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
19 0.1 Introduction/Credits
22 1 Collecting System Information
23 1.1 Process-Specific Subdirectories
25 1.3 IDE devices in /proc/ide
26 1.4 Networking info in /proc/net
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
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
51 ------------------------------------------------------------------------------
53 ------------------------------------------------------------------------------
55 0.1 Introduction/Credits
56 ------------------------
58 This documentation is part of a soon (or so we hope) to be released book on
59 the SuSE Linux distribution. As there is no complete documentation for the
60 /proc file system and we've used many freely available sources to write these
61 chapters, it seems only fair to give the work back to the Linux community.
62 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
63 afraid it's still far from complete, but we hope it will be useful. As far as
64 we know, it is the first 'all-in-one' document about the /proc file system. It
65 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
66 SPARC, AXP, etc., features, you probably won't find what you are looking for.
67 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
68 additions and patches are welcome and will be added to this document if you
71 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
72 other people for help compiling this documentation. We'd also like to extend a
73 special thank you to Andi Kleen for documentation, which we relied on heavily
74 to create this document, as well as the additional information he provided.
75 Thanks to everybody else who contributed source or docs to the Linux kernel
76 and helped create a great piece of software... :)
78 If you have any comments, corrections or additions, please don't hesitate to
79 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
82 The latest version of this document is available online at
83 http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
85 If the above direction does not works for you, you could try the kernel
86 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
87 comandante@zaralinux.com.
92 We don't guarantee the correctness of this document, and if you come to us
93 complaining about how you screwed up your system because of incorrect
94 documentation, we won't feel responsible...
96 ------------------------------------------------------------------------------
97 CHAPTER 1: COLLECTING SYSTEM INFORMATION
98 ------------------------------------------------------------------------------
100 ------------------------------------------------------------------------------
102 ------------------------------------------------------------------------------
103 * Investigating the properties of the pseudo file system /proc and its
104 ability to provide information on the running Linux system
105 * Examining /proc's structure
106 * Uncovering various information about the kernel and the processes running
108 ------------------------------------------------------------------------------
111 The proc file system acts as an interface to internal data structures in the
112 kernel. It can be used to obtain information about the system and to change
113 certain kernel parameters at runtime (sysctl).
115 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
116 show you how you can use /proc/sys to change settings.
118 1.1 Process-Specific Subdirectories
119 -----------------------------------
121 The directory /proc contains (among other things) one subdirectory for each
122 process running on the system, which is named after the process ID (PID).
124 The link self points to the process reading the file system. Each process
125 subdirectory has the entries listed in Table 1-1.
128 Table 1-1: Process specific entries in /proc
129 ..............................................................................
131 clear_refs Clears page referenced bits shown in smaps output
132 cmdline Command line arguments
133 cpu Current and last cpu in which it was executed (2.4)(smp)
134 cwd Link to the current working directory
135 environ Values of environment variables
136 exe Link to the executable of this process
137 fd Directory, which contains all file descriptors
138 maps Memory maps to executables and library files (2.4)
139 mem Memory held by this process
140 root Link to the root directory of this process
142 statm Process memory status information
143 status Process status in human readable form
144 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function
145 symbol the task is blocked in - or "0" if not blocked.
147 stack Report full stack trace, enable via CONFIG_STACKTRACE
148 smaps an extension based on maps, showing the memory consumption of
149 each mapping and flags associated with it
150 numa_maps an extension based on maps, showing the memory locality and
151 binding policy as well as mem usage (in pages) of each mapping.
152 ..............................................................................
154 For example, to get the status information of a process, all you have to do is
155 read the file /proc/PID/status:
157 >cat /proc/self/status
185 SigPnd: 0000000000000000
186 ShdPnd: 0000000000000000
187 SigBlk: 0000000000000000
188 SigIgn: 0000000000000000
189 SigCgt: 0000000000000000
190 CapInh: 00000000fffffeff
191 CapPrm: 0000000000000000
192 CapEff: 0000000000000000
193 CapBnd: ffffffffffffffff
195 voluntary_ctxt_switches: 0
196 nonvoluntary_ctxt_switches: 1
198 This shows you nearly the same information you would get if you viewed it with
199 the ps command. In fact, ps uses the proc file system to obtain its
200 information. But you get a more detailed view of the process by reading the
201 file /proc/PID/status. It fields are described in table 1-2.
203 The statm file contains more detailed information about the process
204 memory usage. Its seven fields are explained in Table 1-3. The stat file
205 contains details information about the process itself. Its fields are
206 explained in Table 1-4.
208 (for SMP CONFIG users)
209 For making accounting scalable, RSS related information are handled in an
210 asynchronous manner and the value may not be very precise. To see a precise
211 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
212 It's slow but very precise.
214 Table 1-2: Contents of the status files (as of 4.1)
215 ..............................................................................
217 Name filename of the executable
218 State state (R is running, S is sleeping, D is sleeping
219 in an uninterruptible wait, Z is zombie,
220 T is traced or stopped)
222 Ngid NUMA group ID (0 if none)
224 PPid process id of the parent process
225 TracerPid PID of process tracing this process (0 if not)
226 Uid Real, effective, saved set, and file system UIDs
227 Gid Real, effective, saved set, and file system GIDs
228 Umask file mode creation mask
229 FDSize number of file descriptor slots currently allocated
230 Groups supplementary group list
231 NStgid descendant namespace thread group ID hierarchy
232 NSpid descendant namespace process ID hierarchy
233 NSpgid descendant namespace process group ID hierarchy
234 NSsid descendant namespace session ID hierarchy
235 VmPeak peak virtual memory size
236 VmSize total program size
237 VmLck locked memory size
238 VmHWM peak resident set size ("high water mark")
239 VmRSS size of memory portions. It contains the three
240 following parts (VmRSS = RssAnon + RssFile + RssShmem)
241 RssAnon size of resident anonymous memory
242 RssFile size of resident file mappings
243 RssShmem size of resident shmem memory (includes SysV shm,
244 mapping of tmpfs and shared anonymous mappings)
245 VmData size of private data segments
246 VmStk size of stack segments
247 VmExe size of text segment
248 VmLib size of shared library code
249 VmPTE size of page table entries
250 VmPMD size of second level page tables
251 VmSwap amount of swap used by anonymous private data
252 (shmem swap usage is not included)
253 HugetlbPages size of hugetlb memory portions
254 Threads number of threads
255 SigQ number of signals queued/max. number for queue
256 SigPnd bitmap of pending signals for the thread
257 ShdPnd bitmap of shared pending signals for the process
258 SigBlk bitmap of blocked signals
259 SigIgn bitmap of ignored signals
260 SigCgt bitmap of caught signals
261 CapInh bitmap of inheritable capabilities
262 CapPrm bitmap of permitted capabilities
263 CapEff bitmap of effective capabilities
264 CapBnd bitmap of capabilities bounding set
265 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
266 Cpus_allowed mask of CPUs on which this process may run
267 Cpus_allowed_list Same as previous, but in "list format"
268 Mems_allowed mask of memory nodes allowed to this process
269 Mems_allowed_list Same as previous, but in "list format"
270 voluntary_ctxt_switches number of voluntary context switches
271 nonvoluntary_ctxt_switches number of non voluntary context switches
272 ..............................................................................
