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
50 ------------------------------------------------------------------------------
52 ------------------------------------------------------------------------------
54 0.1 Introduction/Credits
55 ------------------------
57 This documentation is part of a soon (or so we hope) to be released book on
58 the SuSE Linux distribution. As there is no complete documentation for the
59 /proc file system and we've used many freely available sources to write these
60 chapters, it seems only fair to give the work back to the Linux community.
61 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
62 afraid it's still far from complete, but we hope it will be useful. As far as
63 we know, it is the first 'all-in-one' document about the /proc file system. It
64 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
65 SPARC, AXP, etc., features, you probably won't find what you are looking for.
66 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
67 additions and patches are welcome and will be added to this document if you
70 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
71 other people for help compiling this documentation. We'd also like to extend a
72 special thank you to Andi Kleen for documentation, which we relied on heavily
73 to create this document, as well as the additional information he provided.
74 Thanks to everybody else who contributed source or docs to the Linux kernel
75 and helped create a great piece of software... :)
77 If you have any comments, corrections or additions, please don't hesitate to
78 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
81 The latest version of this document is available online at
82 http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
84 If the above direction does not works for you, you could try the kernel
85 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
86 comandante@zaralinux.com.
91 We don't guarantee the correctness of this document, and if you come to us
92 complaining about how you screwed up your system because of incorrect
93 documentation, we won't feel responsible...
95 ------------------------------------------------------------------------------
96 CHAPTER 1: COLLECTING SYSTEM INFORMATION
97 ------------------------------------------------------------------------------
99 ------------------------------------------------------------------------------
101 ------------------------------------------------------------------------------
102 * Investigating the properties of the pseudo file system /proc and its
103 ability to provide information on the running Linux system
104 * Examining /proc's structure
105 * Uncovering various information about the kernel and the processes running
107 ------------------------------------------------------------------------------
110 The proc file system acts as an interface to internal data structures in the
111 kernel. It can be used to obtain information about the system and to change
112 certain kernel parameters at runtime (sysctl).
114 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
115 show you how you can use /proc/sys to change settings.
117 1.1 Process-Specific Subdirectories
118 -----------------------------------
120 The directory /proc contains (among other things) one subdirectory for each
121 process running on the system, which is named after the process ID (PID).
123 The link self points to the process reading the file system. Each process
124 subdirectory has the entries listed in Table 1-1.
127 Table 1-1: Process specific entries in /proc
128 ..............................................................................
130 clear_refs Clears page referenced bits shown in smaps output
131 cmdline Command line arguments
132 cpu Current and last cpu in which it was executed (2.4)(smp)
133 cwd Link to the current working directory
134 environ Values of environment variables
135 exe Link to the executable of this process
136 fd Directory, which contains all file descriptors
137 maps Memory maps to executables and library files (2.4)
138 mem Memory held by this process
139 root Link to the root directory of this process
141 statm Process memory status information
142 status Process status in human readable form
143 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
145 stack Report full stack trace, enable via CONFIG_STACKTRACE
146 smaps a extension based on maps, showing the memory consumption of
147 each mapping and flags associated with it
148 numa_maps an extension based on maps, showing the memory locality and
149 binding policy as well as mem usage (in pages) of each mapping.
150 ..............................................................................
152 For example, to get the status information of a process, all you have to do is
153 read the file /proc/PID/status:
155 >cat /proc/self/status
179 SigPnd: 0000000000000000
180 ShdPnd: 0000000000000000
181 SigBlk: 0000000000000000
182 SigIgn: 0000000000000000
183 SigCgt: 0000000000000000
184 CapInh: 00000000fffffeff
185 CapPrm: 0000000000000000
186 CapEff: 0000000000000000
187 CapBnd: ffffffffffffffff
189 voluntary_ctxt_switches: 0
190 nonvoluntary_ctxt_switches: 1
192 This shows you nearly the same information you would get if you viewed it with
193 the ps command. In fact, ps uses the proc file system to obtain its
194 information. But you get a more detailed view of the process by reading the
195 file /proc/PID/status. It fields are described in table 1-2.
197 The statm file contains more detailed information about the process
198 memory usage. Its seven fields are explained in Table 1-3. The stat file
199 contains details information about the process itself. Its fields are
200 explained in Table 1-4.
202 (for SMP CONFIG users)
203 For making accounting scalable, RSS related information are handled in an
204 asynchronous manner and the value may not be very precise. To see a precise
205 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
206 It's slow but very precise.
208 Table 1-2: Contents of the status files (as of 4.1)
209 ..............................................................................
211 Name filename of the executable
212 State state (R is running, S is sleeping, D is sleeping
213 in an uninterruptible wait, Z is zombie,
214 T is traced or stopped)
216 Ngid NUMA group ID (0 if none)
218 PPid process id of the parent process
219 TracerPid PID of process tracing this process (0 if not)
220 Uid Real, effective, saved set, and file system UIDs
221 Gid Real, effective, saved set, and file system GIDs
222 FDSize number of file descriptor slots currently allocated
223 Groups supplementary group list
224 NStgid descendant namespace thread group ID hierarchy
225 NSpid descendant namespace process ID hierarchy
226 NSpgid descendant namespace process group ID hierarchy
227 NSsid descendant namespace session ID hierarchy
228 VmPeak peak virtual memory size
229 VmSize total program size
230 VmLck locked memory size
231 VmHWM peak resident set size ("high water mark")
232 VmRSS size of memory portions
233 VmData size of data, stack, and text segments
234 VmStk size of data, stack, and text segments
235 VmExe size of text segment
236 VmLib size of shared library code
237 VmPTE size of page table entries
238 VmPMD size of second level page tables
239 VmSwap size of swap usage (the number of referred swapents)
240 Threads number of threads
241 SigQ number of signals queued/max. number for queue
242 SigPnd bitmap of pending signals for the thread
243 ShdPnd bitmap of shared pending signals for the process
244 SigBlk bitmap of blocked signals
245 SigIgn bitmap of ignored signals
246 SigCgt bitmap of caught signals
247 CapInh bitmap of inheritable capabilities
248 CapPrm bitmap of permitted capabilities
249 CapEff bitmap of effective capabilities
250 CapBnd bitmap of capabilities bounding set
251 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
252 Cpus_allowed mask of CPUs on which this process may run
253 Cpus_allowed_list Same as previous, but in "list format"
254 Mems_allowed mask of memory nodes allowed to this process
255 Mems_allowed_list Same as previous, but in "list format"
256 voluntary_ctxt_switches number of voluntary context switches
257 nonvoluntary_ctxt_switches number of non voluntary context switches
258 ..............................................................................
260 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
261 ..............................................................................
