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 3.20.0)
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 VmSwap size of swap usage (the number of referred swapents)
239 Threads number of threads
240 SigQ number of signals queued/max. number for queue
241 SigPnd bitmap of pending signals for the thread
242 ShdPnd bitmap of shared pending signals for the process
243 SigBlk bitmap of blocked signals
244 SigIgn bitmap of ignored signals
245 SigCgt bitmap of caught signals
246 CapInh bitmap of inheritable capabilities
247 CapPrm bitmap of permitted capabilities
248 CapEff bitmap of effective capabilities
249 CapBnd bitmap of capabilities bounding set
250 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
251 Cpus_allowed mask of CPUs on which this process may run
252 Cpus_allowed_list Same as previous, but in "list format"
253 Mems_allowed mask of memory nodes allowed to this process
254 Mems_allowed_list Same as previous, but in "list format"
255 voluntary_ctxt_switches number of voluntary context switches
256 nonvoluntary_ctxt_switches number of non voluntary context switches
257 ..............................................................................
259 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
260 ..............................................................................
262 size total program size (pages) (same as VmSize in status)
263 resident size of memory portions (pages) (same as VmRSS in status)
264 shared number of pages that are shared (i.e. backed by a file)
265 trs number of pages that are 'code' (not including libs; broken,
266 includes data segment)
267 lrs number of pages of library (always 0 on 2.6)
268 drs number of pages of data/stack (including libs; broken,
269 includes library text)
270 dt number of dirty pages (always 0 on 2.6)
271 ..............................................................................
274 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
275 ..............................................................................
278 tcomm filename of the executable
279 state state (R is running, S is sleeping, D is sleeping in an
280 uninterruptible wait, Z is zombie, T is traced or stopped)
281 ppid process id of the parent process
282 pgrp pgrp of the process
284 tty_nr tty the process uses
285 tty_pgrp pgrp of the tty
287 min_flt number of minor faults
288 cmin_flt number of minor faults with child's
289 maj_flt number of major faults
290 cmaj_flt number of major faults with child's
291 utime user mode jiffies
292 stime kernel mode jiffies
293 cutime user mode jiffies with child's
294 cstime kernel mode jiffies with child's
295 priority priority level
297 num_threads number of threads
298 it_real_value (obsolete, always 0)
299 start_time time the process started after system boot
300 vsize virtual memory size
301 rss resident set memory size
302 rsslim current limit in bytes on the rss
303 start_code address above which program text can run
304 end_code address below which program text can run
305 start_stack address of the start of the main process stack
306 esp current value of ESP
307 eip current value of EIP
308 pending bitmap of pending signals
309 blocked bitmap of blocked signals
310 sigign bitmap of ignored signals
311 sigcatch bitmap of caught signals
312 wchan address where process went to sleep
315 exit_signal signal to send to parent thread on exit
316 task_cpu which CPU the task is scheduled on
317 rt_priority realtime priority
318 policy scheduling policy (man sched_setscheduler)
319 blkio_ticks time spent waiting for block IO
320 gtime guest time of the task in jiffies
321 cgtime guest time of the task children in jiffies
322 start_data address above which program data+bss is placed
323 end_data address below which program data+bss is placed
324 start_brk address above which program heap can be expanded with brk()
325 arg_start address above which program command line is placed
326 arg_end address below which program command line is placed
327 env_start address above which program environment is placed
328 env_end address below which program environment is placed
329 exit_code the thread's exit_code in the form reported by the waitpid system call
330 ..............................................................................
332 The /proc/PID/maps file containing the currently mapped memory regions and
333 their access permissions.
337 address perms offset dev inode pathname
339 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
340 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
341 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
342 a7cb1000-a7cb2000 ---p 00000000 00:00 0
343 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
344 a7eb2000-a7eb3000 ---p 00000000 00:00 0
345 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
346 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
347 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
348 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
349 a800b000-a800e000 rw-p 00000000 00:00 0
350 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
351 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
352 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
353 a8024000-a8027000 rw-p 00000000 00:00 0
354 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
355 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
356 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
357 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
358 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
360 where "address" is the address space in the process that it occupies, "perms"
361 is a set of permissions:
367 p = private (copy on write)
369 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
370 "inode" is the inode on that device. 0 indicates that no inode is associated
371 with the memory region, as the case would be with BSS (uninitialized data).
372 The "pathname" shows the name associated file for this mapping. If the mapping
373 is not associated with a file:
375 [heap] = the heap of the program
376 [stack] = the stack of the main process
377 [stack:1001] = the stack of the thread with tid 1001
378 [vdso] = the "virtual dynamic shared object",
379 the kernel system call handler
381 or if empty, the mapping is anonymous.
383 The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
384 of the individual tasks of a process. In this file you will see a mapping marked
385 as [stack] if that task sees it as a stack. This is a key difference from the
386 content of /proc/PID/maps, where you will see all mappings that are being used
387 as stack by all of those tasks. Hence, for the example above, the task-level
388 map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
390 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
391 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
392 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
393 a7cb1000-a7cb2000 ---p 00000000 00:00 0
394 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
395 a7eb2000-a7eb3000 ---p 00000000 00:00 0
396 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
397 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
398 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
399 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
400 a800b000-a800e000 rw-p 00000000 00:00 0
401 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
402 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
403 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
404 a8024000-a8027000 rw-p 00000000 00:00 0
405 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
406 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
407 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
408 aff35000-aff4a000 rw-p 00000000 00:00 0
409 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
411 The /proc/PID/smaps is an extension based on maps, showing the memory
412 consumption for each of the process's mappings. For each of mappings there
413 is a series of lines such as the following:
415 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
429 VmFlags: rd ex mr mw me de
431 the first of these lines shows the same information as is displayed for the
432 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
433 (size), the amount of the mapping that is currently resident in RAM (RSS), the
434 process' proportional share of this mapping (PSS), the number of clean and
435 dirty private pages in the mapping. Note that even a page which is part of a
436 MAP_SHARED mapping, but has only a single pte mapped, i.e. is currently used
437 by only one process, is accounted as private and not as shared. "Referenced"
438 indicates the amount of memory currently marked as referenced or accessed.