274 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
275 ..............................................................................
277 size total program size (pages) (same as VmSize in status)
278 resident size of memory portions (pages) (same as VmRSS in status)
279 shared number of pages that are shared (i.e. backed by a file, same
280 as RssFile+RssShmem in status)
281 trs number of pages that are 'code' (not including libs; broken,
282 includes data segment)
283 lrs number of pages of library (always 0 on 2.6)
284 drs number of pages of data/stack (including libs; broken,
285 includes library text)
286 dt number of dirty pages (always 0 on 2.6)
287 ..............................................................................
290 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
291 ..............................................................................
294 tcomm filename of the executable
295 state state (R is running, S is sleeping, D is sleeping in an
296 uninterruptible wait, Z is zombie, T is traced or stopped)
297 ppid process id of the parent process
298 pgrp pgrp of the process
300 tty_nr tty the process uses
301 tty_pgrp pgrp of the tty
303 min_flt number of minor faults
304 cmin_flt number of minor faults with child's
305 maj_flt number of major faults
306 cmaj_flt number of major faults with child's
307 utime user mode jiffies
308 stime kernel mode jiffies
309 cutime user mode jiffies with child's
310 cstime kernel mode jiffies with child's
311 priority priority level
313 num_threads number of threads
314 it_real_value (obsolete, always 0)
315 start_time time the process started after system boot
316 vsize virtual memory size
317 rss resident set memory size
318 rsslim current limit in bytes on the rss
319 start_code address above which program text can run
320 end_code address below which program text can run
321 start_stack address of the start of the main process stack
322 esp current value of ESP
323 eip current value of EIP
324 pending bitmap of pending signals
325 blocked bitmap of blocked signals
326 sigign bitmap of ignored signals
327 sigcatch bitmap of caught signals
328 0 (place holder, used to be the wchan address, use /proc/PID/wchan instead)
331 exit_signal signal to send to parent thread on exit
332 task_cpu which CPU the task is scheduled on
333 rt_priority realtime priority
334 policy scheduling policy (man sched_setscheduler)
335 blkio_ticks time spent waiting for block IO
336 gtime guest time of the task in jiffies
337 cgtime guest time of the task children in jiffies
338 start_data address above which program data+bss is placed
339 end_data address below which program data+bss is placed
340 start_brk address above which program heap can be expanded with brk()
341 arg_start address above which program command line is placed
342 arg_end address below which program command line is placed
343 env_start address above which program environment is placed
344 env_end address below which program environment is placed
345 exit_code the thread's exit_code in the form reported by the waitpid system call
346 ..............................................................................
348 The /proc/PID/maps file containing the currently mapped memory regions and
349 their access permissions.
353 address perms offset dev inode pathname
355 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
356 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
357 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
358 a7cb1000-a7cb2000 ---p 00000000 00:00 0
359 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
360 a7eb2000-a7eb3000 ---p 00000000 00:00 0
361 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
362 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
363 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
364 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
365 a800b000-a800e000 rw-p 00000000 00:00 0
366 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
367 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
368 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
369 a8024000-a8027000 rw-p 00000000 00:00 0
370 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
371 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
372 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
373 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
374 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
376 where "address" is the address space in the process that it occupies, "perms"
377 is a set of permissions:
383 p = private (copy on write)
385 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
386 "inode" is the inode on that device. 0 indicates that no inode is associated
387 with the memory region, as the case would be with BSS (uninitialized data).
388 The "pathname" shows the name associated file for this mapping. If the mapping
389 is not associated with a file:
391 [heap] = the heap of the program
392 [stack] = the stack of the main process
393 [vdso] = the "virtual dynamic shared object",
394 the kernel system call handler
396 or if empty, the mapping is anonymous.
398 The /proc/PID/smaps is an extension based on maps, showing the memory
399 consumption for each of the process's mappings. For each of mappings there
400 is a series of lines such as the following:
402 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
415 Private_Hugetlb: 0 kB
421 VmFlags: rd ex mr mw me dw
423 the first of these lines shows the same information as is displayed for the
424 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
425 (size), the amount of the mapping that is currently resident in RAM (RSS), the
426 process' proportional share of this mapping (PSS), the number of clean and
427 dirty private pages in the mapping.
429 The "proportional set size" (PSS) of a process is the count of pages it has
430 in memory, where each page is divided by the number of processes sharing it.
431 So if a process has 1000 pages all to itself, and 1000 shared with one other
432 process, its PSS will be 1500.
433 Note that even a page which is part of a MAP_SHARED mapping, but has only
434 a single pte mapped, i.e. is currently used by only one process, is accounted
435 as private and not as shared.
436 "Referenced" indicates the amount of memory currently marked as referenced or
438 "Anonymous" shows the amount of memory that does not belong to any file. Even
439 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
440 and a page is modified, the file page is replaced by a private anonymous copy.
441 "AnonHugePages" shows the ammount of memory backed by transparent hugepage.
442 "ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by
444 "Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by
445 hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
446 reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
447 "Swap" shows how much would-be-anonymous memory is also used, but out on swap.
448 For shmem mappings, "Swap" includes also the size of the mapped (and not
449 replaced by copy-on-write) part of the underlying shmem object out on swap.
450 "SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this
451 does not take into account swapped out page of underlying shmem objects.
452 "Locked" indicates whether the mapping is locked in memory or not.
454 "VmFlags" field deserves a separate description. This member represents the kernel
455 flags associated with the particular virtual memory area in two letter encoded
456 manner. The codes are the following:
465 gd - stack segment growns down
467 dw - disabled write to the mapped file
468 lo - pages are locked in memory
469 io - memory mapped I/O area
470 sr - sequential read advise provided
471 rr - random read advise provided
472 dc - do not copy area on fork
473 de - do not expand area on remapping
474 ac - area is accountable
475 nr - swap space is not reserved for the area
476 ht - area uses huge tlb pages
477 ar - architecture specific flag
478 dd - do not include area into core dump
481 hg - huge page advise flag
482 nh - no-huge page advise flag
483 mg - mergable advise flag
485 Note that there is no guarantee that every flag and associated mnemonic will
486 be present in all further kernel releases. Things get changed, the flags may
487 be vanished or the reverse -- new added.
489 This file is only present if the CONFIG_MMU kernel configuration option is
492 Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent
493 output can be achieved only in the single read call).
494 This typically manifests when doing partial reads of these files while the
495 memory map is being modified. Despite the races, we do provide the following
498 1) The mapped addresses never go backwards, which implies no two
499 regions will ever overlap.
500 2) If there is something at a given vaddr during the entirety of the
501 life of the smaps/maps walk, there will be some output for it.