263 size total program size (pages) (same as VmSize in status)
264 resident size of memory portions (pages) (same as VmRSS in status)
265 shared number of pages that are shared (i.e. backed by a file)
266 trs number of pages that are 'code' (not including libs; broken,
267 includes data segment)
268 lrs number of pages of library (always 0 on 2.6)
269 drs number of pages of data/stack (including libs; broken,
270 includes library text)
271 dt number of dirty pages (always 0 on 2.6)
272 ..............................................................................
275 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
276 ..............................................................................
279 tcomm filename of the executable
280 state state (R is running, S is sleeping, D is sleeping in an
281 uninterruptible wait, Z is zombie, T is traced or stopped)
282 ppid process id of the parent process
283 pgrp pgrp of the process
285 tty_nr tty the process uses
286 tty_pgrp pgrp of the tty
288 min_flt number of minor faults
289 cmin_flt number of minor faults with child's
290 maj_flt number of major faults
291 cmaj_flt number of major faults with child's
292 utime user mode jiffies
293 stime kernel mode jiffies
294 cutime user mode jiffies with child's
295 cstime kernel mode jiffies with child's
296 priority priority level
298 num_threads number of threads
299 it_real_value (obsolete, always 0)
300 start_time time the process started after system boot
301 vsize virtual memory size
302 rss resident set memory size
303 rsslim current limit in bytes on the rss
304 start_code address above which program text can run
305 end_code address below which program text can run
306 start_stack address of the start of the main process stack
307 esp current value of ESP
308 eip current value of EIP
309 pending bitmap of pending signals
310 blocked bitmap of blocked signals
311 sigign bitmap of ignored signals
312 sigcatch bitmap of caught signals
313 wchan address where process went to sleep
316 exit_signal signal to send to parent thread on exit
317 task_cpu which CPU the task is scheduled on
318 rt_priority realtime priority
319 policy scheduling policy (man sched_setscheduler)
320 blkio_ticks time spent waiting for block IO
321 gtime guest time of the task in jiffies
322 cgtime guest time of the task children in jiffies
323 start_data address above which program data+bss is placed
324 end_data address below which program data+bss is placed
325 start_brk address above which program heap can be expanded with brk()
326 arg_start address above which program command line is placed
327 arg_end address below which program command line is placed
328 env_start address above which program environment is placed
329 env_end address below which program environment is placed
330 exit_code the thread's exit_code in the form reported by the waitpid system call
331 ..............................................................................
333 The /proc/PID/maps file containing the currently mapped memory regions and
334 their access permissions.
338 address perms offset dev inode pathname
340 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
341 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
342 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
343 a7cb1000-a7cb2000 ---p 00000000 00:00 0
344 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
345 a7eb2000-a7eb3000 ---p 00000000 00:00 0
346 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
347 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
348 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
349 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
350 a800b000-a800e000 rw-p 00000000 00:00 0
351 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
352 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
353 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
354 a8024000-a8027000 rw-p 00000000 00:00 0
355 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
356 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
357 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
358 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
359 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
361 where "address" is the address space in the process that it occupies, "perms"
362 is a set of permissions:
368 p = private (copy on write)
370 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
371 "inode" is the inode on that device. 0 indicates that no inode is associated
372 with the memory region, as the case would be with BSS (uninitialized data).
373 The "pathname" shows the name associated file for this mapping. If the mapping
374 is not associated with a file:
376 [heap] = the heap of the program
377 [stack] = the stack of the main process
378 [stack:1001] = the stack of the thread with tid 1001
379 [vdso] = the "virtual dynamic shared object",
380 the kernel system call handler
382 or if empty, the mapping is anonymous.
384 The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
385 of the individual tasks of a process. In this file you will see a mapping marked
386 as [stack] if that task sees it as a stack. This is a key difference from the
387 content of /proc/PID/maps, where you will see all mappings that are being used
388 as stack by all of those tasks. Hence, for the example above, the task-level
389 map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
391 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
392 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
393 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
394 a7cb1000-a7cb2000 ---p 00000000 00:00 0
395 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
396 a7eb2000-a7eb3000 ---p 00000000 00:00 0
397 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
398 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
399 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
400 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
401 a800b000-a800e000 rw-p 00000000 00:00 0
402 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
403 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
404 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
405 a8024000-a8027000 rw-p 00000000 00:00 0
406 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
407 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
408 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
409 aff35000-aff4a000 rw-p 00000000 00:00 0
410 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
412 The /proc/PID/smaps is an extension based on maps, showing the memory
413 consumption for each of the process's mappings. For each of mappings there
414 is a series of lines such as the following:
416 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
430 VmFlags: rd ex mr mw me de
432 the first of these lines shows the same information as is displayed for the
433 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
434 (size), the amount of the mapping that is currently resident in RAM (RSS), the
435 process' proportional share of this mapping (PSS), the number of clean and
436 dirty private pages in the mapping. Note that even a page which is part of a
437 MAP_SHARED mapping, but has only a single pte mapped, i.e. is currently used
438 by only one process, is accounted as private and not as shared. "Referenced"
439 indicates the amount of memory currently marked as referenced or accessed.
440 "Anonymous" shows the amount of memory that does not belong to any file. Even
441 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
442 and a page is modified, the file page is replaced by a private anonymous copy.
443 "Swap" shows how much would-be-anonymous memory is also used, but out on
446 "VmFlags" field deserves a separate description. This member represents the kernel
447 flags associated with the particular virtual memory area in two letter encoded
448 manner. The codes are the following:
457 gd - stack segment growns down
459 dw - disabled write to the mapped file
460 lo - pages are locked in memory
461 io - memory mapped I/O area
462 sr - sequential read advise provided
463 rr - random read advise provided
464 dc - do not copy area on fork
465 de - do not expand area on remapping
466 ac - area is accountable
467 nr - swap space is not reserved for the area
468 ht - area uses huge tlb pages
469 nl - non-linear mapping
470 ar - architecture specific flag
471 dd - do not include area into core dump
474 hg - huge page advise flag
475 nh - no-huge page advise flag
476 mg - mergable advise flag
478 Note that there is no guarantee that every flag and associated mnemonic will
479 be present in all further kernel releases. Things get changed, the flags may
480 be vanished or the reverse -- new added.
482 This file is only present if the CONFIG_MMU kernel configuration option is
485 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
486 bits on both physical and virtual pages associated with a process, and the
487 soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
488 To clear the bits for all the pages associated with the process
489 > echo 1 > /proc/PID/clear_refs
491 To clear the bits for the anonymous pages associated with the process
492 > echo 2 > /proc/PID/clear_refs
494 To clear the bits for the file mapped pages associated with the process
495 > echo 3 > /proc/PID/clear_refs
497 To clear the soft-dirty bit
498 > echo 4 > /proc/PID/clear_refs
500 To reset the peak resident set size ("high water mark") to the process's
502 > echo 5 > /proc/PID/clear_refs
504 Any other value written to /proc/PID/clear_refs will have no effect.
506 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
507 using /proc/kpageflags and number of times a page is mapped using
508 /proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
510 The /proc/pid/numa_maps is an extension based on maps, showing the memory
511 locality and binding policy, as well as the memory usage (in pages) of
512 each mapping. The output follows a general format where mapping details get
513 summarized separated by blank spaces, one mapping per each file line:
515 address policy mapping details
517 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
518 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
519 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
520 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
521 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
522 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
523 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
524 320698b000 default file=/lib64/libc-2.12.so
525 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
526 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
527 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
528 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
529 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
530 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
531 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
532 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
535 "address" is the starting address for the mapping;
536 "policy" reports the NUMA memory policy set for the mapping (see vm/numa_memory_policy.txt);
537 "mapping details" summarizes mapping data such as mapping type, page usage counters,
538 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
539 size, in KB, that is backing the mapping up.
