raw disk blocks
934 shouldn't get tremendously large (20MB or so)
935 Cached: in-memory cache for files read from the disk (the
936 pagecache). Doesn't include SwapCached
937 SwapCached: Memory that once was swapped out, is swapped back in but
938 still also is in the swapfile (if memory is needed it
939 doesn't need to be swapped out AGAIN because it is already
940 in the swapfile. This saves I/O)
941 Active: Memory that has been used more recently and usually not
942 reclaimed unless absolutely necessary.
943 Inactive: Memory which has been less recently used. It is more
944 eligible to be reclaimed for other purposes
945 HighTotal:
946 HighFree: Highmem is all memory above ~860MB of physical memory
947 Highmem areas are for use by userspace programs, or
948 for the pagecache. The kernel must use tricks to access
949 this memory, making it slower to access than lowmem.
950 LowTotal:
951 LowFree: Lowmem is memory which can be used for everything that
952 highmem can be used for, but it is also available for the
953 kernel's use for its own data structures. Among many
954 other things, it is where everything from the Slab is
955 allocated. Bad things happen when you're out of lowmem.
956 SwapTotal: total amount of swap space available
957 SwapFree: Memory which has been evicted from RAM, and is temporarily
958 on the disk
959 Dirty: Memory which is waiting to get written back to the disk
960 Writeback: Memory which is actively being written back to the disk
961 AnonPages: Non-file backed pages mapped into userspace page tables
962 HardwareCorrupted: The amount of RAM/memory in KB, the kernel identifies as
963 corrupted.
964 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
965 Mapped: files which have been mmaped, such as libraries
966 Shmem: Total memory used by shared memory (shmem) and tmpfs
967 ShmemHugePages: Memory used by shared memory (shmem) and tmpfs allocated
968 with huge pages
969 ShmemPmdMapped: Shared memory mapped into userspace with huge pages
970 KReclaimable: Kernel allocations that the kernel will attempt to reclaim
971 under memory pressure. Includes SReclaimable (below), and other
972 direct allocations with a shrinker.
973 Slab: in-kernel data structures cache
974 SReclaimable: Part of Slab, that might be reclaimed, such as caches
975 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
976 PageTables: amount of memory dedicated to the lowest level of page
977 tables.
978 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
979 storage
980 Bounce: Memory used for block device "bounce buffers"
981 WritebackTmp: Memory used by FUSE for temporary writeback buffers
982 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
983 this is the total amount of memory currently available to
984 be allocated on the system. This limit is only adhered to
985 if strict overcommit accounting is enabled (mode 2 in
986 'vm.overcommit_memory').
987 The CommitLimit is calculated with the following formula:
988 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
989 overcommit_ratio / 100 + [total swap pages]
990 For example, on a system with 1G of physical RAM and 7G
991 of swap with a `vm.overcommit_ratio` of 30 it would
992 yield a CommitLimit of 7.3G.
993 For more details, see the memory overcommit documentation
994 in vm/overcommit-accounting.
995 Committed_AS: The amount of memory presently allocated on the system.
996 The committed memory is a sum of all of the memory which
997 has been allocated by processes, even if it has not been
998 "used" by them as of yet. A process which malloc()'s 1G
999 of memory, but only touches 300M of it will show up as
1000 using 1G. This 1G is memory which has been "committed" to
1001 by the VM and can be used at any time by the allocating
1002 application. With strict overcommit enabled on the system
1003 (mode 2 in 'vm.overcommit_memory'),allocations which would
1004 exceed the CommitLimit (detailed above) will not be permitted.
1005 This is useful if one needs to guarantee that processes will
1006 not fail due to lack of memory once that memory has been
1007 successfully allocated.
1008 VmallocTotal: total size of vmalloc memory area
1009 VmallocUsed: amount of vmalloc area which is used
1010 VmallocChunk: largest contiguous block of vmalloc area which is free
1011 Percpu: Memory allocated to the percpu allocator used to back percpu
1012 allocations. This stat excludes the cost of metadata.
1014 ..............................................................................
1016 vmallocinfo:
1018 Provides information about vmalloced/vmaped areas. One line per area,
1019 containing the virtual address range of the area, size in bytes,
1020 caller information of the creator, and optional information depending
1021 on the kind of area :
1023 pages=nr number of pages
1024 phys=addr if a physical address was specified
1025 ioremap I/O mapping (ioremap() and friends)
1026 vmalloc vmalloc() area
1027 vmap vmap()ed pages
1028 user VM_USERMAP area
1029 vpages buffer for pages pointers was vmalloced (huge area)
1030 N<node>=nr (Only on NUMA kernels)
1031 Number of pages allocated on memory node <node>
1033 > cat /proc/vmallocinfo
1034 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
1035 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
1036 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
1037 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
1038 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
1039 phys=7fee8000 ioremap
1040 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
1041 phys=7fee7000 ioremap
1042 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
1043 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
1044 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
1045 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
1046 pages=2 vmalloc N1=2
1047 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
1048 /0x130 [x_tables] pages=4 vmalloc N0=4
1049 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
1050 pages=14 vmalloc N2=14
1051 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
1052 pages=4 vmalloc N1=4
1053 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
1054 pages=2 vmalloc N1=2
1055 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
1056 pages=10 vmalloc N0=10
1058 ..............................................................................