439 "Anonymous" shows the amount of memory that does not belong to any file. Even
440 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
441 and a page is modified, the file page is replaced by a private anonymous copy.
442 "Swap" shows how much would-be-anonymous memory is also used, but out on
445 "VmFlags" field deserves a separate description. This member represents the kernel
446 flags associated with the particular virtual memory area in two letter encoded
447 manner. The codes are the following:
456 gd - stack segment growns down
458 dw - disabled write to the mapped file
459 lo - pages are locked in memory
460 io - memory mapped I/O area
461 sr - sequential read advise provided
462 rr - random read advise provided
463 dc - do not copy area on fork
464 de - do not expand area on remapping
465 ac - area is accountable
466 nr - swap space is not reserved for the area
467 ht - area uses huge tlb pages
468 nl - non-linear mapping
469 ar - architecture specific flag
470 dd - do not include area into core dump
473 hg - huge page advise flag
474 nh - no-huge page advise flag
475 mg - mergable advise flag
477 Note that there is no guarantee that every flag and associated mnemonic will
478 be present in all further kernel releases. Things get changed, the flags may
479 be vanished or the reverse -- new added.
481 This file is only present if the CONFIG_MMU kernel configuration option is
484 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
485 bits on both physical and virtual pages associated with a process, and the
486 soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
487 To clear the bits for all the pages associated with the process
488 > echo 1 > /proc/PID/clear_refs
490 To clear the bits for the anonymous pages associated with the process
491 > echo 2 > /proc/PID/clear_refs
493 To clear the bits for the file mapped pages associated with the process
494 > echo 3 > /proc/PID/clear_refs
496 To clear the soft-dirty bit
497 > echo 4 > /proc/PID/clear_refs
499 To reset the peak resident set size ("high water mark") to the process's
501 > echo 5 > /proc/PID/clear_refs
503 Any other value written to /proc/PID/clear_refs will have no effect.
505 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
506 using /proc/kpageflags and number of times a page is mapped using
507 /proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
509 The /proc/pid/numa_maps is an extension based on maps, showing the memory
510 locality and binding policy, as well as the memory usage (in pages) of
511 each mapping. The output follows a general format where mapping details get
512 summarized separated by blank spaces, one mapping per each file line:
514 address policy mapping details
516 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
517 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
518 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
519 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
520 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
521 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
522 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
523 320698b000 default file=/lib64/libc-2.12.so
524 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
525 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
526 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
527 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
528 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
529 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
530 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
531 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
534 "address" is the starting address for the mapping;
535 "policy" reports the NUMA memory policy set for the mapping (see vm/numa_memory_policy.txt);
536 "mapping details" summarizes mapping data such as mapping type, page usage counters,
537 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
538 size, in KB, that is backing the mapping up.
543 Similar to the process entries, the kernel data files give information about
544 the running kernel. The files used to obtain this information are contained in
545 /proc and are listed in Table 1-5. Not all of these will be present in your
546 system. It depends on the kernel configuration and the loaded modules, which
547 files are there, and which are missing.
549 Table 1-5: Kernel info in /proc
550 ..............................................................................
552 apm Advanced power management info
553 buddyinfo Kernel memory allocator information (see text) (2.5)
554 bus Directory containing bus specific information
555 cmdline Kernel command line
556 cpuinfo Info about the CPU
557 devices Available devices (block and character)
558 dma Used DMS channels
559 filesystems Supported filesystems
560 driver Various drivers grouped here, currently rtc (2.4)
561 execdomains Execdomains, related to security (2.4)
562 fb Frame Buffer devices (2.4)
563 fs File system parameters, currently nfs/exports (2.4)
564 ide Directory containing info about the IDE subsystem
565 interrupts Interrupt usage
566 iomem Memory map (2.4)
567 ioports I/O port usage
568 irq Masks for irq to cpu affinity (2.4)(smp?)
569 isapnp ISA PnP (Plug&Play) Info (2.4)
570 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
572 ksyms Kernel symbol table
573 loadavg Load average of last 1, 5 & 15 minutes
577 modules List of loaded modules
578 mounts Mounted filesystems
579 net Networking info (see text)
580 pagetypeinfo Additional page allocator information (see text) (2.5)
581 partitions Table of partitions known to the system
582 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
583 decoupled by lspci (2.4)
585 scsi SCSI info (see text)
586 slabinfo Slab pool info
587 softirqs softirq usage
588 stat Overall statistics
589 swaps Swap space utilization
591 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
592 tty Info of tty drivers
593 uptime Wall clock since boot, combined idle time of all cpus
594 version Kernel version
595 video bttv info of video resources (2.4)
596 vmallocinfo Show vmalloced areas
597 ..............................................................................
599 You can, for example, check which interrupts are currently in use and what
600 they are used for by looking in the file /proc/interrupts:
602 > cat /proc/interrupts
604 0: 8728810 XT-PIC timer
605 1: 895 XT-PIC keyboard
607 3: 531695 XT-PIC aha152x
608 4: 2014133 XT-PIC serial
609 5: 44401 XT-PIC pcnet_cs
612 12: 182918 XT-PIC PS/2 Mouse
614 14: 1232265 XT-PIC ide0
618 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
619 output of a SMP machine):
621 > cat /proc/interrupts
624 0: 1243498 1214548 IO-APIC-edge timer
625 1: 8949 8958 IO-APIC-edge keyboard
626 2: 0 0 XT-PIC cascade
627 5: 11286 10161 IO-APIC-edge soundblaster
628 8: 1 0 IO-APIC-edge rtc
629 9: 27422 27407 IO-APIC-edge 3c503
630 12: 113645 113873 IO-APIC-edge PS/2 Mouse
632 14: 22491 24012 IO-APIC-edge ide0
633 15: 2183 2415 IO-APIC-edge ide1
634 17: 30564 30414 IO-APIC-level eth0
635 18: 177 164 IO-APIC-level bttv
640 NMI is incremented in this case because every timer interrupt generates a NMI
641 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
643 LOC is the local interrupt counter of the internal APIC of every CPU.
645 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
646 connects the CPUs in a SMP system. This means that an error has been detected,
647 the IO-APIC automatically retry the transmission, so it should not be a big
648 problem, but you should read the SMP-FAQ.