504 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
505 bits on both physical and virtual pages associated with a process, and the
506 soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
507 To clear the bits for all the pages associated with the process
508 > echo 1 > /proc/PID/clear_refs
510 To clear the bits for the anonymous pages associated with the process
511 > echo 2 > /proc/PID/clear_refs
513 To clear the bits for the file mapped pages associated with the process
514 > echo 3 > /proc/PID/clear_refs
516 To clear the soft-dirty bit
517 > echo 4 > /proc/PID/clear_refs
519 To reset the peak resident set size ("high water mark") to the process's
521 > echo 5 > /proc/PID/clear_refs
523 Any other value written to /proc/PID/clear_refs will have no effect.
525 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
526 using /proc/kpageflags and number of times a page is mapped using
527 /proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
529 The /proc/pid/numa_maps is an extension based on maps, showing the memory
530 locality and binding policy, as well as the memory usage (in pages) of
531 each mapping. The output follows a general format where mapping details get
532 summarized separated by blank spaces, one mapping per each file line:
534 address policy mapping details
536 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
537 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
538 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
539 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
540 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
541 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
542 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
543 320698b000 default file=/lib64/libc-2.12.so
544 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
545 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
546 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
547 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
548 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
549 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
550 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
551 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
554 "address" is the starting address for the mapping;
555 "policy" reports the NUMA memory policy set for the mapping (see vm/numa_memory_policy.txt);
556 "mapping details" summarizes mapping data such as mapping type, page usage counters,
557 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
558 size, in KB, that is backing the mapping up.
563 Similar to the process entries, the kernel data files give information about
564 the running kernel. The files used to obtain this information are contained in
565 /proc and are listed in Table 1-5. Not all of these will be present in your
566 system. It depends on the kernel configuration and the loaded modules, which
567 files are there, and which are missing.
569 Table 1-5: Kernel info in /proc
570 ..............................................................................
572 apm Advanced power management info
573 buddyinfo Kernel memory allocator information (see text) (2.5)
574 bus Directory containing bus specific information
575 cmdline Kernel command line
576 cpuinfo Info about the CPU
577 devices Available devices (block and character)
578 dma Used DMS channels
579 filesystems Supported filesystems
580 driver Various drivers grouped here, currently rtc (2.4)
581 execdomains Execdomains, related to security (2.4)
582 fb Frame Buffer devices (2.4)
583 fs File system parameters, currently nfs/exports (2.4)
584 ide Directory containing info about the IDE subsystem
585 interrupts Interrupt usage
586 iomem Memory map (2.4)
587 ioports I/O port usage
588 irq Masks for irq to cpu affinity (2.4)(smp?)
589 isapnp ISA PnP (Plug&Play) Info (2.4)
590 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
592 ksyms Kernel symbol table
593 loadavg Load average of last 1, 5 & 15 minutes
597 modules List of loaded modules
598 mounts Mounted filesystems
599 net Networking info (see text)
600 pagetypeinfo Additional page allocator information (see text) (2.5)
601 partitions Table of partitions known to the system
602 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
603 decoupled by lspci (2.4)
605 scsi SCSI info (see text)
606 slabinfo Slab pool info
607 softirqs softirq usage
608 stat Overall statistics
609 swaps Swap space utilization
611 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
612 tty Info of tty drivers
613 uptime Wall clock since boot, combined idle time of all cpus
614 version Kernel version
615 video bttv info of video resources (2.4)
616 vmallocinfo Show vmalloced areas
617 ..............................................................................
619 You can, for example, check which interrupts are currently in use and what
620 they are used for by looking in the file /proc/interrupts:
622 > cat /proc/interrupts
624 0: 8728810 XT-PIC timer
625 1: 895 XT-PIC keyboard
627 3: 531695 XT-PIC aha152x
628 4: 2014133 XT-PIC serial
629 5: 44401 XT-PIC pcnet_cs
632 12: 182918 XT-PIC PS/2 Mouse
634 14: 1232265 XT-PIC ide0
638 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
639 output of a SMP machine):
641 > cat /proc/interrupts
644 0: 1243498 1214548 IO-APIC-edge timer
645 1: 8949 8958 IO-APIC-edge keyboard
646 2: 0 0 XT-PIC cascade
647 5: 11286 10161 IO-APIC-edge soundblaster
648 8: 1 0 IO-APIC-edge rtc
649 9: 27422 27407 IO-APIC-edge 3c503
650 12: 113645 113873 IO-APIC-edge PS/2 Mouse
652 14: 22491 24012 IO-APIC-edge ide0
653 15: 2183 2415 IO-APIC-edge ide1
654 17: 30564 30414 IO-APIC-level eth0
655 18: 177 164 IO-APIC-level bttv
660 NMI is incremented in this case because every timer interrupt generates a NMI
661 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
663 LOC is the local interrupt counter of the internal APIC of every CPU.
665 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
666 connects the CPUs in a SMP system. This means that an error has been detected,
667 the IO-APIC automatically retry the transmission, so it should not be a big
668 problem, but you should read the SMP-FAQ.
670 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
671 /proc/interrupts to display every IRQ vector in use by the system, not
672 just those considered 'most important'. The new vectors are:
674 THR -- interrupt raised when a machine check threshold counter
675 (typically counting ECC corrected errors of memory or cache) exceeds
676 a configurable threshold. Only available on some systems.
678 TRM -- a thermal event interrupt occurs when a temperature threshold
679 has been exceeded for the CPU. This interrupt may also be generated
680 when the temperature drops back to normal.
682 SPU -- a spurious interrupt is some interrupt that was raised then lowered
683 by some IO device before it could be fully processed by the APIC. Hence
684 the APIC sees the interrupt but does not know what device it came from.
685 For this case the APIC will generate the interrupt with a IRQ vector
686 of 0xff. This might also be generated by chipset bugs.
688 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
689 sent from one CPU to another per the needs of the OS. Typically,
690 their statistics are used by kernel developers and interested users to
691 determine the occurrence of interrupts of the given type.
693 The above IRQ vectors are displayed only when relevant. For example,
694 the threshold vector does not exist on x86_64 platforms. Others are
695 suppressed when the system is a uniprocessor. As of this writing, only
696 i386 and x86_64 platforms support the new IRQ vector displays.
698 Of some interest is the introduction of the /proc/irq directory to 2.4.
699 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
700 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
701 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
706 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
707 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
711 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
712 IRQ, you can set it by doing:
714 > echo 1 > /proc/irq/10/smp_affinity
716 This means that only the first CPU will handle the IRQ, but you can also echo
717 5 which means that only the first and third CPU can handle the IRQ.
719 The contents of each smp_affinity file is the same by default:
721 > cat /proc/irq/0/smp_affinity
724 There is an alternate interface, smp_affinity_list which allows specifying
725 a cpu range instead of a bitmask:
727 > cat /proc/irq/0/smp_affinity_list
730 The default_smp_affinity mask applies to all non-active IRQs, which are the
731 IRQs which have not yet been allocated/activated, and hence which lack a
732 /proc/irq/[0-9]* directory.
734 The node file on an SMP system shows the node to which the device using the IRQ
735 reports itself as being attached. This hardware locality information does not
736 include information about any possible driver locality preference.