1060 softirqs:
1062 Provides counts of softirq handlers serviced since boot time, for each cpu.
1064 > cat /proc/softirqs
1065 CPU0 CPU1 CPU2 CPU3
1066 HI: 0 0 0 0
1067 TIMER: 27166 27120 27097 27034
1068 NET_TX: 0 0 0 17
1069 NET_RX: 42 0 0 39
1070 BLOCK: 0 0 107 1121
1071 TASKLET: 0 0 0 290
1072 SCHED: 27035 26983 26971 26746
1073 HRTIMER: 0 0 0 0
1074 RCU: 1678 1769 2178 2250
1077 1.3 IDE devices in /proc/ide
1078 ----------------------------
1080 The subdirectory /proc/ide contains information about all IDE devices of which
1081 the kernel is aware. There is one subdirectory for each IDE controller, the
1082 file drivers and a link for each IDE device, pointing to the device directory
1083 in the controller specific subtree.
1085 The file drivers contains general information about the drivers used for the
1086 IDE devices:
1088 > cat /proc/ide/drivers
1089 ide-cdrom version 4.53
1090 ide-disk version 1.08
1092 More detailed information can be found in the controller specific
1093 subdirectories. These are named ide0, ide1 and so on. Each of these
1094 directories contains the files shown in table 1-6.
1097 Table 1-6: IDE controller info in /proc/ide/ide?
1098 ..............................................................................
1099 File Content
1100 channel IDE channel (0 or 1)
1101 config Configuration (only for PCI/IDE bridge)
1102 mate Mate name
1103 model Type/Chipset of IDE controller
1104 ..............................................................................
1106 Each device connected to a controller has a separate subdirectory in the
1107 controllers directory. The files listed in table 1-7 are contained in these
1108 directories.
1111 Table 1-7: IDE device information
1112 ..............................................................................
1113 File Content
1114 cache The cache
1115 capacity Capacity of the medium (in 512Byte blocks)
1116 driver driver and version
1117 geometry physical and logical geometry
1118 identify device identify block
1119 media media type
1120 model device identifier
1121 settings device setup
1122 smart_thresholds IDE disk management thresholds
1123 smart_values IDE disk management values
1124 ..............................................................................
1126 The most interesting file is settings. This file contains a nice overview of
1127 the drive parameters:
1129 # cat /proc/ide/ide0/hda/settings
1130 name value min max mode
1131 ---- ----- --- --- ----
1132 bios_cyl 526 0 65535 rw
1133 bios_head 255 0 255 rw
1134 bios_sect 63 0 63 rw
1135 breada_readahead 4 0 127 rw
1136 bswap 0 0 1 r
1137 file_readahead 72 0 2097151 rw
1138 io_32bit 0 0 3 rw
1139 keepsettings 0 0 1 rw
1140 max_kb_per_request 122 1 127 rw
1141 multcount 0 0 8 rw
1142 nice1 1 0 1 rw
1143 nowerr 0 0 1 rw
1144 pio_mode write-only 0 255 w
1145 slow 0 0 1 rw
1146 unmaskirq 0 0 1 rw
1147 using_dma 0 0 1 rw
1150 1.4 Networking info in /proc/net
1151 --------------------------------
1153 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1154 additional values you get for IP version 6 if you configure the kernel to
1155 support this. Table 1-9 lists the files and their meaning.
1158 Table 1-8: IPv6 info in /proc/net
1159 ..............................................................................
1160 File Content
1161 udp6 UDP sockets (IPv6)
1162 tcp6 TCP sockets (IPv6)
1163 raw6 Raw device statistics (IPv6)
1164 igmp6 IP multicast addresses, which this host joined (IPv6)
1165 if_inet6 List of IPv6 interface addresses
1166 ipv6_route Kernel routing table for IPv6
1167 rt6_stats Global IPv6 routing tables statistics
1168 sockstat6 Socket statistics (IPv6)
1169 snmp6 Snmp data (IPv6)
1170 ..............................................................................
1173 Table 1-9: Network info in /proc/net
1174 ..............................................................................
1175 File Content
1176 arp Kernel ARP table
1177 dev network devices with statistics
1178 dev_mcast the Layer2 multicast groups a device is listening too
1179 (interface index, label, number of references, number of bound
1180 addresses).