544 Similar to the process entries, the kernel data files give information about
545 the running kernel. The files used to obtain this information are contained in
546 /proc and are listed in Table 1-5. Not all of these will be present in your
547 system. It depends on the kernel configuration and the loaded modules, which
548 files are there, and which are missing.
550 Table 1-5: Kernel info in /proc
551 ..............................................................................
553 apm Advanced power management info
554 buddyinfo Kernel memory allocator information (see text) (2.5)
555 bus Directory containing bus specific information
556 cmdline Kernel command line
557 cpuinfo Info about the CPU
558 devices Available devices (block and character)
559 dma Used DMS channels
560 filesystems Supported filesystems
561 driver Various drivers grouped here, currently rtc (2.4)
562 execdomains Execdomains, related to security (2.4)
563 fb Frame Buffer devices (2.4)
564 fs File system parameters, currently nfs/exports (2.4)
565 ide Directory containing info about the IDE subsystem
566 interrupts Interrupt usage
567 iomem Memory map (2.4)
568 ioports I/O port usage
569 irq Masks for irq to cpu affinity (2.4)(smp?)
570 isapnp ISA PnP (Plug&Play) Info (2.4)
571 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
573 ksyms Kernel symbol table
574 loadavg Load average of last 1, 5 & 15 minutes
578 modules List of loaded modules
579 mounts Mounted filesystems
580 net Networking info (see text)
581 pagetypeinfo Additional page allocator information (see text) (2.5)
582 partitions Table of partitions known to the system
583 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
584 decoupled by lspci (2.4)
586 scsi SCSI info (see text)
587 slabinfo Slab pool info
588 softirqs softirq usage
589 stat Overall statistics
590 swaps Swap space utilization
592 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
593 tty Info of tty drivers
594 uptime Wall clock since boot, combined idle time of all cpus
595 version Kernel version
596 video bttv info of video resources (2.4)
597 vmallocinfo Show vmalloced areas
598 ..............................................................................
600 You can, for example, check which interrupts are currently in use and what
601 they are used for by looking in the file /proc/interrupts:
603 > cat /proc/interrupts
605 0: 8728810 XT-PIC timer
606 1: 895 XT-PIC keyboard
608 3: 531695 XT-PIC aha152x
609 4: 2014133 XT-PIC serial
610 5: 44401 XT-PIC pcnet_cs
613 12: 182918 XT-PIC PS/2 Mouse
615 14: 1232265 XT-PIC ide0
619 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
620 output of a SMP machine):
622 > cat /proc/interrupts
625 0: 1243498 1214548 IO-APIC-edge timer
626 1: 8949 8958 IO-APIC-edge keyboard
627 2: 0 0 XT-PIC cascade
628 5: 11286 10161 IO-APIC-edge soundblaster
629 8: 1 0 IO-APIC-edge rtc
630 9: 27422 27407 IO-APIC-edge 3c503
631 12: 113645 113873 IO-APIC-edge PS/2 Mouse
633 14: 22491 24012 IO-APIC-edge ide0
634 15: 2183 2415 IO-APIC-edge ide1
635 17: 30564 30414 IO-APIC-level eth0
636 18: 177 164 IO-APIC-level bttv
641 NMI is incremented in this case because every timer interrupt generates a NMI
642 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
644 LOC is the local interrupt counter of the internal APIC of every CPU.
646 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
647 connects the CPUs in a SMP system. This means that an error has been detected,
648 the IO-APIC automatically retry the transmission, so it should not be a big
649 problem, but you should read the SMP-FAQ.
651 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
652 /proc/interrupts to display every IRQ vector in use by the system, not
653 just those considered 'most important'. The new vectors are:
655 THR -- interrupt raised when a machine check threshold counter
656 (typically counting ECC corrected errors of memory or cache) exceeds
657 a configurable threshold. Only available on some systems.
659 TRM -- a thermal event interrupt occurs when a temperature threshold
660 has been exceeded for the CPU. This interrupt may also be generated
661 when the temperature drops back to normal.
663 SPU -- a spurious interrupt is some interrupt that was raised then lowered
664 by some IO device before it could be fully processed by the APIC. Hence
665 the APIC sees the interrupt but does not know what device it came from.
666 For this case the APIC will generate the interrupt with a IRQ vector
667 of 0xff. This might also be generated by chipset bugs.
669 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
670 sent from one CPU to another per the needs of the OS. Typically,
671 their statistics are used by kernel developers and interested users to
672 determine the occurrence of interrupts of the given type.
674 The above IRQ vectors are displayed only when relevant. For example,
675 the threshold vector does not exist on x86_64 platforms. Others are
676 suppressed when the system is a uniprocessor. As of this writing, only
677 i386 and x86_64 platforms support the new IRQ vector displays.
679 Of some interest is the introduction of the /proc/irq directory to 2.4.
680 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
681 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
682 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
687 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
688 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
692 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
693 IRQ, you can set it by doing:
695 > echo 1 > /proc/irq/10/smp_affinity
697 This means that only the first CPU will handle the IRQ, but you can also echo
698 5 which means that only the first and fourth CPU can handle the IRQ.
700 The contents of each smp_affinity file is the same by default:
702 > cat /proc/irq/0/smp_affinity
705 There is an alternate interface, smp_affinity_list which allows specifying
706 a cpu range instead of a bitmask:
708 > cat /proc/irq/0/smp_affinity_list
711 The default_smp_affinity mask applies to all non-active IRQs, which are the
712 IRQs which have not yet been allocated/activated, and hence which lack a
713 /proc/irq/[0-9]* directory.
715 The node file on an SMP system shows the node to which the device using the IRQ
716 reports itself as being attached. This hardware locality information does not
717 include information about any possible driver locality preference.
719 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
720 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
722 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
723 between all the CPUs which are allowed to handle it. As usual the kernel has
724 more info than you and does a better job than you, so the defaults are the
725 best choice for almost everyone. [Note this applies only to those IO-APIC's
726 that support "Round Robin" interrupt distribution.]
