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 Present with CONFIG_KALLSYMS=y: it shows the kernel function
144 symbol the task is blocked in - or "0" if not blocked.
146 stack Report full stack trace, enable via CONFIG_STACKTRACE
147 smaps a extension based on maps, showing the memory consumption of
148 each mapping and flags associated with it
149 numa_maps an extension based on maps, showing the memory locality and
150 binding policy as well as mem usage (in pages) of each mapping.
151 ..............................................................................
153 For example, to get the status information of a process, all you have to do is
154 read the file /proc/PID/status:
156 >cat /proc/self/status
180 SigPnd: 0000000000000000
181 ShdPnd: 0000000000000000
182 SigBlk: 0000000000000000
183 SigIgn: 0000000000000000
184 SigCgt: 0000000000000000
185 CapInh: 00000000fffffeff
186 CapPrm: 0000000000000000
187 CapEff: 0000000000000000
188 CapBnd: ffffffffffffffff
190 voluntary_ctxt_switches: 0
191 nonvoluntary_ctxt_switches: 1
193 This shows you nearly the same information you would get if you viewed it with
194 the ps command. In fact, ps uses the proc file system to obtain its
195 information. But you get a more detailed view of the process by reading the
196 file /proc/PID/status. It fields are described in table 1-2.
198 The statm file contains more detailed information about the process
199 memory usage. Its seven fields are explained in Table 1-3. The stat file
200 contains details information about the process itself. Its fields are
201 explained in Table 1-4.
203 (for SMP CONFIG users)
204 For making accounting scalable, RSS related information are handled in an
205 asynchronous manner and the value may not be very precise. To see a precise
206 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
207 It's slow but very precise.
209 Table 1-2: Contents of the status files (as of 4.1)
210 ..............................................................................
212 Name filename of the executable
213 State state (R is running, S is sleeping, D is sleeping
214 in an uninterruptible wait, Z is zombie,
215 T is traced or stopped)
217 Ngid NUMA group ID (0 if none)
219 PPid process id of the parent process
220 TracerPid PID of process tracing this process (0 if not)
221 Uid Real, effective, saved set, and file system UIDs
222 Gid Real, effective, saved set, and file system GIDs
223 FDSize number of file descriptor slots currently allocated
224 Groups supplementary group list
225 NStgid descendant namespace thread group ID hierarchy
226 NSpid descendant namespace process ID hierarchy
227 NSpgid descendant namespace process group ID hierarchy
228 NSsid descendant namespace session ID hierarchy
229 VmPeak peak virtual memory size
230 VmSize total program size
231 VmLck locked memory size
232 VmHWM peak resident set size ("high water mark")
233 VmRSS size of memory portions
234 VmData size of data, stack, and text segments
235 VmStk size of data, stack, and text segments
236 VmExe size of text segment
237 VmLib size of shared library code
238 VmPTE size of page table entries
239 VmPMD size of second level page tables
240 VmSwap size of swap usage (the number of referred swapents)
241 Threads number of threads
242 SigQ number of signals queued/max. number for queue
243 SigPnd bitmap of pending signals for the thread
244 ShdPnd bitmap of shared pending signals for the process
245 SigBlk bitmap of blocked signals
246 SigIgn bitmap of ignored signals
247 SigCgt bitmap of caught signals
248 CapInh bitmap of inheritable capabilities
249 CapPrm bitmap of permitted capabilities
250 CapEff bitmap of effective capabilities
251 CapBnd bitmap of capabilities bounding set
252 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
253 Cpus_allowed mask of CPUs on which this process may run
254 Cpus_allowed_list Same as previous, but in "list format"
255 Mems_allowed mask of memory nodes allowed to this process
256 Mems_allowed_list Same as previous, but in "list format"
257 voluntary_ctxt_switches number of voluntary context switches
258 nonvoluntary_ctxt_switches number of non voluntary context switches
259 ..............................................................................