1181 dev_stat network device status
1182 ip_fwchains Firewall chain linkage
1183 ip_fwnames Firewall chain names
1184 ip_masq Directory containing the masquerading tables
1185 ip_masquerade Major masquerading table
1186 netstat Network statistics
1187 raw raw device statistics
1188 route Kernel routing table
1189 rpc Directory containing rpc info
1190 rt_cache Routing cache
1191 snmp SNMP data
1192 sockstat Socket statistics
1193 tcp TCP sockets
1194 udp UDP sockets
1195 unix UNIX domain sockets
1196 wireless Wireless interface data (Wavelan etc)
1197 igmp IP multicast addresses, which this host joined
1198 psched Global packet scheduler parameters.
1199 netlink List of PF_NETLINK sockets
1200 ip_mr_vifs List of multicast virtual interfaces
1201 ip_mr_cache List of multicast routing cache
1202 ..............................................................................
1204 You can use this information to see which network devices are available in
1205 your system and how much traffic was routed over those devices:
1207 > cat /proc/net/dev
1208 Inter-|Receive |[...
1209 face |bytes packets errs drop fifo frame compressed multicast|[...
1210 lo: 908188 5596 0 0 0 0 0 0 [...
1211 ppp0:15475140 20721 410 0 0 410 0 0 [...
1212 eth0: 614530 7085 0 0 0 0 0 1 [...
1214 ...] Transmit
1215 ...] bytes packets errs drop fifo colls carrier compressed
1216 ...] 908188 5596 0 0 0 0 0 0
1217 ...] 1375103 17405 0 0 0 0 0 0
1218 ...] 1703981 5535 0 0 0 3 0 0
1220 In addition, each Channel Bond interface has its own directory. For
1221 example, the bond0 device will have a directory called /proc/net/bond0/.
1222 It will contain information that is specific to that bond, such as the
1223 current slaves of the bond, the link status of the slaves, and how
1224 many times the slaves link has failed.
1226 1.5 SCSI info
1227 -------------
1229 If you have a SCSI host adapter in your system, you'll find a subdirectory
1230 named after the driver for this adapter in /proc/scsi. You'll also see a list
1231 of all recognized SCSI devices in /proc/scsi:
1233 >cat /proc/scsi/scsi
1234 Attached devices:
1235 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1236 Vendor: IBM Model: DGHS09U Rev: 03E0
1237 Type: Direct-Access ANSI SCSI revision: 03
1238 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1239 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1240 Type: CD-ROM ANSI SCSI revision: 02
1243 The directory named after the driver has one file for each adapter found in
1244 the system. These files contain information about the controller, including
1245 the used IRQ and the IO address range. The amount of information shown is
1246 dependent on the adapter you use. The example shows the output for an Adaptec
1247 AHA-2940 SCSI adapter:
1249 > cat /proc/scsi/aic7xxx/0
1251 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1252 Compile Options:
1253 TCQ Enabled By Default : Disabled
1254 AIC7XXX_PROC_STATS : Disabled
1255 AIC7XXX_RESET_DELAY : 5
1256 Adapter Configuration:
1257 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1258 Ultra Wide Controller
1259 PCI MMAPed I/O Base: 0xeb001000
1260 Adapter SEEPROM Config: SEEPROM found and used.
1261 Adaptec SCSI BIOS: Enabled
1262 IRQ: 10
1263 SCBs: Active 0, Max Active 2,
1264 Allocated 15, HW 16, Page 255
1265 Interrupts: 160328
1266 BIOS Control Word: 0x18b6
1267 Adapter Control Word: 0x005b
1268 Extended Translation: Enabled
1269 Disconnect Enable Flags: 0xffff
1270 Ultra Enable Flags: 0x0001
1271 Tag Queue Enable Flags: 0x0000
1272 Ordered Queue Tag Flags: 0x0000
1273 Default Tag Queue Depth: 8
1274 Tagged Queue By Device array for aic7xxx host instance 0:
1275 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1276 Actual queue depth per device for aic7xxx host instance 0:
1277 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1278 Statistics:
1279 (scsi0:0:0:0)
1280 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1281 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1282 Total transfers 160151 (74577 reads and 85574 writes)
1283 (scsi0:0:6:0)
1284 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1285 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1286 Total transfers 0 (0 reads and 0 writes)
1289 1.6 Parallel port info in /proc/parport
1290 ---------------------------------------
1292 The directory /proc/parport contains information about the parallel ports of
1293 your system. It has one subdirectory for each port, named after the port
1294 number (0,1,2,...).
1296 These directories contain the four files shown in Table 1-10.
1299 Table 1-10: Files in /proc/parport
1300 ..............................................................................
1301 File Content
1302 autoprobe Any IEEE-1284 device ID information that has been acquired.
1303 devices list of the device drivers using that port. A + will appear by the
1304 name of the device currently using the port (it might not appear
1305 against any).
1306 hardware Parallel port's base address, IRQ line and DMA channel.
1307 irq IRQ that parport is using for that port. This is in a separate
1308 file to allow you to alter it by writing a new value in (IRQ
1309 number or none).
1310 ..............................................................................
1312 1.7 TTY info in /proc/tty
1313 -------------------------
1315 Information about the available and actually used tty's can be found in the
1316 directory /proc/tty.You'll find entries for drivers and line disciplines in
1317 this directory, as shown in Table 1-11.