728 There are three more important subdirectories in /proc: net, scsi, and sys.
729 The general rule is that the contents, or even the existence of these
730 directories, depend on your kernel configuration. If SCSI is not enabled, the
731 directory scsi may not exist. The same is true with the net, which is there
732 only when networking support is present in the running kernel.
734 The slabinfo file gives information about memory usage at the slab level.
735 Linux uses slab pools for memory management above page level in version 2.2.
736 Commonly used objects have their own slab pool (such as network buffers,
737 directory cache, and so on).
739 ..............................................................................
741 > cat /proc/buddyinfo
743 Node 0, zone DMA 0 4 5 4 4 3 ...
744 Node 0, zone Normal 1 0 0 1 101 8 ...
745 Node 0, zone HighMem 2 0 0 1 1 0 ...
747 External fragmentation is a problem under some workloads, and buddyinfo is a
748 useful tool for helping diagnose these problems. Buddyinfo will give you a
749 clue as to how big an area you can safely allocate, or why a previous
752 Each column represents the number of pages of a certain order which are
753 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
754 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
755 available in ZONE_NORMAL, etc...
757 More information relevant to external fragmentation can be found in
760 > cat /proc/pagetypeinfo
764 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
765 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
766 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
767 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
768 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
769 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
770 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
771 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
772 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
773 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
774 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
776 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
777 Node 0, zone DMA 2 0 5 1 0
778 Node 0, zone DMA32 41 6 967 2 0
780 Fragmentation avoidance in the kernel works by grouping pages of different
781 migrate types into the same contiguous regions of memory called page blocks.
782 A page block is typically the size of the default hugepage size e.g. 2MB on
783 X86-64. By keeping pages grouped based on their ability to move, the kernel
784 can reclaim pages within a page block to satisfy a high-order allocation.
786 The pagetypinfo begins with information on the size of a page block. It
787 then gives the same type of information as buddyinfo except broken down
788 by migrate-type and finishes with details on how many page blocks of each
791 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
792 from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
793 make an estimate of the likely number of huge pages that can be allocated
794 at a given point in time. All the "Movable" blocks should be allocatable
795 unless memory has been mlock()'d. Some of the Reclaimable blocks should
796 also be allocatable although a lot of filesystem metadata may have to be
797 reclaimed to achieve this.
799 ..............................................................................
803 Provides information about distribution and utilization of memory. This
804 varies by architecture and compile options. The following is from a
805 16GB PIII, which has highmem enabled. You may not have all of these fields.
809 The "Locked" indicates whether the mapping is locked in memory or not.
812 MemTotal: 16344972 kB
814 MemAvailable: 14836172 kB
820 HighTotal: 15597528 kB
821 HighFree: 13629632 kB
831 SReclaimable: 159856 kB
832 SUnreclaim: 124508 kB
837 CommitLimit: 7669796 kB
838 Committed_AS: 100056 kB
839 VmallocTotal: 112216 kB
841 VmallocChunk: 111088 kB
842 AnonHugePages: 49152 kB
844 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
845 bits and the kernel binary code)
846 MemFree: The sum of LowFree+HighFree
847 MemAvailable: An estimate of how much memory is available for starting new
848 applications, without swapping. Calculated from MemFree,
849 SReclaimable, the size of the file LRU lists, and the low
850 watermarks in each zone.
851 The estimate takes into account that the system needs some
852 page cache to function well, and that not all reclaimable
853 slab will be reclaimable, due to items being in use. The
854 impact of those factors will vary from system to system.
855 Buffers: Relatively temporary storage for raw disk blocks
856 shouldn't get tremendously large (20MB or so)
857 Cached: in-memory cache for files read from the disk (the
858 pagecache). Doesn't include SwapCached
859 SwapCached: Memory that once was swapped out, is swapped back in but
860 still also is in the swapfile (if memory is needed it
861 doesn't need to be swapped out AGAIN because it is already
862 in the swapfile. This saves I/O)
863 Active: Memory that has been used more recently and usually not
864 reclaimed unless absolutely necessary.
865 Inactive: Memory which has been less recently used. It is more
866 eligible to be reclaimed for other purposes
868 HighFree: Highmem is all memory above ~860MB of physical memory
869 Highmem areas are for use by userspace programs, or
870 for the pagecache. The kernel must use tricks to access
871 this memory, making it slower to access than lowmem.
873 LowFree: Lowmem is memory which can be used for everything that
874 highmem can be used for, but it is also available for the
875 kernel's use for its own data structures. Among many
876 other things, it is where everything from the Slab is
877 allocated. Bad things happen when you're out of lowmem.
878 SwapTotal: total amount of swap space available
879 SwapFree: Memory which has been evicted from RAM, and is temporarily
881 Dirty: Memory which is waiting to get written back to the disk
882 Writeback: Memory which is actively being written back to the disk
883 AnonPages: Non-file backed pages mapped into userspace page tables
884 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
885 Mapped: files which have been mmaped, such as libraries
886 Slab: in-kernel data structures cache
887 SReclaimable: Part of Slab, that might be reclaimed, such as caches
888 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
889 PageTables: amount of memory dedicated to the lowest level of page
891 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
893 Bounce: Memory used for block device "bounce buffers"
894 WritebackTmp: Memory used by FUSE for temporary writeback buffers
895 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
896 this is the total amount of memory currently available to
897 be allocated on the system. This limit is only adhered to
898 if strict overcommit accounting is enabled (mode 2 in
899 'vm.overcommit_memory').
900 The CommitLimit is calculated with the following formula:
901 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
902 overcommit_ratio / 100 + [total swap pages]
903 For example, on a system with 1G of physical RAM and 7G
904 of swap with a `vm.overcommit_ratio` of 30 it would
905 yield a CommitLimit of 7.3G.
906 For more details, see the memory overcommit documentation
907 in vm/overcommit-accounting.
908 Committed_AS: The amount of memory presently allocated on the system.
909 The committed memory is a sum of all of the memory which
910 has been allocated by processes, even if it has not been
911 "used" by them as of yet. A process which malloc()'s 1G
912 of memory, but only touches 300M of it will show up as
913 using 1G. This 1G is memory which has been "committed" to
914 by the VM and can be used at any time by the allocating
915 application. With strict overcommit enabled on the system
916 (mode 2 in 'vm.overcommit_memory'),allocations which would
917 exceed the CommitLimit (detailed above) will not be permitted.
918 This is useful if one needs to guarantee that processes will
919 not fail due to lack of memory once that memory has been
920 successfully allocated.
921 VmallocTotal: total size of vmalloc memory area
922 VmallocUsed: amount of vmalloc area which is used
923 VmallocChunk: largest contiguous block of vmalloc area which is free
925 ..............................................................................