738 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
739 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
741 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
742 between all the CPUs which are allowed to handle it. As usual the kernel has
743 more info than you and does a better job than you, so the defaults are the
744 best choice for almost everyone. [Note this applies only to those IO-APIC's
745 that support "Round Robin" interrupt distribution.]
747 There are three more important subdirectories in /proc: net, scsi, and sys.
748 The general rule is that the contents, or even the existence of these
749 directories, depend on your kernel configuration. If SCSI is not enabled, the
750 directory scsi may not exist. The same is true with the net, which is there
751 only when networking support is present in the running kernel.
753 The slabinfo file gives information about memory usage at the slab level.
754 Linux uses slab pools for memory management above page level in version 2.2.
755 Commonly used objects have their own slab pool (such as network buffers,
756 directory cache, and so on).
758 ..............................................................................
760 > cat /proc/buddyinfo
762 Node 0, zone DMA 0 4 5 4 4 3 ...
763 Node 0, zone Normal 1 0 0 1 101 8 ...
764 Node 0, zone HighMem 2 0 0 1 1 0 ...
766 External fragmentation is a problem under some workloads, and buddyinfo is a
767 useful tool for helping diagnose these problems. Buddyinfo will give you a
768 clue as to how big an area you can safely allocate, or why a previous
771 Each column represents the number of pages of a certain order which are
772 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
773 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
774 available in ZONE_NORMAL, etc...
776 More information relevant to external fragmentation can be found in
779 > cat /proc/pagetypeinfo
783 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
784 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
785 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
786 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
787 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
788 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
789 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
790 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
791 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
792 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
793 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
795 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
796 Node 0, zone DMA 2 0 5 1 0
797 Node 0, zone DMA32 41 6 967 2 0
799 Fragmentation avoidance in the kernel works by grouping pages of different
800 migrate types into the same contiguous regions of memory called page blocks.
801 A page block is typically the size of the default hugepage size e.g. 2MB on
802 X86-64. By keeping pages grouped based on their ability to move, the kernel
803 can reclaim pages within a page block to satisfy a high-order allocation.
805 The pagetypinfo begins with information on the size of a page block. It
806 then gives the same type of information as buddyinfo except broken down
807 by migrate-type and finishes with details on how many page blocks of each
810 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
811 from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can
812 make an estimate of the likely number of huge pages that can be allocated
813 at a given point in time. All the "Movable" blocks should be allocatable
814 unless memory has been mlock()'d. Some of the Reclaimable blocks should
815 also be allocatable although a lot of filesystem metadata may have to be
816 reclaimed to achieve this.
818 ..............................................................................
822 Provides information about distribution and utilization of memory. This
823 varies by architecture and compile options. The following is from a
824 16GB PIII, which has highmem enabled. You may not have all of these fields.
828 MemTotal: 16344972 kB
830 MemAvailable: 14836172 kB
836 HighTotal: 15597528 kB
837 HighFree: 13629632 kB
848 SReclaimable: 159856 kB
849 SUnreclaim: 124508 kB
854 CommitLimit: 7669796 kB
855 Committed_AS: 100056 kB
856 VmallocTotal: 112216 kB
858 VmallocChunk: 111088 kB
859 AnonHugePages: 49152 kB
864 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
865 bits and the kernel binary code)
866 MemFree: The sum of LowFree+HighFree
867 MemAvailable: An estimate of how much memory is available for starting new
868 applications, without swapping. Calculated from MemFree,
869 SReclaimable, the size of the file LRU lists, and the low
870 watermarks in each zone.
871 The estimate takes into account that the system needs some
872 page cache to function well, and that not all reclaimable
873 slab will be reclaimable, due to items being in use. The
874 impact of those factors will vary from system to system.
875 Buffers: Relatively temporary storage for raw disk blocks
876 shouldn't get tremendously large (20MB or so)
877 Cached: in-memory cache for files read from the disk (the
878 pagecache). Doesn't include SwapCached
879 SwapCached: Memory that once was swapped out, is swapped back in but
880 still also is in the swapfile (if memory is needed it
881 doesn't need to be swapped out AGAIN because it is already
882 in the swapfile. This saves I/O)
883 Active: Memory that has been used more recently and usually not
884 reclaimed unless absolutely necessary.
885 Inactive: Memory which has been less recently used. It is more
886 eligible to be reclaimed for other purposes
888 HighFree: Highmem is all memory above ~860MB of physical memory
889 Highmem areas are for use by userspace programs, or
890 for the pagecache. The kernel must use tricks to access
891 this memory, making it slower to access than lowmem.
893 LowFree: Lowmem is memory which can be used for everything that
894 highmem can be used for, but it is also available for the
895 kernel's use for its own data structures. Among many
896 other things, it is where everything from the Slab is
897 allocated. Bad things happen when you're out of lowmem.
898 SwapTotal: total amount of swap space available
899 SwapFree: Memory which has been evicted from RAM, and is temporarily
901 Dirty: Memory which is waiting to get written back to the disk
902 Writeback: Memory which is actively being written back to the disk
903 AnonPages: Non-file backed pages mapped into userspace page tables
904 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
905 Mapped: files which have been mmaped, such as libraries
906 Shmem: Total memory used by shared memory (shmem) and tmpfs
907 ShmemHugePages: Memory used by shared memory (shmem) and tmpfs allocated
909 ShmemPmdMapped: Shared memory mapped into userspace with huge pages
910 Slab: in-kernel data structures cache
911 SReclaimable: Part of Slab, that might be reclaimed, such as caches
912 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
913 PageTables: amount of memory dedicated to the lowest level of page
915 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
917 Bounce: Memory used for block device "bounce buffers"
918 WritebackTmp: Memory used by FUSE for temporary writeback buffers
919 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
920 this is the total amount of memory currently available to
921 be allocated on the system. This limit is only adhered to
922 if strict overcommit accounting is enabled (mode 2 in
923 'vm.overcommit_memory').
924 The CommitLimit is calculated with the following formula:
925 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
926 overcommit_ratio / 100 + [total swap pages]
927 For example, on a system with 1G of physical RAM and 7G
928 of swap with a `vm.overcommit_ratio` of 30 it would
929 yield a CommitLimit of 7.3G.
930 For more details, see the memory overcommit documentation
931 in vm/overcommit-accounting.
932 Committed_AS: The amount of memory presently allocated on the system.
933 The committed memory is a sum of all of the memory which
934 has been allocated by processes, even if it has not been
935 "used" by them as of yet. A process which malloc()'s 1G
936 of memory, but only touches 300M of it will show up as
937 using 1G. This 1G is memory which has been "committed" to
938 by the VM and can be used at any time by the allocating
939 application. With strict overcommit enabled on the system
940 (mode 2 in 'vm.overcommit_memory'),allocations which would
941 exceed the CommitLimit (detailed above) will not be permitted.
942 This is useful if one needs to guarantee that processes will
943 not fail due to lack of memory once that memory has been
944 successfully allocated.
945 VmallocTotal: total size of vmalloc memory area
946 VmallocUsed: amount of vmalloc area which is used
947 VmallocChunk: largest contiguous block of vmalloc area which is free
949 ..............................................................................