1320 Table 1-11: Files in /proc/tty
1321 ..............................................................................
1322 File Content
1323 drivers list of drivers and their usage
1324 ldiscs registered line disciplines
1325 driver/serial usage statistic and status of single tty lines
1326 ..............................................................................
1328 To see which tty's are currently in use, you can simply look into the file
1329 /proc/tty/drivers:
1331 > cat /proc/tty/drivers
1332 pty_slave /dev/pts 136 0-255 pty:slave
1333 pty_master /dev/ptm 128 0-255 pty:master
1334 pty_slave /dev/ttyp 3 0-255 pty:slave
1335 pty_master /dev/pty 2 0-255 pty:master
1336 serial /dev/cua 5 64-67 serial:callout
1337 serial /dev/ttyS 4 64-67 serial
1338 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1339 /dev/ptmx /dev/ptmx 5 2 system
1340 /dev/console /dev/console 5 1 system:console
1341 /dev/tty /dev/tty 5 0 system:/dev/tty
1342 unknown /dev/tty 4 1-63 console
1345 1.8 Miscellaneous kernel statistics in /proc/stat
1346 -------------------------------------------------
1348 Various pieces of information about kernel activity are available in the
1349 /proc/stat file. All of the numbers reported in this file are aggregates
1350 since the system first booted. For a quick look, simply cat the file:
1352 > cat /proc/stat
1353 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1354 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1355 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1356 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1357 ctxt 1990473
1358 btime 1062191376
1359 processes 2915
1360 procs_running 1
1361 procs_blocked 0
1362 softirq 183433 0 21755 12 39 1137 231 21459 2263
1364 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1365 lines. These numbers identify the amount of time the CPU has spent performing
1366 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1367 second). The meanings of the columns are as follows, from left to right:
1369 - user: normal processes executing in user mode
1370 - nice: niced processes executing in user mode
1371 - system: processes executing in kernel mode
1372 - idle: twiddling thumbs
1373 - iowait: In a word, iowait stands for waiting for I/O to complete. But there
1374 are several problems:
1375 1. Cpu will not wait for I/O to complete, iowait is the time that a task is
1376 waiting for I/O to complete. When cpu goes into idle state for
1377 outstanding task io, another task will be scheduled on this CPU.
1378 2. In a multi-core CPU, the task waiting for I/O to complete is not running
1379 on any CPU, so the iowait of each CPU is difficult to calculate.
1380 3. The value of iowait field in /proc/stat will decrease in certain
1381 conditions.
1382 So, the iowait is not reliable by reading from /proc/stat.
1383 - irq: servicing interrupts
1384 - softirq: servicing softirqs
1385 - steal: involuntary wait
1386 - guest: running a normal guest
1387 - guest_nice: running a niced guest
1389 The "intr" line gives counts of interrupts serviced since boot time, for each
1390 of the possible system interrupts. The first column is the total of all
1391 interrupts serviced including unnumbered architecture specific interrupts;
1392 each subsequent column is the total for that particular numbered interrupt.
1393 Unnumbered interrupts are not shown, only summed into the total.
1395 The "ctxt" line gives the total number of context switches across all CPUs.
1397 The "btime" line gives the time at which the system booted, in seconds since
1398 the Unix epoch.
1400 The "processes" line gives the number of processes and threads created, which
1401 includes (but is not limited to) those created by calls to the fork() and
1402 clone() system calls.
1404 The "procs_running" line gives the total number of threads that are
1405 running or ready to run (i.e., the total number of runnable threads).
1407 The "procs_blocked" line gives the number of processes currently blocked,
1408 waiting for I/O to complete.
1410 The "softirq" line gives counts of softirqs serviced since boot time, for each
1411 of the possible system softirqs. The first column is the total of all
1412 softirqs serviced; each subsequent column is the total for that particular
1413 softirq.
1416 1.9 Ext4 file system parameters
1417 -------------------------------
1419 Information about mounted ext4 file systems can be found in
1420 /proc/fs/ext4. Each mounted filesystem will have a directory in
1421 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1422 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1423 in Table 1-12, below.
1425 Table 1-12: Files in /proc/fs/ext4/<devname>
1426 ..............................................................................
1427 File Content
1428 mb_groups details of multiblock allocator buddy cache of free blocks
1429 ..............................................................................
1431 2.0 /proc/consoles
1432 ------------------
1433 Shows registered system console lines.
1435 To see which character device lines are currently used for the system console
1436 /dev/console, you may simply look into the file /proc/consoles:
1438 > cat /proc/consoles
1439 tty0 -WU (ECp) 4:7
1440 ttyS0 -W- (Ep) 4:64
1442 The columns are:
1444 device name of the device
1445 operations R = can do read operations
1446 W = can do write operations
1447 U = can do unblank
1448 flags E = it is enabled
1449 C = it is preferred console
1450 B = it is primary boot console
1451 p = it is used for printk buffer
1452 b = it is not a TTY but a Braille device
1453 a = it is safe to use when cpu is offline
1454 major:minor major and minor number of the device separated by a colon
1456 ------------------------------------------------------------------------------
1457 Summary
1458 ------------------------------------------------------------------------------
1459 The /proc file system serves information about the running system. It not only
1460 allows access to process data but also allows you to request the kernel status
1461 by reading files in the hierarchy.