929 Provides information about vmalloced/vmaped areas. One line per area,
930 containing the virtual address range of the area, size in bytes,
931 caller information of the creator, and optional information depending
932 on the kind of area :
934 pages=nr number of pages
935 phys=addr if a physical address was specified
936 ioremap I/O mapping (ioremap() and friends)
937 vmalloc vmalloc() area
940 vpages buffer for pages pointers was vmalloced (huge area)
941 N<node>=nr (Only on NUMA kernels)
942 Number of pages allocated on memory node <node>
944 > cat /proc/vmallocinfo
945 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
946 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
947 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
948 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
949 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
950 phys=7fee8000 ioremap
951 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
952 phys=7fee7000 ioremap
953 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
954 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
955 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
956 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
958 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
959 /0x130 [x_tables] pages=4 vmalloc N0=4
960 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
961 pages=14 vmalloc N2=14
962 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
964 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
966 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
967 pages=10 vmalloc N0=10
969 ..............................................................................
973 Provides counts of softirq handlers serviced since boot time, for each cpu.
978 TIMER: 27166 27120 27097 27034
983 SCHED: 27035 26983 26971 26746
985 RCU: 1678 1769 2178 2250
988 1.3 IDE devices in /proc/ide
989 ----------------------------
991 The subdirectory /proc/ide contains information about all IDE devices of which
992 the kernel is aware. There is one subdirectory for each IDE controller, the
993 file drivers and a link for each IDE device, pointing to the device directory
994 in the controller specific subtree.
996 The file drivers contains general information about the drivers used for the
999 > cat /proc/ide/drivers
1000 ide-cdrom version 4.53
1001 ide-disk version 1.08
1003 More detailed information can be found in the controller specific
1004 subdirectories. These are named ide0, ide1 and so on. Each of these
1005 directories contains the files shown in table 1-6.
1008 Table 1-6: IDE controller info in /proc/ide/ide?
1009 ..............................................................................
1011 channel IDE channel (0 or 1)
1012 config Configuration (only for PCI/IDE bridge)
1014 model Type/Chipset of IDE controller
1015 ..............................................................................
1017 Each device connected to a controller has a separate subdirectory in the
1018 controllers directory. The files listed in table 1-7 are contained in these
1022 Table 1-7: IDE device information
1023 ..............................................................................
1026 capacity Capacity of the medium (in 512Byte blocks)
1027 driver driver and version
1028 geometry physical and logical geometry
1029 identify device identify block
1031 model device identifier
1032 settings device setup
1033 smart_thresholds IDE disk management thresholds
1034 smart_values IDE disk management values
1035 ..............................................................................
1037 The most interesting file is settings. This file contains a nice overview of
1038 the drive parameters:
1040 # cat /proc/ide/ide0/hda/settings
1041 name value min max mode
1042 ---- ----- --- --- ----
1043 bios_cyl 526 0 65535 rw
1044 bios_head 255 0 255 rw
1045 bios_sect 63 0 63 rw
1046 breada_readahead 4 0 127 rw
1048 file_readahead 72 0 2097151 rw
1050 keepsettings 0 0 1 rw
1051 max_kb_per_request 122 1 127 rw
1055 pio_mode write-only 0 255 w
1061 1.4 Networking info in /proc/net
1062 --------------------------------
1064 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1065 additional values you get for IP version 6 if you configure the kernel to
1066 support this. Table 1-9 lists the files and their meaning.
1069 Table 1-8: IPv6 info in /proc/net
1070 ..............................................................................
1072 udp6 UDP sockets (IPv6)
1073 tcp6 TCP sockets (IPv6)
1074 raw6 Raw device statistics (IPv6)
1075 igmp6 IP multicast addresses, which this host joined (IPv6)
1076 if_inet6 List of IPv6 interface addresses
1077 ipv6_route Kernel routing table for IPv6
1078 rt6_stats Global IPv6 routing tables statistics
1079 sockstat6 Socket statistics (IPv6)
1080 snmp6 Snmp data (IPv6)
1081 ..............................................................................
1084 Table 1-9: Network info in /proc/net
1085 ..............................................................................
1087 arp Kernel ARP table
1088 dev network devices with statistics
1089 dev_mcast the Layer2 multicast groups a device is listening too
1090 (interface index, label, number of references, number of bound
1092 dev_stat network device status
1093 ip_fwchains Firewall chain linkage
1094 ip_fwnames Firewall chain names
1095 ip_masq Directory containing the masquerading tables
1096 ip_masquerade Major masquerading table
1097 netstat Network statistics
1098 raw raw device statistics
1099 route Kernel routing table
1100 rpc Directory containing rpc info
1101 rt_cache Routing cache
1103 sockstat Socket statistics
1106 unix UNIX domain sockets
1107 wireless Wireless interface data (Wavelan etc)
1108 igmp IP multicast addresses, which this host joined
1109 psched Global packet scheduler parameters.
1110 netlink List of PF_NETLINK sockets
1111 ip_mr_vifs List of multicast virtual interfaces
1112 ip_mr_cache List of multicast routing cache
1113 ..............................................................................
1115 You can use this information to see which network devices are available in
1116 your system and how much traffic was routed over those devices:
1119 Inter-|Receive |[...
1120 face |bytes packets errs drop fifo frame compressed multicast|[...
1121 lo: 908188 5596 0 0 0 0 0 0 [...
1122 ppp0:15475140 20721 410 0 0 410 0 0 [...
1123 eth0: 614530 7085 0 0 0 0 0 1 [...
1126 ...] bytes packets errs drop fifo colls carrier compressed
1127 ...] 908188 5596 0 0 0 0 0 0
1128 ...] 1375103 17405 0 0 0 0 0 0
1129 ...] 1703981 5535 0 0 0 3 0 0
1131 In addition, each Channel Bond interface has its own directory. For
1132 example, the bond0 device will have a directory called /proc/net/bond0/.
1133 It will contain information that is specific to that bond, such as the
1134 current slaves of the bond, the link status of the slaves, and how
1135 many times the slaves link has failed.
1140 If you have a SCSI host adapter in your system, you'll find a subdirectory
1141 named after the driver for this adapter in /proc/scsi. You'll also see a list
1142 of all recognized SCSI devices in /proc/scsi:
1144 >cat /proc/scsi/scsi
1146 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1147 Vendor: IBM Model: DGHS09U Rev: 03E0
1148 Type: Direct-Access ANSI SCSI revision: 03
1149 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1150 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1151 Type: CD-ROM ANSI SCSI revision: 02
1154 The directory named after the driver has one file for each adapter found in
1155 the system. These files contain information about the controller, including
1156 the used IRQ and the IO address range. The amount of information shown is
1157 dependent on the adapter you use. The example shows the output for an Adaptec
1158 AHA-2940 SCSI adapter:
1160 > cat /proc/scsi/aic7xxx/0
1162 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1164 TCQ Enabled By Default : Disabled
1165 AIC7XXX_PROC_STATS : Disabled
1166 AIC7XXX_RESET_DELAY : 5
1167 Adapter Configuration:
1168 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1169 Ultra Wide Controller