650 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
651 /proc/interrupts to display every IRQ vector in use by the system, not
652 just those considered 'most important'. The new vectors are:
654 THR -- interrupt raised when a machine check threshold counter
655 (typically counting ECC corrected errors of memory or cache) exceeds
656 a configurable threshold. Only available on some systems.
658 TRM -- a thermal event interrupt occurs when a temperature threshold
659 has been exceeded for the CPU. This interrupt may also be generated
660 when the temperature drops back to normal.
662 SPU -- a spurious interrupt is some interrupt that was raised then lowered
663 by some IO device before it could be fully processed by the APIC. Hence
664 the APIC sees the interrupt but does not know what device it came from.
665 For this case the APIC will generate the interrupt with a IRQ vector
666 of 0xff. This might also be generated by chipset bugs.
668 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
669 sent from one CPU to another per the needs of the OS. Typically,
670 their statistics are used by kernel developers and interested users to
671 determine the occurrence of interrupts of the given type.
673 The above IRQ vectors are displayed only when relevant. For example,
674 the threshold vector does not exist on x86_64 platforms. Others are
675 suppressed when the system is a uniprocessor. As of this writing, only
676 i386 and x86_64 platforms support the new IRQ vector displays.
678 Of some interest is the introduction of the /proc/irq directory to 2.4.
679 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
680 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
681 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
686 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
687 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
691 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
692 IRQ, you can set it by doing:
694 > echo 1 > /proc/irq/10/smp_affinity
696 This means that only the first CPU will handle the IRQ, but you can also echo
697 5 which means that only the first and fourth CPU can handle the IRQ.
699 The contents of each smp_affinity file is the same by default:
701 > cat /proc/irq/0/smp_affinity
704 There is an alternate interface, smp_affinity_list which allows specifying
705 a cpu range instead of a bitmask:
707 > cat /proc/irq/0/smp_affinity_list
710 The default_smp_affinity mask applies to all non-active IRQs, which are the
711 IRQs which have not yet been allocated/activated, and hence which lack a
712 /proc/irq/[0-9]* directory.
714 The node file on an SMP system shows the node to which the device using the IRQ
715 reports itself as being attached. This hardware locality information does not
716 include information about any possible driver locality preference.
718 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
719 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
721 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
722 between all the CPUs which are allowed to handle it. As usual the kernel has
723 more info than you and does a better job than you, so the defaults are the
724 best choice for almost everyone. [Note this applies only to those IO-APIC's
725 that support "Round Robin" interrupt distribution.]
727 There are three more important subdirectories in /proc: net, scsi, and sys.
728 The general rule is that the contents, or even the existence of these
729 directories, depend on your kernel configuration. If SCSI is not enabled, the
730 directory scsi may not exist. The same is true with the net, which is there
731 only when networking support is present in the running kernel.
733 The slabinfo file gives information about memory usage at the slab level.
734 Linux uses slab pools for memory management above page level in version 2.2.
735 Commonly used objects have their own slab pool (such as network buffers,
736 directory cache, and so on).
738 ..............................................................................
740 > cat /proc/buddyinfo
742 Node 0, zone DMA 0 4 5 4 4 3 ...
743 Node 0, zone Normal 1 0 0 1 101 8 ...
744 Node 0, zone HighMem 2 0 0 1 1 0 ...
746 External fragmentation is a problem under some workloads, and buddyinfo is a
747 useful tool for helping diagnose these problems. Buddyinfo will give you a
748 clue as to how big an area you can safely allocate, or why a previous
751 Each column represents the number of pages of a certain order which are
752 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
753 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
754 available in ZONE_NORMAL, etc...
756 More information relevant to external fragmentation can be found in
759 > cat /proc/pagetypeinfo
763 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
764 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
765 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
766 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
767 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
768 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
769 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
770 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
771 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
772 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
773 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
775 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
776 Node 0, zone DMA 2 0 5 1 0
777 Node 0, zone DMA32 41 6 967 2 0
779 Fragmentation avoidance in the kernel works by grouping pages of different
780 migrate types into the same contiguous regions of memory called page blocks.
781 A page block is typically the size of the default hugepage size e.g. 2MB on
782 X86-64. By keeping pages grouped based on their ability to move, the kernel
783 can reclaim pages within a page block to satisfy a high-order allocation.
785 The pagetypinfo begins with information on the size of a page block. It
786 then gives the same type of information as buddyinfo except broken down
787 by migrate-type and finishes with details on how many page blocks of each
790 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
791 from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
792 make an estimate of the likely number of huge pages that can be allocated
793 at a given point in time. All the "Movable" blocks should be allocatable
794 unless memory has been mlock()'d. Some of the Reclaimable blocks should
795 also be allocatable although a lot of filesystem metadata may have to be
796 reclaimed to achieve this.
798 ..............................................................................
802 Provides information about distribution and utilization of memory. This
803 varies by architecture and compile options. The following is from a
804 16GB PIII, which has highmem enabled. You may not have all of these fields.
808 The "Locked" indicates whether the mapping is locked in memory or not.
811 MemTotal: 16344972 kB
813 MemAvailable: 14836172 kB
819 HighTotal: 15597528 kB
820 HighFree: 13629632 kB
830 SReclaimable: 159856 kB
831 SUnreclaim: 124508 kB
836 CommitLimit: 7669796 kB
837 Committed_AS: 100056 kB
838 VmallocTotal: 112216 kB
840 VmallocChunk: 111088 kB
841 AnonHugePages: 49152 kB
843 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
844 bits and the kernel binary code)
845 MemFree: The sum of LowFree+HighFree
846 MemAvailable: An estimate of how much memory is available for starting new
847 applications, without swapping. Calculated from MemFree,
848 SReclaimable, the size of the file LRU lists, and the low
849 watermarks in each zone.
850 The estimate takes into account that the system needs some
851 page cache to function well, and that not all reclaimable
852 slab will be reclaimable, due to items being in use. The
853 impact of those factors will vary from system to system.