1463 The directory structure of /proc reflects the types of information and makes
1464 it easy, if not obvious, where to look for specific data.
1465 ------------------------------------------------------------------------------
1467 ------------------------------------------------------------------------------
1468 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1469 ------------------------------------------------------------------------------
1471 ------------------------------------------------------------------------------
1472 In This Chapter
1473 ------------------------------------------------------------------------------
1474 * Modifying kernel parameters by writing into files found in /proc/sys
1475 * Exploring the files which modify certain parameters
1476 * Review of the /proc/sys file tree
1477 ------------------------------------------------------------------------------
1480 A very interesting part of /proc is the directory /proc/sys. This is not only
1481 a source of information, it also allows you to change parameters within the
1482 kernel. Be very careful when attempting this. You can optimize your system,
1483 but you can also cause it to crash. Never alter kernel parameters on a
1484 production system. Set up a development machine and test to make sure that
1485 everything works the way you want it to. You may have no alternative but to
1486 reboot the machine once an error has been made.
1488 To change a value, simply echo the new value into the file. An example is
1489 given below in the section on the file system data. You need to be root to do
1490 this. You can create your own boot script to perform this every time your
1491 system boots.
1493 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1494 general things in the operation of the Linux kernel. Since some of the files
1495 can inadvertently disrupt your system, it is advisable to read both
1496 documentation and source before actually making adjustments. In any case, be
1497 very careful when writing to any of these files. The entries in /proc may
1498 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1499 review the kernel documentation in the directory /usr/src/linux/Documentation.
1500 This chapter is heavily based on the documentation included in the pre 2.2
1501 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1503 Please see: Documentation/admin-guide/sysctl/ directory for descriptions of these
1504 entries.
1506 ------------------------------------------------------------------------------
1507 Summary
1508 ------------------------------------------------------------------------------
1509 Certain aspects of kernel behavior can be modified at runtime, without the
1510 need to recompile the kernel, or even to reboot the system. The files in the
1511 /proc/sys tree can not only be read, but also modified. You can use the echo
1512 command to write value into these files, thereby changing the default settings
1513 of the kernel.
1514 ------------------------------------------------------------------------------
1516 ------------------------------------------------------------------------------
1517 CHAPTER 3: PER-PROCESS PARAMETERS
1518 ------------------------------------------------------------------------------
1520 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1521 --------------------------------------------------------------------------------
1523 These file can be used to adjust the badness heuristic used to select which
1524 process gets killed in out of memory conditions.
1526 The badness heuristic assigns a value to each candidate task ranging from 0
1527 (never kill) to 1000 (always kill) to determine which process is targeted. The
1528 units are roughly a proportion along that range of allowed memory the process
1529 may allocate from based on an estimation of its current memory and swap use.
1530 For example, if a task is using all allowed memory, its badness score will be
1531 1000. If it is using half of its allowed memory, its score will be 500.
1533 There is an additional factor included in the badness score: the current memory
1534 and swap usage is discounted by 3% for root processes.
1536 The amount of "allowed" memory depends on the context in which the oom killer
1537 was called. If it is due to the memory assigned to the allocating task's cpuset
1538 being exhausted, the allowed memory represents the set of mems assigned to that
1539 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1540 memory represents the set of mempolicy nodes. If it is due to a memory
1541 limit (or swap limit) being reached, the allowed memory is that configured
1542 limit. Finally, if it is due to the entire system being out of memory, the
1543 allowed memory represents all allocatable resources.
1545 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1546 is used to determine which task to kill. Acceptable values range from -1000
1547 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1548 polarize the preference for oom killing either by always preferring a certain
1549 task or completely disabling it. The lowest possible value, -1000, is
1550 equivalent to disabling oom killing entirely for that task since it will always
1551 report a badness score of 0.
1553 Consequently, it is very simple for userspace to define the amount of memory to
1554 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1555 example, is roughly equivalent to allowing the remainder of tasks sharing the
1556 same system, cpuset, mempolicy, or memory controller resources to use at least
1557 50% more memory. A value of -500, on the other hand, would be roughly
1558 equivalent to discounting 50% of the task's allowed memory from being considered
1559 as scoring against the task.
1561 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1562 be used to tune the badness score. Its acceptable values range from -16
1563 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1564 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1565 scaled linearly with /proc/<pid>/oom_score_adj.
1567 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1568 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1569 requires CAP_SYS_RESOURCE.
1571 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1572 generation children with separate address spaces instead, if possible. This
1573 avoids servers and important system daemons from being killed and loses the
1574 minimal amount of work.