953 Provides information about vmalloced/vmaped areas. One line per area,
954 containing the virtual address range of the area, size in bytes,
955 caller information of the creator, and optional information depending
956 on the kind of area :
958 pages=nr number of pages
959 phys=addr if a physical address was specified
960 ioremap I/O mapping (ioremap() and friends)
961 vmalloc vmalloc() area
964 vpages buffer for pages pointers was vmalloced (huge area)
965 N<node>=nr (Only on NUMA kernels)
966 Number of pages allocated on memory node <node>
968 > cat /proc/vmallocinfo
969 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
970 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
971 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
972 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
973 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
974 phys=7fee8000 ioremap
975 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
976 phys=7fee7000 ioremap
977 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
978 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
979 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
980 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
982 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
983 /0x130 [x_tables] pages=4 vmalloc N0=4
984 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
985 pages=14 vmalloc N2=14
986 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
988 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
990 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
991 pages=10 vmalloc N0=10
993 ..............................................................................
997 Provides counts of softirq handlers serviced since boot time, for each cpu.
1002 TIMER: 27166 27120 27097 27034
1007 SCHED: 27035 26983 26971 26746
1009 RCU: 1678 1769 2178 2250
1012 1.3 IDE devices in /proc/ide
1013 ----------------------------
1015 The subdirectory /proc/ide contains information about all IDE devices of which
1016 the kernel is aware. There is one subdirectory for each IDE controller, the
1017 file drivers and a link for each IDE device, pointing to the device directory
1018 in the controller specific subtree.
1020 The file drivers contains general information about the drivers used for the
1023 > cat /proc/ide/drivers
1024 ide-cdrom version 4.53
1025 ide-disk version 1.08
1027 More detailed information can be found in the controller specific
1028 subdirectories. These are named ide0, ide1 and so on. Each of these
1029 directories contains the files shown in table 1-6.
1032 Table 1-6: IDE controller info in /proc/ide/ide?
1033 ..............................................................................
1035 channel IDE channel (0 or 1)
1036 config Configuration (only for PCI/IDE bridge)
1038 model Type/Chipset of IDE controller
1039 ..............................................................................
1041 Each device connected to a controller has a separate subdirectory in the
1042 controllers directory. The files listed in table 1-7 are contained in these
1046 Table 1-7: IDE device information
1047 ..............................................................................
1050 capacity Capacity of the medium (in 512Byte blocks)
1051 driver driver and version
1052 geometry physical and logical geometry
1053 identify device identify block
1055 model device identifier
1056 settings device setup
1057 smart_thresholds IDE disk management thresholds
1058 smart_values IDE disk management values
1059 ..............................................................................
1061 The most interesting file is settings. This file contains a nice overview of
1062 the drive parameters:
1064 # cat /proc/ide/ide0/hda/settings
1065 name value min max mode
1066 ---- ----- --- --- ----
1067 bios_cyl 526 0 65535 rw
1068 bios_head 255 0 255 rw
1069 bios_sect 63 0 63 rw
1070 breada_readahead 4 0 127 rw
1072 file_readahead 72 0 2097151 rw
1074 keepsettings 0 0 1 rw
1075 max_kb_per_request 122 1 127 rw
1079 pio_mode write-only 0 255 w
1085 1.4 Networking info in /proc/net
1086 --------------------------------
1088 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1089 additional values you get for IP version 6 if you configure the kernel to
1090 support this. Table 1-9 lists the files and their meaning.
1093 Table 1-8: IPv6 info in /proc/net
1094 ..............................................................................
1096 udp6 UDP sockets (IPv6)
1097 tcp6 TCP sockets (IPv6)
1098 raw6 Raw device statistics (IPv6)
1099 igmp6 IP multicast addresses, which this host joined (IPv6)
1100 if_inet6 List of IPv6 interface addresses
1101 ipv6_route Kernel routing table for IPv6
1102 rt6_stats Global IPv6 routing tables statistics
1103 sockstat6 Socket statistics (IPv6)
1104 snmp6 Snmp data (IPv6)
1105 ..............................................................................
1108 Table 1-9: Network info in /proc/net
1109 ..............................................................................
1111 arp Kernel ARP table
1112 dev network devices with statistics
1113 dev_mcast the Layer2 multicast groups a device is listening too
1114 (interface index, label, number of references, number of bound
1116 dev_stat network device status
1117 ip_fwchains Firewall chain linkage
1118 ip_fwnames Firewall chain names
1119 ip_masq Directory containing the masquerading tables
1120 ip_masquerade Major masquerading table
1121 netstat Network statistics
1122 raw raw device statistics
1123 route Kernel routing table
1124 rpc Directory containing rpc info
1125 rt_cache Routing cache
1127 sockstat Socket statistics
1130 unix UNIX domain sockets
1131 wireless Wireless interface data (Wavelan etc)
1132 igmp IP multicast addresses, which this host joined
1133 psched Global packet scheduler parameters.
1134 netlink List of PF_NETLINK sockets
1135 ip_mr_vifs List of multicast virtual interfaces
1136 ip_mr_cache List of multicast routing cache
1137 ..............................................................................
1139 You can use this information to see which network devices are available in
1140 your system and how much traffic was routed over those devices:
1143 Inter-|Receive |[...
1144 face |bytes packets errs drop fifo frame compressed multicast|[...
1145 lo: 908188 5596 0 0 0 0 0 0 [...
1146 ppp0:15475140 20721 410 0 0 410 0 0 [...
1147 eth0: 614530 7085 0 0 0 0 0 1 [...
1150 ...] bytes packets errs drop fifo colls carrier compressed
1151 ...] 908188 5596 0 0 0 0 0 0
1152 ...] 1375103 17405 0 0 0 0 0 0
1153 ...] 1703981 5535 0 0 0 3 0 0
1155 In addition, each Channel Bond interface has its own directory. For
1156 example, the bond0 device will have a directory called /proc/net/bond0/.
1157 It will contain information that is specific to that bond, such as the
1158 current slaves of the bond, the link status of the slaves, and how
1159 many times the slaves link has failed.