854 Buffers: Relatively temporary storage for raw disk blocks
855 shouldn't get tremendously large (20MB or so)
856 Cached: in-memory cache for files read from the disk (the
857 pagecache). Doesn't include SwapCached
858 SwapCached: Memory that once was swapped out, is swapped back in but
859 still also is in the swapfile (if memory is needed it
860 doesn't need to be swapped out AGAIN because it is already
861 in the swapfile. This saves I/O)
862 Active: Memory that has been used more recently and usually not
863 reclaimed unless absolutely necessary.
864 Inactive: Memory which has been less recently used. It is more
865 eligible to be reclaimed for other purposes
867 HighFree: Highmem is all memory above ~860MB of physical memory
868 Highmem areas are for use by userspace programs, or
869 for the pagecache. The kernel must use tricks to access
870 this memory, making it slower to access than lowmem.
872 LowFree: Lowmem is memory which can be used for everything that
873 highmem can be used for, but it is also available for the
874 kernel's use for its own data structures. Among many
875 other things, it is where everything from the Slab is
876 allocated. Bad things happen when you're out of lowmem.
877 SwapTotal: total amount of swap space available
878 SwapFree: Memory which has been evicted from RAM, and is temporarily
880 Dirty: Memory which is waiting to get written back to the disk
881 Writeback: Memory which is actively being written back to the disk
882 AnonPages: Non-file backed pages mapped into userspace page tables
883 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
884 Mapped: files which have been mmaped, such as libraries
885 Slab: in-kernel data structures cache
886 SReclaimable: Part of Slab, that might be reclaimed, such as caches
887 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
888 PageTables: amount of memory dedicated to the lowest level of page
890 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
892 Bounce: Memory used for block device "bounce buffers"
893 WritebackTmp: Memory used by FUSE for temporary writeback buffers
894 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
895 this is the total amount of memory currently available to
896 be allocated on the system. This limit is only adhered to
897 if strict overcommit accounting is enabled (mode 2 in
898 'vm.overcommit_memory').
899 The CommitLimit is calculated with the following formula:
900 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
901 overcommit_ratio / 100 + [total swap pages]
902 For example, on a system with 1G of physical RAM and 7G
903 of swap with a `vm.overcommit_ratio` of 30 it would
904 yield a CommitLimit of 7.3G.
905 For more details, see the memory overcommit documentation
906 in vm/overcommit-accounting.
907 Committed_AS: The amount of memory presently allocated on the system.
908 The committed memory is a sum of all of the memory which
909 has been allocated by processes, even if it has not been
910 "used" by them as of yet. A process which malloc()'s 1G
911 of memory, but only touches 300M of it will show up as
912 using 1G. This 1G is memory which has been "committed" to
913 by the VM and can be used at any time by the allocating
914 application. With strict overcommit enabled on the system
915 (mode 2 in 'vm.overcommit_memory'),allocations which would
916 exceed the CommitLimit (detailed above) will not be permitted.
917 This is useful if one needs to guarantee that processes will
918 not fail due to lack of memory once that memory has been
919 successfully allocated.
920 VmallocTotal: total size of vmalloc memory area
921 VmallocUsed: amount of vmalloc area which is used
922 VmallocChunk: largest contiguous block of vmalloc area which is free
924 ..............................................................................
928 Provides information about vmalloced/vmaped areas. One line per area,
929 containing the virtual address range of the area, size in bytes,
930 caller information of the creator, and optional information depending
931 on the kind of area :
933 pages=nr number of pages
934 phys=addr if a physical address was specified
935 ioremap I/O mapping (ioremap() and friends)
936 vmalloc vmalloc() area
939 vpages buffer for pages pointers was vmalloced (huge area)
940 N<node>=nr (Only on NUMA kernels)
941 Number of pages allocated on memory node <node>
943 > cat /proc/vmallocinfo
944 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
945 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
946 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
947 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
948 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
949 phys=7fee8000 ioremap
950 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
951 phys=7fee7000 ioremap
952 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
953 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
954 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
955 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
957 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
958 /0x130 [x_tables] pages=4 vmalloc N0=4
959 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
960 pages=14 vmalloc N2=14
961 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
963 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
965 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
966 pages=10 vmalloc N0=10
968 ..............................................................................
972 Provides counts of softirq handlers serviced since boot time, for each cpu.
977 TIMER: 27166 27120 27097 27034
982 SCHED: 27035 26983 26971 26746
984 RCU: 1678 1769 2178 2250
987 1.3 IDE devices in /proc/ide
988 ----------------------------
990 The subdirectory /proc/ide contains information about all IDE devices of which
991 the kernel is aware. There is one subdirectory for each IDE controller, the
992 file drivers and a link for each IDE device, pointing to the device directory
993 in the controller specific subtree.
995 The file drivers contains general information about the drivers used for the
998 > cat /proc/ide/drivers
999 ide-cdrom version 4.53
1000 ide-disk version 1.08
1002 More detailed information can be found in the controller specific
1003 subdirectories. These are named ide0, ide1 and so on. Each of these
1004 directories contains the files shown in table 1-6.
1007 Table 1-6: IDE controller info in /proc/ide/ide?
1008 ..............................................................................
1010 channel IDE channel (0 or 1)
1011 config Configuration (only for PCI/IDE bridge)
1013 model Type/Chipset of IDE controller
1014 ..............................................................................
1016 Each device connected to a controller has a separate subdirectory in the
1017 controllers directory. The files listed in table 1-7 are contained in these
1021 Table 1-7: IDE device information
1022 ..............................................................................
1025 capacity Capacity of the medium (in 512Byte blocks)
1026 driver driver and version
1027 geometry physical and logical geometry
1028 identify device identify block
1030 model device identifier
1031 settings device setup
1032 smart_thresholds IDE disk management thresholds
1033 smart_values IDE disk management values
1034 ..............................................................................
1036 The most interesting file is settings. This file contains a nice overview of
1037 the drive parameters:
1039 # cat /proc/ide/ide0/hda/settings
1040 name value min max mode
1041 ---- ----- --- --- ----
1042 bios_cyl 526 0 65535 rw
1043 bios_head 255 0 255 rw
1044 bios_sect 63 0 63 rw
1045 breada_readahead 4 0 127 rw
1047 file_readahead 72 0 2097151 rw
1049 keepsettings 0 0 1 rw
1050 max_kb_per_request 122 1 127 rw
1054 pio_mode write-only 0 255 w
1060 1.4 Networking info in /proc/net
1061 --------------------------------
1063 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1064 additional values you get for IP version 6 if you configure the kernel to
1065 support this. Table 1-9 lists the files and their meaning.