1170 PCI MMAPed I/O Base: 0xeb001000
1171 Adapter SEEPROM Config: SEEPROM found and used.
1172 Adaptec SCSI BIOS: Enabled
1174 SCBs: Active 0, Max Active 2,
1175 Allocated 15, HW 16, Page 255
1177 BIOS Control Word: 0x18b6
1178 Adapter Control Word: 0x005b
1179 Extended Translation: Enabled
1180 Disconnect Enable Flags: 0xffff
1181 Ultra Enable Flags: 0x0001
1182 Tag Queue Enable Flags: 0x0000
1183 Ordered Queue Tag Flags: 0x0000
1184 Default Tag Queue Depth: 8
1185 Tagged Queue By Device array for aic7xxx host instance 0:
1186 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1187 Actual queue depth per device for aic7xxx host instance 0:
1188 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1191 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1192 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1193 Total transfers 160151 (74577 reads and 85574 writes)
1195 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1196 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1197 Total transfers 0 (0 reads and 0 writes)
1200 1.6 Parallel port info in /proc/parport
1201 ---------------------------------------
1203 The directory /proc/parport contains information about the parallel ports of
1204 your system. It has one subdirectory for each port, named after the port
1207 These directories contain the four files shown in Table 1-10.
1210 Table 1-10: Files in /proc/parport
1211 ..............................................................................
1213 autoprobe Any IEEE-1284 device ID information that has been acquired.
1214 devices list of the device drivers using that port. A + will appear by the
1215 name of the device currently using the port (it might not appear
1217 hardware Parallel port's base address, IRQ line and DMA channel.
1218 irq IRQ that parport is using for that port. This is in a separate
1219 file to allow you to alter it by writing a new value in (IRQ
1221 ..............................................................................
1223 1.7 TTY info in /proc/tty
1224 -------------------------
1226 Information about the available and actually used tty's can be found in the
1227 directory /proc/tty.You'll find entries for drivers and line disciplines in
1228 this directory, as shown in Table 1-11.
1231 Table 1-11: Files in /proc/tty
1232 ..............................................................................
1234 drivers list of drivers and their usage
1235 ldiscs registered line disciplines
1236 driver/serial usage statistic and status of single tty lines
1237 ..............................................................................
1239 To see which tty's are currently in use, you can simply look into the file
1242 > cat /proc/tty/drivers
1243 pty_slave /dev/pts 136 0-255 pty:slave
1244 pty_master /dev/ptm 128 0-255 pty:master
1245 pty_slave /dev/ttyp 3 0-255 pty:slave
1246 pty_master /dev/pty 2 0-255 pty:master
1247 serial /dev/cua 5 64-67 serial:callout
1248 serial /dev/ttyS 4 64-67 serial
1249 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1250 /dev/ptmx /dev/ptmx 5 2 system
1251 /dev/console /dev/console 5 1 system:console
1252 /dev/tty /dev/tty 5 0 system:/dev/tty
1253 unknown /dev/tty 4 1-63 console
1256 1.8 Miscellaneous kernel statistics in /proc/stat
1257 -------------------------------------------------
1259 Various pieces of information about kernel activity are available in the
1260 /proc/stat file. All of the numbers reported in this file are aggregates
1261 since the system first booted. For a quick look, simply cat the file:
1264 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1265 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1266 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1267 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1273 softirq 183433 0 21755 12 39 1137 231 21459 2263
1275 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1276 lines. These numbers identify the amount of time the CPU has spent performing
1277 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1278 second). The meanings of the columns are as follows, from left to right:
1280 - user: normal processes executing in user mode
1281 - nice: niced processes executing in user mode
1282 - system: processes executing in kernel mode
1283 - idle: twiddling thumbs
1284 - iowait: waiting for I/O to complete
1285 - irq: servicing interrupts
1286 - softirq: servicing softirqs
1287 - steal: involuntary wait
1288 - guest: running a normal guest
1289 - guest_nice: running a niced guest
1291 The "intr" line gives counts of interrupts serviced since boot time, for each
1292 of the possible system interrupts. The first column is the total of all
1293 interrupts serviced including unnumbered architecture specific interrupts;
1294 each subsequent column is the total for that particular numbered interrupt.
1295 Unnumbered interrupts are not shown, only summed into the total.
1297 The "ctxt" line gives the total number of context switches across all CPUs.
1299 The "btime" line gives the time at which the system booted, in seconds since
1302 The "processes" line gives the number of processes and threads created, which
1303 includes (but is not limited to) those created by calls to the fork() and
1304 clone() system calls.
1306 The "procs_running" line gives the total number of threads that are
1307 running or ready to run (i.e., the total number of runnable threads).
1309 The "procs_blocked" line gives the number of processes currently blocked,
1310 waiting for I/O to complete.
1312 The "softirq" line gives counts of softirqs serviced since boot time, for each
1313 of the possible system softirqs. The first column is the total of all
1314 softirqs serviced; each subsequent column is the total for that particular
1318 1.9 Ext4 file system parameters
1319 -------------------------------
1321 Information about mounted ext4 file systems can be found in
1322 /proc/fs/ext4. Each mounted filesystem will have a directory in
1323 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1324 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1325 in Table 1-12, below.
1327 Table 1-12: Files in /proc/fs/ext4/<devname>
1328 ..............................................................................
1330 mb_groups details of multiblock allocator buddy cache of free blocks
1331 ..............................................................................
1335 Shows registered system console lines.
1337 To see which character device lines are currently used for the system console
1338 /dev/console, you may simply look into the file /proc/consoles:
1340 > cat /proc/consoles
1346 device name of the device
1347 operations R = can do read operations
1348 W = can do write operations
1350 flags E = it is enabled
1351 C = it is preferred console
1352 B = it is primary boot console
1353 p = it is used for printk buffer
1354 b = it is not a TTY but a Braille device
1355 a = it is safe to use when cpu is offline
1356 major:minor major and minor number of the device separated by a colon
1358 ------------------------------------------------------------------------------
1360 ------------------------------------------------------------------------------
1361 The /proc file system serves information about the running system. It not only
1362 allows access to process data but also allows you to request the kernel status
1363 by reading files in the hierarchy.
1365 The directory structure of /proc reflects the types of information and makes
1366 it easy, if not obvious, where to look for specific data.
1367 ------------------------------------------------------------------------------
1369 ------------------------------------------------------------------------------
1370 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1371 ------------------------------------------------------------------------------
1373 ------------------------------------------------------------------------------
1375 ------------------------------------------------------------------------------
1376 * Modifying kernel parameters by writing into files found in /proc/sys
1377 * Exploring the files which modify certain parameters
1378 * Review of the /proc/sys file tree
1379 ------------------------------------------------------------------------------
1382 A very interesting part of /proc is the directory /proc/sys. This is not only
1383 a source of information, it also allows you to change parameters within the
1384 kernel. Be very careful when attempting this. You can optimize your system,
1385 but you can also cause it to crash. Never alter kernel parameters on a
1386 production system. Set up a development machine and test to make sure that
1387 everything works the way you want it to. You may have no alternative but to
1388 reboot the machine once an error has been made.