1164 If you have a SCSI host adapter in your system, you'll find a subdirectory
1165 named after the driver for this adapter in /proc/scsi. You'll also see a list
1166 of all recognized SCSI devices in /proc/scsi:
1168 >cat /proc/scsi/scsi
1170 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1171 Vendor: IBM Model: DGHS09U Rev: 03E0
1172 Type: Direct-Access ANSI SCSI revision: 03
1173 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1174 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1175 Type: CD-ROM ANSI SCSI revision: 02
1178 The directory named after the driver has one file for each adapter found in
1179 the system. These files contain information about the controller, including
1180 the used IRQ and the IO address range. The amount of information shown is
1181 dependent on the adapter you use. The example shows the output for an Adaptec
1182 AHA-2940 SCSI adapter:
1184 > cat /proc/scsi/aic7xxx/0
1186 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1188 TCQ Enabled By Default : Disabled
1189 AIC7XXX_PROC_STATS : Disabled
1190 AIC7XXX_RESET_DELAY : 5
1191 Adapter Configuration:
1192 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1193 Ultra Wide Controller
1194 PCI MMAPed I/O Base: 0xeb001000
1195 Adapter SEEPROM Config: SEEPROM found and used.
1196 Adaptec SCSI BIOS: Enabled
1198 SCBs: Active 0, Max Active 2,
1199 Allocated 15, HW 16, Page 255
1201 BIOS Control Word: 0x18b6
1202 Adapter Control Word: 0x005b
1203 Extended Translation: Enabled
1204 Disconnect Enable Flags: 0xffff
1205 Ultra Enable Flags: 0x0001
1206 Tag Queue Enable Flags: 0x0000
1207 Ordered Queue Tag Flags: 0x0000
1208 Default Tag Queue Depth: 8
1209 Tagged Queue By Device array for aic7xxx host instance 0:
1210 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1211 Actual queue depth per device for aic7xxx host instance 0:
1212 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1215 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1216 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1217 Total transfers 160151 (74577 reads and 85574 writes)
1219 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1220 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1221 Total transfers 0 (0 reads and 0 writes)
1224 1.6 Parallel port info in /proc/parport
1225 ---------------------------------------
1227 The directory /proc/parport contains information about the parallel ports of
1228 your system. It has one subdirectory for each port, named after the port
1231 These directories contain the four files shown in Table 1-10.
1234 Table 1-10: Files in /proc/parport
1235 ..............................................................................
1237 autoprobe Any IEEE-1284 device ID information that has been acquired.
1238 devices list of the device drivers using that port. A + will appear by the
1239 name of the device currently using the port (it might not appear
1241 hardware Parallel port's base address, IRQ line and DMA channel.
1242 irq IRQ that parport is using for that port. This is in a separate
1243 file to allow you to alter it by writing a new value in (IRQ
1245 ..............................................................................
1247 1.7 TTY info in /proc/tty
1248 -------------------------
1250 Information about the available and actually used tty's can be found in the
1251 directory /proc/tty.You'll find entries for drivers and line disciplines in
1252 this directory, as shown in Table 1-11.
1255 Table 1-11: Files in /proc/tty
1256 ..............................................................................
1258 drivers list of drivers and their usage
1259 ldiscs registered line disciplines
1260 driver/serial usage statistic and status of single tty lines
1261 ..............................................................................
1263 To see which tty's are currently in use, you can simply look into the file
1266 > cat /proc/tty/drivers
1267 pty_slave /dev/pts 136 0-255 pty:slave
1268 pty_master /dev/ptm 128 0-255 pty:master
1269 pty_slave /dev/ttyp 3 0-255 pty:slave
1270 pty_master /dev/pty 2 0-255 pty:master
1271 serial /dev/cua 5 64-67 serial:callout
1272 serial /dev/ttyS 4 64-67 serial
1273 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1274 /dev/ptmx /dev/ptmx 5 2 system
1275 /dev/console /dev/console 5 1 system:console
1276 /dev/tty /dev/tty 5 0 system:/dev/tty
1277 unknown /dev/tty 4 1-63 console
1280 1.8 Miscellaneous kernel statistics in /proc/stat
1281 -------------------------------------------------
1283 Various pieces of information about kernel activity are available in the
1284 /proc/stat file. All of the numbers reported in this file are aggregates
1285 since the system first booted. For a quick look, simply cat the file:
1288 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1289 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1290 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1291 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1297 softirq 183433 0 21755 12 39 1137 231 21459 2263
1299 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1300 lines. These numbers identify the amount of time the CPU has spent performing
1301 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1302 second). The meanings of the columns are as follows, from left to right:
1304 - user: normal processes executing in user mode
1305 - nice: niced processes executing in user mode
1306 - system: processes executing in kernel mode
1307 - idle: twiddling thumbs
1308 - iowait: waiting for I/O to complete
1309 - irq: servicing interrupts
1310 - softirq: servicing softirqs
1311 - steal: involuntary wait
1312 - guest: running a normal guest
1313 - guest_nice: running a niced guest
1315 The "intr" line gives counts of interrupts serviced since boot time, for each
1316 of the possible system interrupts. The first column is the total of all
1317 interrupts serviced including unnumbered architecture specific interrupts;
1318 each subsequent column is the total for that particular numbered interrupt.
1319 Unnumbered interrupts are not shown, only summed into the total.
1321 The "ctxt" line gives the total number of context switches across all CPUs.
1323 The "btime" line gives the time at which the system booted, in seconds since
1326 The "processes" line gives the number of processes and threads created, which
1327 includes (but is not limited to) those created by calls to the fork() and
1328 clone() system calls.
1330 The "procs_running" line gives the total number of threads that are
1331 running or ready to run (i.e., the total number of runnable threads).
1333 The "procs_blocked" line gives the number of processes currently blocked,
1334 waiting for I/O to complete.
1336 The "softirq" line gives counts of softirqs serviced since boot time, for each
1337 of the possible system softirqs. The first column is the total of all
1338 softirqs serviced; each subsequent column is the total for that particular
1342 1.9 Ext4 file system parameters
1343 -------------------------------
1345 Information about mounted ext4 file systems can be found in
1346 /proc/fs/ext4. Each mounted filesystem will have a directory in
1347 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1348 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1349 in Table 1-12, below.
1351 Table 1-12: Files in /proc/fs/ext4/<devname>
1352 ..............................................................................
1354 mb_groups details of multiblock allocator buddy cache of free blocks
1355 ..............................................................................
1359 Shows registered system console lines.
1361 To see which character device lines are currently used for the system console
1362 /dev/console, you may simply look into the file /proc/consoles:
1364 > cat /proc/consoles
1370 device name of the device
1371 operations R = can do read operations
1372 W = can do write operations
1374 flags E = it is enabled
1375 C = it is preferred console
1376 B = it is primary boot console
1377 p = it is used for printk buffer
1378 b = it is not a TTY but a Braille device
1379 a = it is safe to use when cpu is offline
1380 major:minor major and minor number of the device separated by a colon
1382 ------------------------------------------------------------------------------
1384 ------------------------------------------------------------------------------
1385 The /proc file system serves information about the running system. It not only
1386 allows access to process data but also allows you to request the kernel status
1387 by reading files in the hierarchy.
1389 The directory structure of /proc reflects the types of information and makes
1390 it easy, if not obvious, where to look for specific data.
1391 ------------------------------------------------------------------------------
1393 ------------------------------------------------------------------------------
1394 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1395 ------------------------------------------------------------------------------
1397 ------------------------------------------------------------------------------
1399 ------------------------------------------------------------------------------
1400 * Modifying kernel parameters by writing into files found in /proc/sys
1401 * Exploring the files which modify certain parameters
1402 * Review of the /proc/sys file tree
1403 ------------------------------------------------------------------------------
1406 A very interesting part of /proc is the directory /proc/sys. This is not only
1407 a source of information, it also allows you to change parameters within the
1408 kernel. Be very careful when attempting this. You can optimize your system,
1409 but you can also cause it to crash. Never alter kernel parameters on a
1410 production system. Set up a development machine and test to make sure that
1411 everything works the way you want it to. You may have no alternative but to
1412 reboot the machine once an error has been made.