1068 Table 1-8: IPv6 info in /proc/net
1069 ..............................................................................
1071 udp6 UDP sockets (IPv6)
1072 tcp6 TCP sockets (IPv6)
1073 raw6 Raw device statistics (IPv6)
1074 igmp6 IP multicast addresses, which this host joined (IPv6)
1075 if_inet6 List of IPv6 interface addresses
1076 ipv6_route Kernel routing table for IPv6
1077 rt6_stats Global IPv6 routing tables statistics
1078 sockstat6 Socket statistics (IPv6)
1079 snmp6 Snmp data (IPv6)
1080 ..............................................................................
1083 Table 1-9: Network info in /proc/net
1084 ..............................................................................
1086 arp Kernel ARP table
1087 dev network devices with statistics
1088 dev_mcast the Layer2 multicast groups a device is listening too
1089 (interface index, label, number of references, number of bound
1091 dev_stat network device status
1092 ip_fwchains Firewall chain linkage
1093 ip_fwnames Firewall chain names
1094 ip_masq Directory containing the masquerading tables
1095 ip_masquerade Major masquerading table
1096 netstat Network statistics
1097 raw raw device statistics
1098 route Kernel routing table
1099 rpc Directory containing rpc info
1100 rt_cache Routing cache
1102 sockstat Socket statistics
1105 unix UNIX domain sockets
1106 wireless Wireless interface data (Wavelan etc)
1107 igmp IP multicast addresses, which this host joined
1108 psched Global packet scheduler parameters.
1109 netlink List of PF_NETLINK sockets
1110 ip_mr_vifs List of multicast virtual interfaces
1111 ip_mr_cache List of multicast routing cache
1112 ..............................................................................
1114 You can use this information to see which network devices are available in
1115 your system and how much traffic was routed over those devices:
1118 Inter-|Receive |[...
1119 face |bytes packets errs drop fifo frame compressed multicast|[...
1120 lo: 908188 5596 0 0 0 0 0 0 [...
1121 ppp0:15475140 20721 410 0 0 410 0 0 [...
1122 eth0: 614530 7085 0 0 0 0 0 1 [...
1125 ...] bytes packets errs drop fifo colls carrier compressed
1126 ...] 908188 5596 0 0 0 0 0 0
1127 ...] 1375103 17405 0 0 0 0 0 0
1128 ...] 1703981 5535 0 0 0 3 0 0
1130 In addition, each Channel Bond interface has its own directory. For
1131 example, the bond0 device will have a directory called /proc/net/bond0/.
1132 It will contain information that is specific to that bond, such as the
1133 current slaves of the bond, the link status of the slaves, and how
1134 many times the slaves link has failed.
1139 If you have a SCSI host adapter in your system, you'll find a subdirectory
1140 named after the driver for this adapter in /proc/scsi. You'll also see a list
1141 of all recognized SCSI devices in /proc/scsi:
1143 >cat /proc/scsi/scsi
1145 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1146 Vendor: IBM Model: DGHS09U Rev: 03E0
1147 Type: Direct-Access ANSI SCSI revision: 03
1148 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1149 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1150 Type: CD-ROM ANSI SCSI revision: 02
1153 The directory named after the driver has one file for each adapter found in
1154 the system. These files contain information about the controller, including
1155 the used IRQ and the IO address range. The amount of information shown is
1156 dependent on the adapter you use. The example shows the output for an Adaptec
1157 AHA-2940 SCSI adapter:
1159 > cat /proc/scsi/aic7xxx/0
1161 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1163 TCQ Enabled By Default : Disabled
1164 AIC7XXX_PROC_STATS : Disabled
1165 AIC7XXX_RESET_DELAY : 5
1166 Adapter Configuration:
1167 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1168 Ultra Wide Controller
1169 PCI MMAPed I/O Base: 0xeb001000
1170 Adapter SEEPROM Config: SEEPROM found and used.
1171 Adaptec SCSI BIOS: Enabled
1173 SCBs: Active 0, Max Active 2,
1174 Allocated 15, HW 16, Page 255
1176 BIOS Control Word: 0x18b6
1177 Adapter Control Word: 0x005b
1178 Extended Translation: Enabled
1179 Disconnect Enable Flags: 0xffff
1180 Ultra Enable Flags: 0x0001
1181 Tag Queue Enable Flags: 0x0000
1182 Ordered Queue Tag Flags: 0x0000
1183 Default Tag Queue Depth: 8
1184 Tagged Queue By Device array for aic7xxx host instance 0:
1185 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1186 Actual queue depth per device for aic7xxx host instance 0:
1187 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1190 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1191 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1192 Total transfers 160151 (74577 reads and 85574 writes)
1194 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1195 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1196 Total transfers 0 (0 reads and 0 writes)
1199 1.6 Parallel port info in /proc/parport
1200 ---------------------------------------
1202 The directory /proc/parport contains information about the parallel ports of
1203 your system. It has one subdirectory for each port, named after the port
1206 These directories contain the four files shown in Table 1-10.
1209 Table 1-10: Files in /proc/parport
1210 ..............................................................................
1212 autoprobe Any IEEE-1284 device ID information that has been acquired.
1213 devices list of the device drivers using that port. A + will appear by the
1214 name of the device currently using the port (it might not appear
1216 hardware Parallel port's base address, IRQ line and DMA channel.
1217 irq IRQ that parport is using for that port. This is in a separate
1218 file to allow you to alter it by writing a new value in (IRQ
1220 ..............................................................................
1222 1.7 TTY info in /proc/tty
1223 -------------------------
1225 Information about the available and actually used tty's can be found in the
1226 directory /proc/tty.You'll find entries for drivers and line disciplines in
1227 this directory, as shown in Table 1-11.
1230 Table 1-11: Files in /proc/tty
1231 ..............................................................................