261 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
262 ..............................................................................
264 size total program size (pages) (same as VmSize in status)
265 resident size of memory portions (pages) (same as VmRSS in status)
266 shared number of pages that are shared (i.e. backed by a file)
267 trs number of pages that are 'code' (not including libs; broken,
268 includes data segment)
269 lrs number of pages of library (always 0 on 2.6)
270 drs number of pages of data/stack (including libs; broken,
271 includes library text)
272 dt number of dirty pages (always 0 on 2.6)
273 ..............................................................................
276 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
277 ..............................................................................
280 tcomm filename of the executable
281 state state (R is running, S is sleeping, D is sleeping in an
282 uninterruptible wait, Z is zombie, T is traced or stopped)
283 ppid process id of the parent process
284 pgrp pgrp of the process
286 tty_nr tty the process uses
287 tty_pgrp pgrp of the tty
289 min_flt number of minor faults
290 cmin_flt number of minor faults with child's
291 maj_flt number of major faults
292 cmaj_flt number of major faults with child's
293 utime user mode jiffies
294 stime kernel mode jiffies
295 cutime user mode jiffies with child's
296 cstime kernel mode jiffies with child's
297 priority priority level
299 num_threads number of threads
300 it_real_value (obsolete, always 0)
301 start_time time the process started after system boot
302 vsize virtual memory size
303 rss resident set memory size
304 rsslim current limit in bytes on the rss
305 start_code address above which program text can run
306 end_code address below which program text can run
307 start_stack address of the start of the main process stack
308 esp current value of ESP
309 eip current value of EIP
310 pending bitmap of pending signals
311 blocked bitmap of blocked signals
312 sigign bitmap of ignored signals
313 sigcatch bitmap of caught signals
314 0 (place holder, used to be the wchan address, use /proc/PID/wchan instead)
317 exit_signal signal to send to parent thread on exit
318 task_cpu which CPU the task is scheduled on
319 rt_priority realtime priority
320 policy scheduling policy (man sched_setscheduler)
321 blkio_ticks time spent waiting for block IO
322 gtime guest time of the task in jiffies
323 cgtime guest time of the task children in jiffies
324 start_data address above which program data+bss is placed
325 end_data address below which program data+bss is placed
326 start_brk address above which program heap can be expanded with brk()
327 arg_start address above which program command line is placed
328 arg_end address below which program command line is placed
329 env_start address above which program environment is placed
330 env_end address below which program environment is placed
331 exit_code the thread's exit_code in the form reported by the waitpid system call
332 ..............................................................................
334 The /proc/PID/maps file containing the currently mapped memory regions and
335 their access permissions.
339 address perms offset dev inode pathname
341 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
342 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
343 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
344 a7cb1000-a7cb2000 ---p 00000000 00:00 0
345 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
346 a7eb2000-a7eb3000 ---p 00000000 00:00 0
347 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
348 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
349 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
350 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
351 a800b000-a800e000 rw-p 00000000 00:00 0
352 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
353 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
354 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
355 a8024000-a8027000 rw-p 00000000 00:00 0
356 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
357 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
358 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
359 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
360 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
362 where "address" is the address space in the process that it occupies, "perms"
363 is a set of permissions:
369 p = private (copy on write)
371 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
372 "inode" is the inode on that device. 0 indicates that no inode is associated
373 with the memory region, as the case would be with BSS (uninitialized data).
374 The "pathname" shows the name associated file for this mapping. If the mapping
375 is not associated with a file:
377 [heap] = the heap of the program
378 [stack] = the stack of the main process
379 [stack:1001] = the stack of the thread with tid 1001
380 [vdso] = the "virtual dynamic shared object",
381 the kernel system call handler
383 or if empty, the mapping is anonymous.