1414 To change a value, simply echo the new value into the file. An example is
1415 given below in the section on the file system data. You need to be root to do
1416 this. You can create your own boot script to perform this every time your
1419 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1420 general things in the operation of the Linux kernel. Since some of the files
1421 can inadvertently disrupt your system, it is advisable to read both
1422 documentation and source before actually making adjustments. In any case, be
1423 very careful when writing to any of these files. The entries in /proc may
1424 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1425 review the kernel documentation in the directory /usr/src/linux/Documentation.
1426 This chapter is heavily based on the documentation included in the pre 2.2
1427 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1429 Please see: Documentation/sysctl/ directory for descriptions of these
1432 ------------------------------------------------------------------------------
1434 ------------------------------------------------------------------------------
1435 Certain aspects of kernel behavior can be modified at runtime, without the
1436 need to recompile the kernel, or even to reboot the system. The files in the
1437 /proc/sys tree can not only be read, but also modified. You can use the echo
1438 command to write value into these files, thereby changing the default settings
1440 ------------------------------------------------------------------------------
1442 ------------------------------------------------------------------------------
1443 CHAPTER 3: PER-PROCESS PARAMETERS
1444 ------------------------------------------------------------------------------
1446 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1447 --------------------------------------------------------------------------------
1449 These file can be used to adjust the badness heuristic used to select which
1450 process gets killed in out of memory conditions.
1452 The badness heuristic assigns a value to each candidate task ranging from 0
1453 (never kill) to 1000 (always kill) to determine which process is targeted. The
1454 units are roughly a proportion along that range of allowed memory the process
1455 may allocate from based on an estimation of its current memory and swap use.
1456 For example, if a task is using all allowed memory, its badness score will be
1457 1000. If it is using half of its allowed memory, its score will be 500.
1459 There is an additional factor included in the badness score: the current memory
1460 and swap usage is discounted by 3% for root processes.
1462 The amount of "allowed" memory depends on the context in which the oom killer
1463 was called. If it is due to the memory assigned to the allocating task's cpuset
1464 being exhausted, the allowed memory represents the set of mems assigned to that
1465 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1466 memory represents the set of mempolicy nodes. If it is due to a memory
1467 limit (or swap limit) being reached, the allowed memory is that configured
1468 limit. Finally, if it is due to the entire system being out of memory, the
1469 allowed memory represents all allocatable resources.
1471 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1472 is used to determine which task to kill. Acceptable values range from -1000
1473 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1474 polarize the preference for oom killing either by always preferring a certain
1475 task or completely disabling it. The lowest possible value, -1000, is
1476 equivalent to disabling oom killing entirely for that task since it will always
1477 report a badness score of 0.
1479 Consequently, it is very simple for userspace to define the amount of memory to
1480 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1481 example, is roughly equivalent to allowing the remainder of tasks sharing the
1482 same system, cpuset, mempolicy, or memory controller resources to use at least
1483 50% more memory. A value of -500, on the other hand, would be roughly
1484 equivalent to discounting 50% of the task's allowed memory from being considered
1485 as scoring against the task.
1487 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1488 be used to tune the badness score. Its acceptable values range from -16
1489 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1490 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1491 scaled linearly with /proc/<pid>/oom_score_adj.
1493 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1494 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1495 requires CAP_SYS_RESOURCE.
1497 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1498 generation children with separate address spaces instead, if possible. This
1499 avoids servers and important system daemons from being killed and loses the
1500 minimal amount of work.
1503 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1504 -------------------------------------------------------------
1506 This file can be used to check the current score used by the oom-killer is for
1507 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1508 process should be killed in an out-of-memory situation.
1511 3.3 /proc/<pid>/io - Display the IO accounting fields
1512 -------------------------------------------------------
1514 This file contains IO statistics for each running process
1519 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1522 test:/tmp # cat /proc/3828/io
1528 write_bytes: 323932160
1529 cancelled_write_bytes: 0
1538 I/O counter: chars read
1539 The number of bytes which this task has caused to be read from storage. This
1540 is simply the sum of bytes which this process passed to read() and pread().
1541 It includes things like tty IO and it is unaffected by whether or not actual
1542 physical disk IO was required (the read might have been satisfied from
1549 I/O counter: chars written
1550 The number of bytes which this task has caused, or shall cause to be written
1551 to disk. Similar caveats apply here as with rchar.
1557 I/O counter: read syscalls
1558 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1565 I/O counter: write syscalls
1566 Attempt to count the number of write I/O operations, i.e. syscalls like
1567 write() and pwrite().
1573 I/O counter: bytes read
1574 Attempt to count the number of bytes which this process really did cause to
1575 be fetched from the storage layer. Done at the submit_bio() level, so it is
1576 accurate for block-backed filesystems. <please add status regarding NFS and
1577 CIFS at a later time>
1583 I/O counter: bytes written
1584 Attempt to count the number of bytes which this process caused to be sent to
1585 the storage layer. This is done at page-dirtying time.
1588 cancelled_write_bytes
1589 ---------------------
1591 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1592 then deletes the file, it will in fact perform no writeout. But it will have
1593 been accounted as having caused 1MB of write.
1594 In other words: The number of bytes which this process caused to not happen,
1595 by truncating pagecache. A task can cause "negative" IO too. If this task
1596 truncates some dirty pagecache, some IO which another task has been accounted
1597 for (in its write_bytes) will not be happening. We _could_ just subtract that
1598 from the truncating task's write_bytes, but there is information loss in doing
1605 At its current implementation state, this is a bit racy on 32-bit machines: if
1606 process A reads process B's /proc/pid/io while process B is updating one of
1607 those 64-bit counters, process A could see an intermediate result.
1610 More information about this can be found within the taskstats documentation in
1611 Documentation/accounting.
1613 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1614 ---------------------------------------------------------------
1615 When a process is dumped, all anonymous memory is written to a core file as
1616 long as the size of the core file isn't limited. But sometimes we don't want
1617 to dump some memory segments, for example, huge shared memory or DAX.
1618 Conversely, sometimes we want to save file-backed memory segments into a core
1619 file, not only the individual files.
1621 /proc/<pid>/coredump_filter allows you to customize which memory segments
1622 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1623 of memory types. If a bit of the bitmask is set, memory segments of the
1624 corresponding memory type are dumped, otherwise they are not dumped.
1626 The following 9 memory types are supported:
1627 - (bit 0) anonymous private memory
1628 - (bit 1) anonymous shared memory
1629 - (bit 2) file-backed private memory
1630 - (bit 3) file-backed shared memory
1631 - (bit 4) ELF header pages in file-backed private memory areas (it is
1632 effective only if the bit 2 is cleared)
1633 - (bit 5) hugetlb private memory
1634 - (bit 6) hugetlb shared memory
1635 - (bit 7) DAX private memory
1636 - (bit 8) DAX shared memory
1638 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1639 are always dumped regardless of the bitmask status.
1641 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is
1642 only affected by bit 5-6, and DAX is only affected by bits 7-8.