1233 drivers list of drivers and their usage
1234 ldiscs registered line disciplines
1235 driver/serial usage statistic and status of single tty lines
1236 ..............................................................................
1238 To see which tty's are currently in use, you can simply look into the file
1241 > cat /proc/tty/drivers
1242 pty_slave /dev/pts 136 0-255 pty:slave
1243 pty_master /dev/ptm 128 0-255 pty:master
1244 pty_slave /dev/ttyp 3 0-255 pty:slave
1245 pty_master /dev/pty 2 0-255 pty:master
1246 serial /dev/cua 5 64-67 serial:callout
1247 serial /dev/ttyS 4 64-67 serial
1248 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1249 /dev/ptmx /dev/ptmx 5 2 system
1250 /dev/console /dev/console 5 1 system:console
1251 /dev/tty /dev/tty 5 0 system:/dev/tty
1252 unknown /dev/tty 4 1-63 console
1255 1.8 Miscellaneous kernel statistics in /proc/stat
1256 -------------------------------------------------
1258 Various pieces of information about kernel activity are available in the
1259 /proc/stat file. All of the numbers reported in this file are aggregates
1260 since the system first booted. For a quick look, simply cat the file:
1263 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1264 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1265 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1266 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1272 softirq 183433 0 21755 12 39 1137 231 21459 2263
1274 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1275 lines. These numbers identify the amount of time the CPU has spent performing
1276 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1277 second). The meanings of the columns are as follows, from left to right:
1279 - user: normal processes executing in user mode
1280 - nice: niced processes executing in user mode
1281 - system: processes executing in kernel mode
1282 - idle: twiddling thumbs
1283 - iowait: waiting for I/O to complete
1284 - irq: servicing interrupts
1285 - softirq: servicing softirqs
1286 - steal: involuntary wait
1287 - guest: running a normal guest
1288 - guest_nice: running a niced guest
1290 The "intr" line gives counts of interrupts serviced since boot time, for each
1291 of the possible system interrupts. The first column is the total of all
1292 interrupts serviced including unnumbered architecture specific interrupts;
1293 each subsequent column is the total for that particular numbered interrupt.
1294 Unnumbered interrupts are not shown, only summed into the total.
1296 The "ctxt" line gives the total number of context switches across all CPUs.
1298 The "btime" line gives the time at which the system booted, in seconds since
1301 The "processes" line gives the number of processes and threads created, which
1302 includes (but is not limited to) those created by calls to the fork() and
1303 clone() system calls.
1305 The "procs_running" line gives the total number of threads that are
1306 running or ready to run (i.e., the total number of runnable threads).
1308 The "procs_blocked" line gives the number of processes currently blocked,
1309 waiting for I/O to complete.
1311 The "softirq" line gives counts of softirqs serviced since boot time, for each
1312 of the possible system softirqs. The first column is the total of all
1313 softirqs serviced; each subsequent column is the total for that particular
1317 1.9 Ext4 file system parameters
1318 -------------------------------
1320 Information about mounted ext4 file systems can be found in
1321 /proc/fs/ext4. Each mounted filesystem will have a directory in
1322 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1323 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1324 in Table 1-12, below.
1326 Table 1-12: Files in /proc/fs/ext4/<devname>
1327 ..............................................................................
1329 mb_groups details of multiblock allocator buddy cache of free blocks
1330 ..............................................................................
1334 Shows registered system console lines.
1336 To see which character device lines are currently used for the system console
1337 /dev/console, you may simply look into the file /proc/consoles:
1339 > cat /proc/consoles
1345 device name of the device
1346 operations R = can do read operations
1347 W = can do write operations
1349 flags E = it is enabled
1350 C = it is preferred console
1351 B = it is primary boot console
1352 p = it is used for printk buffer
1353 b = it is not a TTY but a Braille device
1354 a = it is safe to use when cpu is offline
1355 major:minor major and minor number of the device separated by a colon
1357 ------------------------------------------------------------------------------
1359 ------------------------------------------------------------------------------
1360 The /proc file system serves information about the running system. It not only
1361 allows access to process data but also allows you to request the kernel status
1362 by reading files in the hierarchy.
1364 The directory structure of /proc reflects the types of information and makes
1365 it easy, if not obvious, where to look for specific data.
1366 ------------------------------------------------------------------------------
1368 ------------------------------------------------------------------------------
1369 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1370 ------------------------------------------------------------------------------
1372 ------------------------------------------------------------------------------
1374 ------------------------------------------------------------------------------
1375 * Modifying kernel parameters by writing into files found in /proc/sys
1376 * Exploring the files which modify certain parameters
1377 * Review of the /proc/sys file tree
1378 ------------------------------------------------------------------------------
1381 A very interesting part of /proc is the directory /proc/sys. This is not only
1382 a source of information, it also allows you to change parameters within the
1383 kernel. Be very careful when attempting this. You can optimize your system,
1384 but you can also cause it to crash. Never alter kernel parameters on a
1385 production system. Set up a development machine and test to make sure that
1386 everything works the way you want it to. You may have no alternative but to
1387 reboot the machine once an error has been made.
1389 To change a value, simply echo the new value into the file. An example is
1390 given below in the section on the file system data. You need to be root to do
1391 this. You can create your own boot script to perform this every time your
1394 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1395 general things in the operation of the Linux kernel. Since some of the files
1396 can inadvertently disrupt your system, it is advisable to read both
1397 documentation and source before actually making adjustments. In any case, be
1398 very careful when writing to any of these files. The entries in /proc may
1399 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1400 review the kernel documentation in the directory /usr/src/linux/Documentation.
1401 This chapter is heavily based on the documentation included in the pre 2.2
1402 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1404 Please see: Documentation/sysctl/ directory for descriptions of these
1407 ------------------------------------------------------------------------------
1409 ------------------------------------------------------------------------------
1410 Certain aspects of kernel behavior can be modified at runtime, without the
1411 need to recompile the kernel, or even to reboot the system. The files in the
1412 /proc/sys tree can not only be read, but also modified. You can use the echo
1413 command to write value into these files, thereby changing the default settings
1415 ------------------------------------------------------------------------------
1417 ------------------------------------------------------------------------------
1418 CHAPTER 3: PER-PROCESS PARAMETERS
1419 ------------------------------------------------------------------------------
1421 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1422 --------------------------------------------------------------------------------
1424 These file can be used to adjust the badness heuristic used to select which
1425 process gets killed in out of memory conditions.