1390 To change a value, simply echo the new value into the file. An example is
1391 given below in the section on the file system data. You need to be root to do
1392 this. You can create your own boot script to perform this every time your
1395 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1396 general things in the operation of the Linux kernel. Since some of the files
1397 can inadvertently disrupt your system, it is advisable to read both
1398 documentation and source before actually making adjustments. In any case, be
1399 very careful when writing to any of these files. The entries in /proc may
1400 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1401 review the kernel documentation in the directory /usr/src/linux/Documentation.
1402 This chapter is heavily based on the documentation included in the pre 2.2
1403 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1405 Please see: Documentation/sysctl/ directory for descriptions of these
1408 ------------------------------------------------------------------------------
1410 ------------------------------------------------------------------------------
1411 Certain aspects of kernel behavior can be modified at runtime, without the
1412 need to recompile the kernel, or even to reboot the system. The files in the
1413 /proc/sys tree can not only be read, but also modified. You can use the echo
1414 command to write value into these files, thereby changing the default settings
1416 ------------------------------------------------------------------------------
1418 ------------------------------------------------------------------------------
1419 CHAPTER 3: PER-PROCESS PARAMETERS
1420 ------------------------------------------------------------------------------
1422 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1423 --------------------------------------------------------------------------------
1425 These file can be used to adjust the badness heuristic used to select which
1426 process gets killed in out of memory conditions.
1428 The badness heuristic assigns a value to each candidate task ranging from 0
1429 (never kill) to 1000 (always kill) to determine which process is targeted. The
1430 units are roughly a proportion along that range of allowed memory the process
1431 may allocate from based on an estimation of its current memory and swap use.
1432 For example, if a task is using all allowed memory, its badness score will be
1433 1000. If it is using half of its allowed memory, its score will be 500.
1435 There is an additional factor included in the badness score: the current memory
1436 and swap usage is discounted by 3% for root processes.
1438 The amount of "allowed" memory depends on the context in which the oom killer
1439 was called. If it is due to the memory assigned to the allocating task's cpuset
1440 being exhausted, the allowed memory represents the set of mems assigned to that
1441 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1442 memory represents the set of mempolicy nodes. If it is due to a memory
1443 limit (or swap limit) being reached, the allowed memory is that configured
1444 limit. Finally, if it is due to the entire system being out of memory, the
1445 allowed memory represents all allocatable resources.
1447 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1448 is used to determine which task to kill. Acceptable values range from -1000
1449 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1450 polarize the preference for oom killing either by always preferring a certain
1451 task or completely disabling it. The lowest possible value, -1000, is
1452 equivalent to disabling oom killing entirely for that task since it will always
1453 report a badness score of 0.
1455 Consequently, it is very simple for userspace to define the amount of memory to
1456 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1457 example, is roughly equivalent to allowing the remainder of tasks sharing the
1458 same system, cpuset, mempolicy, or memory controller resources to use at least
1459 50% more memory. A value of -500, on the other hand, would be roughly
1460 equivalent to discounting 50% of the task's allowed memory from being considered
1461 as scoring against the task.
1463 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1464 be used to tune the badness score. Its acceptable values range from -16
1465 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1466 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1467 scaled linearly with /proc/<pid>/oom_score_adj.
1469 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1470 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1471 requires CAP_SYS_RESOURCE.
1473 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1474 generation children with separate address spaces instead, if possible. This
1475 avoids servers and important system daemons from being killed and loses the
1476 minimal amount of work.
1479 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1480 -------------------------------------------------------------
1482 This file can be used to check the current score used by the oom-killer is for
1483 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1484 process should be killed in an out-of-memory situation.
1487 3.3 /proc/<pid>/io - Display the IO accounting fields
1488 -------------------------------------------------------
1490 This file contains IO statistics for each running process
1495 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1498 test:/tmp # cat /proc/3828/io
1504 write_bytes: 323932160
1505 cancelled_write_bytes: 0
1514 I/O counter: chars read
1515 The number of bytes which this task has caused to be read from storage. This
1516 is simply the sum of bytes which this process passed to read() and pread().
1517 It includes things like tty IO and it is unaffected by whether or not actual
1518 physical disk IO was required (the read might have been satisfied from
1525 I/O counter: chars written
1526 The number of bytes which this task has caused, or shall cause to be written
1527 to disk. Similar caveats apply here as with rchar.
1533 I/O counter: read syscalls
1534 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1541 I/O counter: write syscalls
1542 Attempt to count the number of write I/O operations, i.e. syscalls like
1543 write() and pwrite().
1549 I/O counter: bytes read
1550 Attempt to count the number of bytes which this process really did cause to
1551 be fetched from the storage layer. Done at the submit_bio() level, so it is
1552 accurate for block-backed filesystems. <please add status regarding NFS and
1553 CIFS at a later time>
1559 I/O counter: bytes written
1560 Attempt to count the number of bytes which this process caused to be sent to
1561 the storage layer. This is done at page-dirtying time.
1564 cancelled_write_bytes
1565 ---------------------
1567 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1568 then deletes the file, it will in fact perform no writeout. But it will have
1569 been accounted as having caused 1MB of write.
1570 In other words: The number of bytes which this process caused to not happen,
1571 by truncating pagecache. A task can cause "negative" IO too. If this task
1572 truncates some dirty pagecache, some IO which another task has been accounted
1573 for (in its write_bytes) will not be happening. We _could_ just subtract that
1574 from the truncating task's write_bytes, but there is information loss in doing
1581 At its current implementation state, this is a bit racy on 32-bit machines: if
1582 process A reads process B's /proc/pid/io while process B is updating one of
1583 those 64-bit counters, process A could see an intermediate result.
1586 More information about this can be found within the taskstats documentation in
1587 Documentation/accounting.
1589 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1590 ---------------------------------------------------------------
1591 When a process is dumped, all anonymous memory is written to a core file as
1592 long as the size of the core file isn't limited. But sometimes we don't want
1593 to dump some memory segments, for example, huge shared memory. Conversely,
1594 sometimes we want to save file-backed memory segments into a core file, not
1595 only the individual files.
1597 /proc/<pid>/coredump_filter allows you to customize which memory segments
1598 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1599 of memory types. If a bit of the bitmask is set, memory segments of the
1600 corresponding memory type are dumped, otherwise they are not dumped.
1602 The following 7 memory types are supported:
1603 - (bit 0) anonymous private memory
1604 - (bit 1) anonymous shared memory
1605 - (bit 2) file-backed private memory
1606 - (bit 3) file-backed shared memory
1607 - (bit 4) ELF header pages in file-backed private memory areas (it is
1608 effective only if the bit 2 is cleared)
1609 - (bit 5) hugetlb private memory
1610 - (bit 6) hugetlb shared memory
1612 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1613 are always dumped regardless of the bitmask status.