1644 The default value of coredump_filter is 0x33; this means all anonymous memory
1645 segments, ELF header pages and hugetlb private memory are dumped.
1647 If you don't want to dump all shared memory segments attached to pid 1234,
1648 write 0x31 to the process's proc file.
1650 $ echo 0x31 > /proc/1234/coredump_filter
1652 When a new process is created, the process inherits the bitmask status from its
1653 parent. It is useful to set up coredump_filter before the program runs.
1656 $ echo 0x7 > /proc/self/coredump_filter
1659 3.5 /proc/<pid>/mountinfo - Information about mounts
1660 --------------------------------------------------------
1662 This file contains lines of the form:
1664 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1665 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1667 (1) mount ID: unique identifier of the mount (may be reused after umount)
1668 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1669 (3) major:minor: value of st_dev for files on filesystem
1670 (4) root: root of the mount within the filesystem
1671 (5) mount point: mount point relative to the process's root
1672 (6) mount options: per mount options
1673 (7) optional fields: zero or more fields of the form "tag[:value]"
1674 (8) separator: marks the end of the optional fields
1675 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1676 (10) mount source: filesystem specific information or "none"
1677 (11) super options: per super block options
1679 Parsers should ignore all unrecognised optional fields. Currently the
1680 possible optional fields are:
1682 shared:X mount is shared in peer group X
1683 master:X mount is slave to peer group X
1684 propagate_from:X mount is slave and receives propagation from peer group X (*)
1685 unbindable mount is unbindable
1687 (*) X is the closest dominant peer group under the process's root. If
1688 X is the immediate master of the mount, or if there's no dominant peer
1689 group under the same root, then only the "master:X" field is present
1690 and not the "propagate_from:X" field.
1692 For more information on mount propagation see:
1694 Documentation/filesystems/sharedsubtree.txt
1697 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1698 --------------------------------------------------------
1699 These files provide a method to access a tasks comm value. It also allows for
1700 a task to set its own or one of its thread siblings comm value. The comm value
1701 is limited in size compared to the cmdline value, so writing anything longer
1702 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1706 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1707 -------------------------------------------------------------------------
1708 This file provides a fast way to retrieve first level children pids
1709 of a task pointed by <pid>/<tid> pair. The format is a space separated
1712 Note the "first level" here -- if a child has own children they will
1713 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1714 to obtain the descendants.
1716 Since this interface is intended to be fast and cheap it doesn't
1717 guarantee to provide precise results and some children might be
1718 skipped, especially if they've exited right after we printed their
1719 pids, so one need to either stop or freeze processes being inspected
1720 if precise results are needed.
1723 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1724 ---------------------------------------------------------------
1725 This file provides information associated with an opened file. The regular
1726 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1727 represents the current offset of the opened file in decimal form [see lseek(2)
1728 for details], 'flags' denotes the octal O_xxx mask the file has been
1729 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1730 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1739 All locks associated with a file descriptor are shown in its fdinfo too.
1741 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1743 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1744 pair provide additional information particular to the objects they represent.
1753 where 'eventfd-count' is hex value of a counter.
1760 sigmask: 0000000000000200
1762 where 'sigmask' is hex value of the signal mask associated
1770 tfd: 5 events: 1d data: ffffffffffffffff
1772 where 'tfd' is a target file descriptor number in decimal form,
1773 'events' is events mask being watched and the 'data' is data
1774 associated with a target [see epoll(7) for more details].
1778 For inotify files the format is the following
1782 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1784 where 'wd' is a watch descriptor in decimal form, ie a target file
1785 descriptor number, 'ino' and 'sdev' are inode and device where the
1786 target file resides and the 'mask' is the mask of events, all in hex
1787 form [see inotify(7) for more details].
1789 If the kernel was built with exportfs support, the path to the target
1790 file is encoded as a file handle. The file handle is provided by three
1791 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1794 If the kernel is built without exportfs support the file handle won't be
1797 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1799 For fanotify files the format is
1804 fanotify flags:10 event-flags:0
1805 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1806 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1808 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1809 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1810 flags associated with mark which are tracked separately from events
1811 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1812 mask and 'ignored_mask' is the mask of events which are to be ignored.
1813 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1814 does provide information about flags and mask used in fanotify_mark
1815 call [see fsnotify manpage for details].
1817 While the first three lines are mandatory and always printed, the rest is
1818 optional and may be omitted if no marks created yet.
1829 it_value: (0, 49406829)
1832 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1833 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1834 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1835 details]. 'it_value' is remaining time until the timer exiration.
1836 'it_interval' is the interval for the timer. Note the timer might be set up
1837 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1838 still exhibits timer's remaining time.
1840 3.9 /proc/<pid>/map_files - Information about memory mapped files
1841 ---------------------------------------------------------------------
1842 This directory contains symbolic links which represent memory mapped files
1843 the process is maintaining. Example output:
1845 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1846 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1847 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1849 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1850 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1852 The name of a link represents the virtual memory bounds of a mapping, i.e.
1853 vm_area_struct::vm_start-vm_area_struct::vm_end.
1855 The main purpose of the map_files is to retrieve a set of memory mapped
1856 files in a fast way instead of parsing /proc/<pid>/maps or
1857 /proc/<pid>/smaps, both of which contain many more records. At the same
1858 time one can open(2) mappings from the listings of two processes and
1859 comparing their inode numbers to figure out which anonymous memory areas
1860 are actually shared.
1862 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
1863 ---------------------------------------------------------
1864 This file provides the value of the task's timerslack value in nanoseconds.
1865 This value specifies a amount of time that normal timers may be deferred
1866 in order to coalesce timers and avoid unnecessary wakeups.
1868 This allows a task's interactivity vs power consumption trade off to be
1871 Writing 0 to the file will set the tasks timerslack to the default value.
1873 Valid values are from 0 - ULLONG_MAX
1875 An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level
1876 permissions on the task specified to change its timerslack_ns value.
1879 ------------------------------------------------------------------------------
1881 ------------------------------------------------------------------------------
1884 ---------------------
1886 The following mount options are supported:
1888 hidepid= Set /proc/<pid>/ access mode.
1889 gid= Set the group authorized to learn processes information.
1891 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1894 hidepid=1 means users may not access any /proc/<pid>/ directories but their
1895 own. Sensitive files like cmdline, sched*, status are now protected against
1896 other users. This makes it impossible to learn whether any user runs
1897 specific program (given the program doesn't reveal itself by its behaviour).
1898 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1899 poorly written programs passing sensitive information via program arguments are
1900 now protected against local eavesdroppers.
1902 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1903 users. It doesn't mean that it hides a fact whether a process with a specific
1904 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1905 but it hides process' uid and gid, which may be learned by stat()'ing
1906 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1907 information about running processes, whether some daemon runs with elevated
1908 privileges, whether other user runs some sensitive program, whether other users
1909 run any program at all, etc.
1911 gid= defines a group authorized to learn processes information otherwise
1912 prohibited by hidepid=. If you use some daemon like identd which needs to learn
1913 information about processes information, just add identd to this group.
projects / cascardo / linux.git / blob