1427 The badness heuristic assigns a value to each candidate task ranging from 0
1428 (never kill) to 1000 (always kill) to determine which process is targeted. The
1429 units are roughly a proportion along that range of allowed memory the process
1430 may allocate from based on an estimation of its current memory and swap use.
1431 For example, if a task is using all allowed memory, its badness score will be
1432 1000. If it is using half of its allowed memory, its score will be 500.
1434 There is an additional factor included in the badness score: the current memory
1435 and swap usage is discounted by 3% for root processes.
1437 The amount of "allowed" memory depends on the context in which the oom killer
1438 was called. If it is due to the memory assigned to the allocating task's cpuset
1439 being exhausted, the allowed memory represents the set of mems assigned to that
1440 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1441 memory represents the set of mempolicy nodes. If it is due to a memory
1442 limit (or swap limit) being reached, the allowed memory is that configured
1443 limit. Finally, if it is due to the entire system being out of memory, the
1444 allowed memory represents all allocatable resources.
1446 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1447 is used to determine which task to kill. Acceptable values range from -1000
1448 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1449 polarize the preference for oom killing either by always preferring a certain
1450 task or completely disabling it. The lowest possible value, -1000, is
1451 equivalent to disabling oom killing entirely for that task since it will always
1452 report a badness score of 0.
1454 Consequently, it is very simple for userspace to define the amount of memory to
1455 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1456 example, is roughly equivalent to allowing the remainder of tasks sharing the
1457 same system, cpuset, mempolicy, or memory controller resources to use at least
1458 50% more memory. A value of -500, on the other hand, would be roughly
1459 equivalent to discounting 50% of the task's allowed memory from being considered
1460 as scoring against the task.
1462 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1463 be used to tune the badness score. Its acceptable values range from -16
1464 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1465 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1466 scaled linearly with /proc/<pid>/oom_score_adj.
1468 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1469 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1470 requires CAP_SYS_RESOURCE.
1472 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1473 generation children with separate address spaces instead, if possible. This
1474 avoids servers and important system daemons from being killed and loses the
1475 minimal amount of work.
1478 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1479 -------------------------------------------------------------
1481 This file can be used to check the current score used by the oom-killer is for
1482 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1483 process should be killed in an out-of-memory situation.
1486 3.3 /proc/<pid>/io - Display the IO accounting fields
1487 -------------------------------------------------------
1489 This file contains IO statistics for each running process
1494 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1497 test:/tmp # cat /proc/3828/io
1503 write_bytes: 323932160
1504 cancelled_write_bytes: 0
1513 I/O counter: chars read
1514 The number of bytes which this task has caused to be read from storage. This
1515 is simply the sum of bytes which this process passed to read() and pread().
1516 It includes things like tty IO and it is unaffected by whether or not actual
1517 physical disk IO was required (the read might have been satisfied from
1524 I/O counter: chars written
1525 The number of bytes which this task has caused, or shall cause to be written
1526 to disk. Similar caveats apply here as with rchar.
1532 I/O counter: read syscalls
1533 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1540 I/O counter: write syscalls
1541 Attempt to count the number of write I/O operations, i.e. syscalls like
1542 write() and pwrite().
1548 I/O counter: bytes read
1549 Attempt to count the number of bytes which this process really did cause to
1550 be fetched from the storage layer. Done at the submit_bio() level, so it is
1551 accurate for block-backed filesystems. <please add status regarding NFS and
1552 CIFS at a later time>
1558 I/O counter: bytes written
1559 Attempt to count the number of bytes which this process caused to be sent to
1560 the storage layer. This is done at page-dirtying time.
1563 cancelled_write_bytes
1564 ---------------------
1566 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1567 then deletes the file, it will in fact perform no writeout. But it will have
1568 been accounted as having caused 1MB of write.
1569 In other words: The number of bytes which this process caused to not happen,
1570 by truncating pagecache. A task can cause "negative" IO too. If this task
1571 truncates some dirty pagecache, some IO which another task has been accounted
1572 for (in its write_bytes) will not be happening. We _could_ just subtract that
1573 from the truncating task's write_bytes, but there is information loss in doing
1580 At its current implementation state, this is a bit racy on 32-bit machines: if
1581 process A reads process B's /proc/pid/io while process B is updating one of
1582 those 64-bit counters, process A could see an intermediate result.
1585 More information about this can be found within the taskstats documentation in
1586 Documentation/accounting.
1588 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1589 ---------------------------------------------------------------
1590 When a process is dumped, all anonymous memory is written to a core file as
1591 long as the size of the core file isn't limited. But sometimes we don't want
1592 to dump some memory segments, for example, huge shared memory. Conversely,
1593 sometimes we want to save file-backed memory segments into a core file, not
1594 only the individual files.
1596 /proc/<pid>/coredump_filter allows you to customize which memory segments
1597 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1598 of memory types. If a bit of the bitmask is set, memory segments of the
1599 corresponding memory type are dumped, otherwise they are not dumped.
1601 The following 7 memory types are supported:
1602 - (bit 0) anonymous private memory
1603 - (bit 1) anonymous shared memory
1604 - (bit 2) file-backed private memory
1605 - (bit 3) file-backed shared memory
1606 - (bit 4) ELF header pages in file-backed private memory areas (it is
1607 effective only if the bit 2 is cleared)
1608 - (bit 5) hugetlb private memory
1609 - (bit 6) hugetlb shared memory
1611 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1612 are always dumped regardless of the bitmask status.
1614 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1615 effected by bit 5-6.
1617 Default value of coredump_filter is 0x23; this means all anonymous memory
1618 segments and hugetlb private memory are dumped.
1620 If you don't want to dump all shared memory segments attached to pid 1234,
1621 write 0x21 to the process's proc file.
1623 $ echo 0x21 > /proc/1234/coredump_filter
1625 When a new process is created, the process inherits the bitmask status from its
1626 parent. It is useful to set up coredump_filter before the program runs.