1577 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1578 -------------------------------------------------------------
1580 This file can be used to check the current score used by the oom-killer is for
1581 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1582 process should be killed in an out-of-memory situation.
1585 3.3 /proc/<pid>/io - Display the IO accounting fields
1586 -------------------------------------------------------
1588 This file contains IO statistics for each running process
1590 Example
1591 -------
1593 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1594 [1] 3828
1596 test:/tmp # cat /proc/3828/io
1597 rchar: 323934931
1598 wchar: 323929600
1599 syscr: 632687
1600 syscw: 632675
1601 read_bytes: 0
1602 write_bytes: 323932160
1603 cancelled_write_bytes: 0
1606 Description
1607 -----------
1609 rchar
1610 -----
1612 I/O counter: chars read
1613 The number of bytes which this task has caused to be read from storage. This
1614 is simply the sum of bytes which this process passed to read() and pread().
1615 It includes things like tty IO and it is unaffected by whether or not actual
1616 physical disk IO was required (the read might have been satisfied from
1617 pagecache)
1620 wchar
1621 -----
1623 I/O counter: chars written
1624 The number of bytes which this task has caused, or shall cause to be written
1625 to disk. Similar caveats apply here as with rchar.
1628 syscr
1629 -----
1631 I/O counter: read syscalls
1632 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1633 and pread().
1636 syscw
1637 -----
1639 I/O counter: write syscalls
1640 Attempt to count the number of write I/O operations, i.e. syscalls like
1641 write() and pwrite().
1644 read_bytes
1645 ----------
1647 I/O counter: bytes read
1648 Attempt to count the number of bytes which this process really did cause to
1649 be fetched from the storage layer. Done at the submit_bio() level, so it is
1650 accurate for block-backed filesystems. <please add status regarding NFS and
1651 CIFS at a later time>
1654 write_bytes
1655 -----------
1657 I/O counter: bytes written
1658 Attempt to count the number of bytes which this process caused to be sent to
1659 the storage layer. This is done at page-dirtying time.
1662 cancelled_write_bytes
1663 ---------------------
1665 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1666 then deletes the file, it will in fact perform no writeout. But it will have
1667 been accounted as having caused 1MB of write.
1668 In other words: The number of bytes which this process caused to not happen,
1669 by truncating pagecache. A task can cause "negative" IO too. If this task
1670 truncates some dirty pagecache, some IO which another task has been accounted
1671 for (in its write_bytes) will not be happening. We _could_ just subtract that
1672 from the truncating task's write_bytes, but there is information loss in doing
1673 that.
1676 Note
1677 ----
1679 At its current implementation state, this is a bit racy on 32-bit machines: if
1680 process A reads process B's /proc/pid/io while process B is updating one of
1681 those 64-bit counters, process A could see an intermediate result.
1684 More information about this can be found within the taskstats documentation in
1685 Documentation/accounting.
1687 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1688 ---------------------------------------------------------------
1689 When a process is dumped, all anonymous memory is written to a core file as
1690 long as the size of the core file isn't limited. But sometimes we don't want
1691 to dump some memory segments, for example, huge shared memory or DAX.
1692 Conversely, sometimes we want to save file-backed memory segments into a core
1693 file, not only the individual files.
1695 /proc/<pid>/coredump_filter allows you to customize which memory segments
1696 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1697 of memory types. If a bit of the bitmask is set, memory segments of the
1698 corresponding memory type are dumped, otherwise they are not dumped.
1700 The following 9 memory types are supported:
1701 - (bit 0) anonymous private memory
1702 - (bit 1) anonymous shared memory
1703 - (bit 2) file-backed private memory
1704 - (bit 3) file-backed shared memory
1705 - (bit 4) ELF header pages in file-backed private memory areas (it is
1706 effective only if the bit 2 is cleared)
1707 - (bit 5) hugetlb private memory
1708 - (bit 6) hugetlb shared memory
1709 - (bit 7) DAX private memory
1710 - (bit 8) DAX shared memory
1712 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1713 are always dumped regardless of the bitmask status.
1715 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is
1716 only affected by bit 5-6, and DAX is only affected by bits 7-8.
1718 The default value of coredump_filter is 0x33; this means all anonymous memory
1719 segments, ELF header pages and hugetlb private memory are dumped.
1721 If you don't want to dump all shared memory segments attached to pid 1234,
1722 write 0x31 to the process's proc file.
1724 $ echo 0x31 > /proc/1234/coredump_filter
1726 When a new process is created, the process inherits the bitmask status from its
1727 parent. It is useful to set up coredump_filter before the program runs.