385 The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
386 of the individual tasks of a process. In this file you will see a mapping marked
387 as [stack] if that task sees it as a stack. This is a key difference from the
388 content of /proc/PID/maps, where you will see all mappings that are being used
389 as stack by all of those tasks. Hence, for the example above, the task-level
390 map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
392 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
393 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
394 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
395 a7cb1000-a7cb2000 ---p 00000000 00:00 0
396 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
397 a7eb2000-a7eb3000 ---p 00000000 00:00 0
398 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
399 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
400 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
401 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
402 a800b000-a800e000 rw-p 00000000 00:00 0
403 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
404 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
405 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
406 a8024000-a8027000 rw-p 00000000 00:00 0
407 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
408 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
409 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
410 aff35000-aff4a000 rw-p 00000000 00:00 0
411 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
413 The /proc/PID/smaps is an extension based on maps, showing the memory
414 consumption for each of the process's mappings. For each of mappings there
415 is a series of lines such as the following:
417 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
429 Private_Hugetlb: 0 kB
435 VmFlags: rd ex mr mw me de
437 the first of these lines shows the same information as is displayed for the
438 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
439 (size), the amount of the mapping that is currently resident in RAM (RSS), the
440 process' proportional share of this mapping (PSS), the number of clean and
441 dirty private pages in the mapping.
443 The "proportional set size" (PSS) of a process is the count of pages it has
444 in memory, where each page is divided by the number of processes sharing it.
445 So if a process has 1000 pages all to itself, and 1000 shared with one other
446 process, its PSS will be 1500.
447 Note that even a page which is part of a MAP_SHARED mapping, but has only
448 a single pte mapped, i.e. is currently used by only one process, is accounted
449 as private and not as shared.
450 "Referenced" indicates the amount of memory currently marked as referenced or
452 "Anonymous" shows the amount of memory that does not belong to any file. Even
453 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
454 and a page is modified, the file page is replaced by a private anonymous copy.
455 "Swap" shows how much would-be-anonymous memory is also used, but out on
457 "SwapPss" shows proportional swap share of this mapping.
458 "AnonHugePages" shows the ammount of memory backed by transparent hugepage.
459 "Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by
460 hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
461 reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
463 "VmFlags" field deserves a separate description. This member represents the kernel
464 flags associated with the particular virtual memory area in two letter encoded
465 manner. The codes are the following:
474 gd - stack segment growns down
476 dw - disabled write to the mapped file
477 lo - pages are locked in memory
478 io - memory mapped I/O area
479 sr - sequential read advise provided
480 rr - random read advise provided
481 dc - do not copy area on fork
482 de - do not expand area on remapping
483 ac - area is accountable
484 nr - swap space is not reserved for the area
485 ht - area uses huge tlb pages
486 nl - non-linear mapping
487 ar - architecture specific flag
488 dd - do not include area into core dump
491 hg - huge page advise flag
492 nh - no-huge page advise flag
493 mg - mergable advise flag
495 Note that there is no guarantee that every flag and associated mnemonic will
496 be present in all further kernel releases. Things get changed, the flags may
497 be vanished or the reverse -- new added.