1629 $ echo 0x7 > /proc/self/coredump_filter
1632 3.5 /proc/<pid>/mountinfo - Information about mounts
1633 --------------------------------------------------------
1635 This file contains lines of the form:
1637 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1638 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1640 (1) mount ID: unique identifier of the mount (may be reused after umount)
1641 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1642 (3) major:minor: value of st_dev for files on filesystem
1643 (4) root: root of the mount within the filesystem
1644 (5) mount point: mount point relative to the process's root
1645 (6) mount options: per mount options
1646 (7) optional fields: zero or more fields of the form "tag[:value]"
1647 (8) separator: marks the end of the optional fields
1648 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1649 (10) mount source: filesystem specific information or "none"
1650 (11) super options: per super block options
1652 Parsers should ignore all unrecognised optional fields. Currently the
1653 possible optional fields are:
1655 shared:X mount is shared in peer group X
1656 master:X mount is slave to peer group X
1657 propagate_from:X mount is slave and receives propagation from peer group X (*)
1658 unbindable mount is unbindable
1660 (*) X is the closest dominant peer group under the process's root. If
1661 X is the immediate master of the mount, or if there's no dominant peer
1662 group under the same root, then only the "master:X" field is present
1663 and not the "propagate_from:X" field.
1665 For more information on mount propagation see:
1667 Documentation/filesystems/sharedsubtree.txt
1670 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1671 --------------------------------------------------------
1672 These files provide a method to access a tasks comm value. It also allows for
1673 a task to set its own or one of its thread siblings comm value. The comm value
1674 is limited in size compared to the cmdline value, so writing anything longer
1675 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1679 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1680 -------------------------------------------------------------------------
1681 This file provides a fast way to retrieve first level children pids
1682 of a task pointed by <pid>/<tid> pair. The format is a space separated
1685 Note the "first level" here -- if a child has own children they will
1686 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1687 to obtain the descendants.
1689 Since this interface is intended to be fast and cheap it doesn't
1690 guarantee to provide precise results and some children might be
1691 skipped, especially if they've exited right after we printed their
1692 pids, so one need to either stop or freeze processes being inspected
1693 if precise results are needed.
1696 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1697 ---------------------------------------------------------------
1698 This file provides information associated with an opened file. The regular
1699 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1700 represents the current offset of the opened file in decimal form [see lseek(2)
1701 for details], 'flags' denotes the octal O_xxx mask the file has been
1702 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1703 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1712 All locks associated with a file descriptor are shown in its fdinfo too.
1714 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1716 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1717 pair provide additional information particular to the objects they represent.
1726 where 'eventfd-count' is hex value of a counter.
1733 sigmask: 0000000000000200
1735 where 'sigmask' is hex value of the signal mask associated
1743 tfd: 5 events: 1d data: ffffffffffffffff
1745 where 'tfd' is a target file descriptor number in decimal form,
1746 'events' is events mask being watched and the 'data' is data
1747 associated with a target [see epoll(7) for more details].
1751 For inotify files the format is the following
1755 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1757 where 'wd' is a watch descriptor in decimal form, ie a target file
1758 descriptor number, 'ino' and 'sdev' are inode and device where the
1759 target file resides and the 'mask' is the mask of events, all in hex
1760 form [see inotify(7) for more details].
1762 If the kernel was built with exportfs support, the path to the target
1763 file is encoded as a file handle. The file handle is provided by three
1764 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1767 If the kernel is built without exportfs support the file handle won't be
1770 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1772 For fanotify files the format is
1777 fanotify flags:10 event-flags:0
1778 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1779 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1781 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1782 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1783 flags associated with mark which are tracked separately from events
1784 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1785 mask and 'ignored_mask' is the mask of events which are to be ignored.
1786 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1787 does provide information about flags and mask used in fanotify_mark
1788 call [see fsnotify manpage for details].
1790 While the first three lines are mandatory and always printed, the rest is
1791 optional and may be omitted if no marks created yet.
1802 it_value: (0, 49406829)
1805 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1806 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1807 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1808 details]. 'it_value' is remaining time until the timer exiration.
1809 'it_interval' is the interval for the timer. Note the timer might be set up
1810 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1811 still exhibits timer's remaining time.
1813 3.9 /proc/<pid>/map_files - Information about memory mapped files
1814 ---------------------------------------------------------------------
1815 This directory contains symbolic links which represent memory mapped files
1816 the process is maintaining. Example output:
1818 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1819 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1820 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1822 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1823 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1825 The name of a link represents the virtual memory bounds of a mapping, i.e.
1826 vm_area_struct::vm_start-vm_area_struct::vm_end.
1828 The main purpose of the map_files is to retrieve a set of memory mapped
1829 files in a fast way instead of parsing /proc/<pid>/maps or
1830 /proc/<pid>/smaps, both of which contain many more records. At the same
1831 time one can open(2) mappings from the listings of two processes and
1832 comparing their inode numbers to figure out which anonymous memory areas
1833 are actually shared.
1835 ------------------------------------------------------------------------------
1837 ------------------------------------------------------------------------------
1840 ---------------------
1842 The following mount options are supported:
1844 hidepid= Set /proc/<pid>/ access mode.
1845 gid= Set the group authorized to learn processes information.
1847 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1850 hidepid=1 means users may not access any /proc/<pid>/ directories but their
1851 own. Sensitive files like cmdline, sched*, status are now protected against
1852 other users. This makes it impossible to learn whether any user runs
1853 specific program (given the program doesn't reveal itself by its behaviour).
1854 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1855 poorly written programs passing sensitive information via program arguments are
1856 now protected against local eavesdroppers.
1858 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1859 users. It doesn't mean that it hides a fact whether a process with a specific
1860 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1861 but it hides process' uid and gid, which may be learned by stat()'ing
1862 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1863 information about running processes, whether some daemon runs with elevated
1864 privileges, whether other user runs some sensitive program, whether other users
1865 run any program at all, etc.
1867 gid= defines a group authorized to learn processes information otherwise
1868 prohibited by hidepid=. If you use some daemon like identd which needs to learn
1869 information about processes information, just add identd to this group.
Code Review / kvmfornfv.git / blob