1615 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1616 effected by bit 5-6.
1618 Default value of coredump_filter is 0x23; this means all anonymous memory
1619 segments and hugetlb private memory are dumped.
1621 If you don't want to dump all shared memory segments attached to pid 1234,
1622 write 0x21 to the process's proc file.
1624 $ echo 0x21 > /proc/1234/coredump_filter
1626 When a new process is created, the process inherits the bitmask status from its
1627 parent. It is useful to set up coredump_filter before the program runs.
1630 $ echo 0x7 > /proc/self/coredump_filter
1633 3.5 /proc/<pid>/mountinfo - Information about mounts
1634 --------------------------------------------------------
1636 This file contains lines of the form:
1638 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1639 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1641 (1) mount ID: unique identifier of the mount (may be reused after umount)
1642 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1643 (3) major:minor: value of st_dev for files on filesystem
1644 (4) root: root of the mount within the filesystem
1645 (5) mount point: mount point relative to the process's root
1646 (6) mount options: per mount options
1647 (7) optional fields: zero or more fields of the form "tag[:value]"
1648 (8) separator: marks the end of the optional fields
1649 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1650 (10) mount source: filesystem specific information or "none"
1651 (11) super options: per super block options
1653 Parsers should ignore all unrecognised optional fields. Currently the
1654 possible optional fields are:
1656 shared:X mount is shared in peer group X
1657 master:X mount is slave to peer group X
1658 propagate_from:X mount is slave and receives propagation from peer group X (*)
1659 unbindable mount is unbindable
1661 (*) X is the closest dominant peer group under the process's root. If
1662 X is the immediate master of the mount, or if there's no dominant peer
1663 group under the same root, then only the "master:X" field is present
1664 and not the "propagate_from:X" field.
1666 For more information on mount propagation see:
1668 Documentation/filesystems/sharedsubtree.txt
1671 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1672 --------------------------------------------------------
1673 These files provide a method to access a tasks comm value. It also allows for
1674 a task to set its own or one of its thread siblings comm value. The comm value
1675 is limited in size compared to the cmdline value, so writing anything longer
1676 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1680 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1681 -------------------------------------------------------------------------
1682 This file provides a fast way to retrieve first level children pids
1683 of a task pointed by <pid>/<tid> pair. The format is a space separated
1686 Note the "first level" here -- if a child has own children they will
1687 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1688 to obtain the descendants.
1690 Since this interface is intended to be fast and cheap it doesn't
1691 guarantee to provide precise results and some children might be
1692 skipped, especially if they've exited right after we printed their
1693 pids, so one need to either stop or freeze processes being inspected
1694 if precise results are needed.
1697 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1698 ---------------------------------------------------------------
1699 This file provides information associated with an opened file. The regular
1700 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1701 represents the current offset of the opened file in decimal form [see lseek(2)
1702 for details], 'flags' denotes the octal O_xxx mask the file has been
1703 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1704 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1713 All locks associated with a file descriptor are shown in its fdinfo too.
1715 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1717 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1718 pair provide additional information particular to the objects they represent.
1727 where 'eventfd-count' is hex value of a counter.
1734 sigmask: 0000000000000200
1736 where 'sigmask' is hex value of the signal mask associated
1744 tfd: 5 events: 1d data: ffffffffffffffff
1746 where 'tfd' is a target file descriptor number in decimal form,
1747 'events' is events mask being watched and the 'data' is data
1748 associated with a target [see epoll(7) for more details].
1752 For inotify files the format is the following
1756 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1758 where 'wd' is a watch descriptor in decimal form, ie a target file
1759 descriptor number, 'ino' and 'sdev' are inode and device where the
1760 target file resides and the 'mask' is the mask of events, all in hex
1761 form [see inotify(7) for more details].
1763 If the kernel was built with exportfs support, the path to the target
1764 file is encoded as a file handle. The file handle is provided by three
1765 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1768 If the kernel is built without exportfs support the file handle won't be
1771 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1773 For fanotify files the format is
1778 fanotify flags:10 event-flags:0
1779 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1780 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1782 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1783 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1784 flags associated with mark which are tracked separately from events
1785 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1786 mask and 'ignored_mask' is the mask of events which are to be ignored.
1787 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1788 does provide information about flags and mask used in fanotify_mark
1789 call [see fsnotify manpage for details].
1791 While the first three lines are mandatory and always printed, the rest is
1792 optional and may be omitted if no marks created yet.
1803 it_value: (0, 49406829)
1806 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1807 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1808 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1809 details]. 'it_value' is remaining time until the timer exiration.
1810 'it_interval' is the interval for the timer. Note the timer might be set up
1811 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1812 still exhibits timer's remaining time.
1814 3.9 /proc/<pid>/map_files - Information about memory mapped files
1815 ---------------------------------------------------------------------
1816 This directory contains symbolic links which represent memory mapped files
1817 the process is maintaining. Example output:
1819 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1820 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1821 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1823 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1824 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1826 The name of a link represents the virtual memory bounds of a mapping, i.e.
1827 vm_area_struct::vm_start-vm_area_struct::vm_end.
1829 The main purpose of the map_files is to retrieve a set of memory mapped
1830 files in a fast way instead of parsing /proc/<pid>/maps or
1831 /proc/<pid>/smaps, both of which contain many more records. At the same
1832 time one can open(2) mappings from the listings of two processes and
1833 comparing their inode numbers to figure out which anonymous memory areas
1834 are actually shared.
1836 ------------------------------------------------------------------------------
1838 ------------------------------------------------------------------------------
1841 ---------------------
1843 The following mount options are supported:
1845 hidepid= Set /proc/<pid>/ access mode.
1846 gid= Set the group authorized to learn processes information.
1848 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1851 hidepid=1 means users may not access any /proc/<pid>/ directories but their
1852 own. Sensitive files like cmdline, sched*, status are now protected against
1853 other users. This makes it impossible to learn whether any user runs
1854 specific program (given the program doesn't reveal itself by its behaviour).
1855 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1856 poorly written programs passing sensitive information via program arguments are
1857 now protected against local eavesdroppers.
1859 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1860 users. It doesn't mean that it hides a fact whether a process with a specific
1861 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1862 but it hides process' uid and gid, which may be learned by stat()'ing
1863 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1864 information about running processes, whether some daemon runs with elevated
1865 privileges, whether other user runs some sensitive program, whether other users
1866 run any program at all, etc.
1868 gid= defines a group authorized to learn processes information otherwise
1869 prohibited by hidepid=. If you use some daemon like identd which needs to learn
1870 information about processes information, just add identd to this group.
projects / deliverable / linux.git / blob