499 This file is only present if the CONFIG_MMU kernel configuration option is
502 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
503 bits on both physical and virtual pages associated with a process, and the
504 soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
505 To clear the bits for all the pages associated with the process
506 > echo 1 > /proc/PID/clear_refs
508 To clear the bits for the anonymous pages associated with the process
509 > echo 2 > /proc/PID/clear_refs
511 To clear the bits for the file mapped pages associated with the process
512 > echo 3 > /proc/PID/clear_refs
514 To clear the soft-dirty bit
515 > echo 4 > /proc/PID/clear_refs
517 To reset the peak resident set size ("high water mark") to the process's
519 > echo 5 > /proc/PID/clear_refs
521 Any other value written to /proc/PID/clear_refs will have no effect.
523 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
524 using /proc/kpageflags and number of times a page is mapped using
525 /proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
527 The /proc/pid/numa_maps is an extension based on maps, showing the memory
528 locality and binding policy, as well as the memory usage (in pages) of
529 each mapping. The output follows a general format where mapping details get
530 summarized separated by blank spaces, one mapping per each file line:
532 address policy mapping details
534 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
535 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
536 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
537 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
538 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
539 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
540 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
541 320698b000 default file=/lib64/libc-2.12.so
542 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
543 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
544 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
545 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
546 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
547 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
548 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
549 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
552 "address" is the starting address for the mapping;
553 "policy" reports the NUMA memory policy set for the mapping (see vm/numa_memory_policy.txt);
554 "mapping details" summarizes mapping data such as mapping type, page usage counters,
555 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
556 size, in KB, that is backing the mapping up.
561 Similar to the process entries, the kernel data files give information about
562 the running kernel. The files used to obtain this information are contained in
563 /proc and are listed in Table 1-5. Not all of these will be present in your
564 system. It depends on the kernel configuration and the loaded modules, which
565 files are there, and which are missing.
567 Table 1-5: Kernel info in /proc
568 ..............................................................................
570 apm Advanced power management info
571 buddyinfo Kernel memory allocator information (see text) (2.5)
572 bus Directory containing bus specific information
573 cmdline Kernel command line
574 cpuinfo Info about the CPU
575 devices Available devices (block and character)
576 dma Used DMS channels
577 filesystems Supported filesystems
578 driver Various drivers grouped here, currently rtc (2.4)
579 execdomains Execdomains, related to security (2.4)
580 fb Frame Buffer devices (2.4)
581 fs File system parameters, currently nfs/exports (2.4)
582 ide Directory containing info about the IDE subsystem
583 interrupts Interrupt usage
584 iomem Memory map (2.4)
585 ioports I/O port usage
586 irq Masks for irq to cpu affinity (2.4)(smp?)
587 isapnp ISA PnP (Plug&Play) Info (2.4)
588 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
590 ksyms Kernel symbol table
591 loadavg Load average of last 1, 5 & 15 minutes
595 modules List of loaded modules
596 mounts Mounted filesystems
597 net Networking info (see text)
598 pagetypeinfo Additional page allocator information (see text) (2.5)
599 partitions Table of partitions known to the system
600 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
601 decoupled by lspci (2.4)
603 scsi SCSI info (see text)
604 slabinfo Slab pool info
605 softirqs softirq usage
606 stat Overall statistics
607 swaps Swap space utilization
609 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
610 tty Info of tty drivers
611 uptime Wall clock since boot, combined idle time of all cpus
612 version Kernel version
613 video bttv info of video resources (2.4)
614 vmallocinfo Show vmalloced areas
615 ..............................................................................
617 You can, for example, check which interrupts are currently in use and what
618 they are used for by looking in the file /proc/interrupts:
620 > cat /proc/interrupts
622 0: 8728810 XT-PIC timer
623 1: 895 XT-PIC keyboard
625 3: 531695 XT-PIC aha152x
626 4: 2014133 XT-PIC serial
627 5: 44401 XT-PIC pcnet_cs
630 12: 182918 XT-PIC PS/2 Mouse
632 14: 1232265 XT-PIC ide0
636 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
637 output of a SMP machine):
639 > cat /proc/interrupts
642 0: 1243498 1214548 IO-APIC-edge timer
643 1: 8949 8958 IO-APIC-edge keyboard
644 2: 0 0 XT-PIC cascade
645 5: 11286 10161 IO-APIC-edge soundblaster
646 8: 1 0 IO-APIC-edge rtc
647 9: 27422 27407 IO-APIC-edge 3c503
648 12: 113645 113873 IO-APIC-edge PS/2 Mouse
650 14: 22491 24012 IO-APIC-edge ide0
651 15: 2183 2415 IO-APIC-edge ide1
652 17: 30564 30414 IO-APIC-level eth0
653 18: 177 164 IO-APIC-level bttv
658 NMI is incremented in this case because every timer interrupt generates a NMI
659 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
661 LOC is the local interrupt counter of the internal APIC of every CPU.
663 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
664 connects the CPUs in a SMP system. This means that an error has been detected,
665 the IO-APIC automatically retry the transmission, so it should not be a big
666 problem, but you should read the SMP-FAQ.