1728 For example:
1730 $ echo 0x7 > /proc/self/coredump_filter
1731 $ ./some_program
1733 3.5 /proc/<pid>/mountinfo - Information about mounts
1734 --------------------------------------------------------
1736 This file contains lines of the form:
1738 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1739 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1741 (1) mount ID: unique identifier of the mount (may be reused after umount)
1742 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1743 (3) major:minor: value of st_dev for files on filesystem
1744 (4) root: root of the mount within the filesystem
1745 (5) mount point: mount point relative to the process's root
1746 (6) mount options: per mount options
1747 (7) optional fields: zero or more fields of the form "tag[:value]"
1748 (8) separator: marks the end of the optional fields
1749 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1750 (10) mount source: filesystem specific information or "none"
1751 (11) super options: per super block options
1753 Parsers should ignore all unrecognised optional fields. Currently the
1754 possible optional fields are:
1756 shared:X mount is shared in peer group X
1757 master:X mount is slave to peer group X
1758 propagate_from:X mount is slave and receives propagation from peer group X (*)
1759 unbindable mount is unbindable
1761 (*) X is the closest dominant peer group under the process's root. If
1762 X is the immediate master of the mount, or if there's no dominant peer
1763 group under the same root, then only the "master:X" field is present
1764 and not the "propagate_from:X" field.
1766 For more information on mount propagation see:
1768 Documentation/filesystems/sharedsubtree.txt
1771 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1772 --------------------------------------------------------
1773 These files provide a method to access a tasks comm value. It also allows for
1774 a task to set its own or one of its thread siblings comm value. The comm value
1775 is limited in size compared to the cmdline value, so writing anything longer
1776 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1777 comm value.
1780 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1781 -------------------------------------------------------------------------
1782 This file provides a fast way to retrieve first level children pids
1783 of a task pointed by <pid>/<tid> pair. The format is a space separated
1784 stream of pids.
1786 Note the "first level" here -- if a child has own children they will
1787 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1788 to obtain the descendants.
1790 Since this interface is intended to be fast and cheap it doesn't
1791 guarantee to provide precise results and some children might be
1792 skipped, especially if they've exited right after we printed their
1793 pids, so one need to either stop or freeze processes being inspected
1794 if precise results are needed.
1797 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1798 ---------------------------------------------------------------
1799 This file provides information associated with an opened file. The regular
1800 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1801 represents the current offset of the opened file in decimal form [see lseek(2)
1802 for details], 'flags' denotes the octal O_xxx mask the file has been
1803 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1804 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1805 for details].
1807 A typical output is
1809 pos: 0
1810 flags: 0100002
1811 mnt_id: 19
1813 All locks associated with a file descriptor are shown in its fdinfo too.
1815 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1817 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1818 pair provide additional information particular to the objects they represent.
1820 Eventfd files
1821 ~~~~~~~~~~~~~
1822 pos: 0
1823 flags: 04002
1824 mnt_id: 9
1825 eventfd-count: 5a
1827 where 'eventfd-count' is hex value of a counter.
1829 Signalfd files
1830 ~~~~~~~~~~~~~~
1831 pos: 0
1832 flags: 04002
1833 mnt_id: 9
1834 sigmask: 0000000000000200
1836 where 'sigmask' is hex value of the signal mask associated
1837 with a file.
1839 Epoll files
1840 ~~~~~~~~~~~
1841 pos: 0
1842 flags: 02
1843 mnt_id: 9
1844 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7
1846 where 'tfd' is a target file descriptor number in decimal form,
1847 'events' is events mask being watched and the 'data' is data
1848 associated with a target [see epoll(7) for more details].
1850 The 'pos' is current offset of the target file in decimal form
1851 [see lseek(2)], 'ino' and 'sdev' are inode and device numbers
1852 where target file resides, all in hex format.
1854 Fsnotify files
1855 ~~~~~~~~~~~~~~
1856 For inotify files the format is the following
1858 pos: 0
1859 flags: 02000000
1860 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1862 where 'wd' is a watch descriptor in decimal form, ie a target file
1863 descriptor number, 'ino' and 'sdev' are inode and device where the
1864 target file resides and the 'mask' is the mask of events, all in hex
1865 form [see inotify(7) for more details].
1867 If the kernel was built with exportfs support, the path to the target
1868 file is encoded as a file handle. The file handle is provided by three
1869 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1870 format.
1872 If the kernel is built without exportfs support the file handle won't be
1873 printed out.
1875 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1877 For fanotify files the format is
1879 pos: 0
1880 flags: 02
1881 mnt_id: 9
1882 fanotify flags:10 event-flags:0
1883 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1884 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1886 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1887 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1888 flags associated with mark which are tracked separately from events
1889 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1890 mask and 'ignored_mask' is the mask of events which are to be ignored.
1891 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1892 does provide information about flags and mask used in fanotify_mark
1893 call [see fsnotify manpage for details].
1895 While the first three lines are mandatory and always printed, the rest is
1896 optional and may be omitted if no marks created yet.
1898 Timerfd files
1899 ~~~~~~~~~~~~~
1901 pos: 0
1902 flags: 02
1903 mnt_id: 9
1904 clockid: 0
1905 ticks: 0
1906 settime flags: 01
1907 it_value: (0, 49406829)
1908 it_interval: (1, 0)
1910 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1911 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1912 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1913 details]. 'it_value' is remaining time until the timer exiration.