668 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
669 /proc/interrupts to display every IRQ vector in use by the system, not
670 just those considered 'most important'. The new vectors are:
672 THR -- interrupt raised when a machine check threshold counter
673 (typically counting ECC corrected errors of memory or cache) exceeds
674 a configurable threshold. Only available on some systems.
676 TRM -- a thermal event interrupt occurs when a temperature threshold
677 has been exceeded for the CPU. This interrupt may also be generated
678 when the temperature drops back to normal.
680 SPU -- a spurious interrupt is some interrupt that was raised then lowered
681 by some IO device before it could be fully processed by the APIC. Hence
682 the APIC sees the interrupt but does not know what device it came from.
683 For this case the APIC will generate the interrupt with a IRQ vector
684 of 0xff. This might also be generated by chipset bugs.
686 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
687 sent from one CPU to another per the needs of the OS. Typically,
688 their statistics are used by kernel developers and interested users to
689 determine the occurrence of interrupts of the given type.
691 The above IRQ vectors are displayed only when relevant. For example,
692 the threshold vector does not exist on x86_64 platforms. Others are
693 suppressed when the system is a uniprocessor. As of this writing, only
694 i386 and x86_64 platforms support the new IRQ vector displays.
696 Of some interest is the introduction of the /proc/irq directory to 2.4.
697 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
698 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
699 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
704 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
705 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
709 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
710 IRQ, you can set it by doing:
712 > echo 1 > /proc/irq/10/smp_affinity
714 This means that only the first CPU will handle the IRQ, but you can also echo
715 5 which means that only the first and fourth CPU can handle the IRQ.
717 The contents of each smp_affinity file is the same by default:
719 > cat /proc/irq/0/smp_affinity
722 There is an alternate interface, smp_affinity_list which allows specifying
723 a cpu range instead of a bitmask:
725 > cat /proc/irq/0/smp_affinity_list
728 The default_smp_affinity mask applies to all non-active IRQs, which are the
729 IRQs which have not yet been allocated/activated, and hence which lack a
730 /proc/irq/[0-9]* directory.
732 The node file on an SMP system shows the node to which the device using the IRQ
733 reports itself as being attached. This hardware locality information does not
734 include information about any possible driver locality preference.
736 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
737 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
739 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
740 between all the CPUs which are allowed to handle it. As usual the kernel has
741 more info than you and does a better job than you, so the defaults are the
742 best choice for almost everyone. [Note this applies only to those IO-APIC's
743 that support "Round Robin" interrupt distribution.]
745 There are three more important subdirectories in /proc: net, scsi, and sys.
746 The general rule is that the contents, or even the existence of these
747 directories, depend on your kernel configuration. If SCSI is not enabled, the
748 directory scsi may not exist. The same is true with the net, which is there
749 only when networking support is present in the running kernel.
751 The slabinfo file gives information about memory usage at the slab level.
752 Linux uses slab pools for memory management above page level in version 2.2.
753 Commonly used objects have their own slab pool (such as network buffers,
754 directory cache, and so on).
756 ..............................................................................
758 > cat /proc/buddyinfo
760 Node 0, zone DMA 0 4 5 4 4 3 ...
761 Node 0, zone Normal 1 0 0 1 101 8 ...
762 Node 0, zone HighMem 2 0 0 1 1 0 ...
764 External fragmentation is a problem under some workloads, and buddyinfo is a
765 useful tool for helping diagnose these problems. Buddyinfo will give you a
766 clue as to how big an area you can safely allocate, or why a previous
769 Each column represents the number of pages of a certain order which are
770 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
771 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
772 available in ZONE_NORMAL, etc...