1914 'it_interval' is the interval for the timer. Note the timer might be set up
1915 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1916 still exhibits timer's remaining time.
1918 3.9 /proc/<pid>/map_files - Information about memory mapped files
1919 ---------------------------------------------------------------------
1920 This directory contains symbolic links which represent memory mapped files
1921 the process is maintaining. Example output:
1923 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1924 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1925 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1926 | ...
1927 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1928 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1930 The name of a link represents the virtual memory bounds of a mapping, i.e.
1931 vm_area_struct::vm_start-vm_area_struct::vm_end.
1933 The main purpose of the map_files is to retrieve a set of memory mapped
1934 files in a fast way instead of parsing /proc/<pid>/maps or
1935 /proc/<pid>/smaps, both of which contain many more records. At the same
1936 time one can open(2) mappings from the listings of two processes and
1937 comparing their inode numbers to figure out which anonymous memory areas
1938 are actually shared.
1940 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
1941 ---------------------------------------------------------
1942 This file provides the value of the task's timerslack value in nanoseconds.
1943 This value specifies a amount of time that normal timers may be deferred
1944 in order to coalesce timers and avoid unnecessary wakeups.
1946 This allows a task's interactivity vs power consumption trade off to be
1947 adjusted.
1949 Writing 0 to the file will set the tasks timerslack to the default value.
1951 Valid values are from 0 - ULLONG_MAX
1953 An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level
1954 permissions on the task specified to change its timerslack_ns value.
1956 3.11 /proc/<pid>/patch_state - Livepatch patch operation state
1957 -----------------------------------------------------------------
1958 When CONFIG_LIVEPATCH is enabled, this file displays the value of the
1959 patch state for the task.
1961 A value of '-1' indicates that no patch is in transition.
1963 A value of '0' indicates that a patch is in transition and the task is
1964 unpatched. If the patch is being enabled, then the task hasn't been
1965 patched yet. If the patch is being disabled, then the task has already
1966 been unpatched.
1968 A value of '1' indicates that a patch is in transition and the task is
1969 patched. If the patch is being enabled, then the task has already been
1970 patched. If the patch is being disabled, then the task hasn't been
1971 unpatched yet.
1973 3.12 /proc/<pid>/arch_status - task architecture specific status
1974 -------------------------------------------------------------------
1975 When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the
1976 architecture specific status of the task.
1978 Example
1979 -------
1980 $ cat /proc/6753/arch_status
1981 AVX512_elapsed_ms: 8
1983 Description
1984 -----------
1986 x86 specific entries:
1987 ---------------------
1988 AVX512_elapsed_ms:
1989 ------------------
1990 If AVX512 is supported on the machine, this entry shows the milliseconds
1991 elapsed since the last time AVX512 usage was recorded. The recording
1992 happens on a best effort basis when a task is scheduled out. This means
1993 that the value depends on two factors:
1995 1) The time which the task spent on the CPU without being scheduled
1996 out. With CPU isolation and a single runnable task this can take
1997 several seconds.
1999 2) The time since the task was scheduled out last. Depending on the
2000 reason for being scheduled out (time slice exhausted, syscall ...)
2001 this can be arbitrary long time.
2003 As a consequence the value cannot be considered precise and authoritative
2004 information. The application which uses this information has to be aware
2005 of the overall scenario on the system in order to determine whether a
2006 task is a real AVX512 user or not. Precise information can be obtained
2007 with performance counters.
2009 A special value of '-1' indicates that no AVX512 usage was recorded, thus
2010 the task is unlikely an AVX512 user, but depends on the workload and the
2011 scheduling scenario, it also could be a false negative mentioned above.
2013 ------------------------------------------------------------------------------
2014 Configuring procfs
2015 ------------------------------------------------------------------------------
2017 4.1 Mount options
2018 ---------------------
2020 The following mount options are supported:
2022 hidepid= Set /proc/<pid>/ access mode.
2023 gid= Set the group authorized to learn processes information.
2025 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
2026 (default).
2028 hidepid=1 means users may not access any /proc/<pid>/ directories but their
2029 own. Sensitive files like cmdline, sched*, status are now protected against
2030 other users. This makes it impossible to learn whether any user runs
2031 specific program (given the program doesn't reveal itself by its behaviour).
2032 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
2033 poorly written programs passing sensitive information via program arguments are
2034 now protected against local eavesdroppers.
2036 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
2037 users. It doesn't mean that it hides a fact whether a process with a specific
2038 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
2039 but it hides process' uid and gid, which may be learned by stat()'ing
2040 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
2041 information about running processes, whether some daemon runs with elevated
2042 privileges, whether other user runs some sensitive program, whether other users
2043 run any program at all, etc.
2045 gid= defines a group authorized to learn processes information otherwise
2046 prohibited by hidepid=. If you use some daemon like identd which needs to learn
2047 information about processes information, just add identd to this group.