774 More information relevant to external fragmentation can be found in
777 > cat /proc/pagetypeinfo
781 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
782 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
783 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
784 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
785 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
786 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
787 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
788 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
789 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
790 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
791 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
793 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
794 Node 0, zone DMA 2 0 5 1 0
795 Node 0, zone DMA32 41 6 967 2 0
797 Fragmentation avoidance in the kernel works by grouping pages of different
798 migrate types into the same contiguous regions of memory called page blocks.
799 A page block is typically the size of the default hugepage size e.g. 2MB on
800 X86-64. By keeping pages grouped based on their ability to move, the kernel
801 can reclaim pages within a page block to satisfy a high-order allocation.
803 The pagetypinfo begins with information on the size of a page block. It
804 then gives the same type of information as buddyinfo except broken down
805 by migrate-type and finishes with details on how many page blocks of each
808 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
809 from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
810 make an estimate of the likely number of huge pages that can be allocated
811 at a given point in time. All the "Movable" blocks should be allocatable
812 unless memory has been mlock()'d. Some of the Reclaimable blocks should
813 also be allocatable although a lot of filesystem metadata may have to be
814 reclaimed to achieve this.
816 ..............................................................................
820 Provides information about distribution and utilization of memory. This
821 varies by architecture and compile options. The following is from a
822 16GB PIII, which has highmem enabled. You may not have all of these fields.
826 The "Locked" indicates whether the mapping is locked in memory or not.
829 MemTotal: 16344972 kB
831 MemAvailable: 14836172 kB
837 HighTotal: 15597528 kB
838 HighFree: 13629632 kB
848 SReclaimable: 159856 kB
849 SUnreclaim: 124508 kB
854 CommitLimit: 7669796 kB
855 Committed_AS: 100056 kB
856 VmallocTotal: 112216 kB
858 VmallocChunk: 111088 kB
859 AnonHugePages: 49152 kB
861 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
862 bits and the kernel binary code)
863 MemFree: The sum of LowFree+HighFree
864 MemAvailable: An estimate of how much memory is available for starting new
865 applications, without swapping. Calculated from MemFree,
866 SReclaimable, the size of the file LRU lists, and the low
867 watermarks in each zone.
868 The estimate takes into account that the system needs some
869 page cache to function well, and that not all reclaimable
870 slab will be reclaimable, due to items being in use. The
871 impact of those factors will vary from system to system.
872 Buffers: Relatively temporary storage for raw disk blocks
873 shouldn't get tremendously large (20MB or so)
874 Cached: in-memory cache for files read from the disk (the
875 pagecache). Doesn't include SwapCached
876 SwapCached: Memory that once was swapped out, is swapped back in but
877 still also is in the swapfile (if memory is needed it
878 doesn't need to be swapped out AGAIN because it is already
879 in the swapfile. This saves I/O)
880 Active: Memory that has been used more recently and usually not
881 reclaimed unless absolutely necessary.
882 Inactive: Memory which has been less recently used. It is more
883 eligible to be reclaimed for other purposes
885 HighFree: Highmem is all memory above ~860MB of physical memory
886 Highmem areas are for use by userspace programs, or
887 for the pagecache. The kernel must use tricks to access
888 this memory, making it slower to access than lowmem.
890 LowFree: Lowmem is memory which can be used for everything that
891 highmem can be used for, but it is also available for the
892 kernel's use for its own data structures. Among many
893 other things, it is where everything from the Slab is
894 allocated. Bad things happen when you're out of lowmem.
895 SwapTotal: total amount of swap space available
896 SwapFree: Memory which has been evicted from RAM, and is temporarily
898 Dirty: Memory which is waiting to get written back to the disk
899 Writeback: Memory which is actively being written back to the disk
900 AnonPages: Non-file backed pages mapped into userspace page tables
901 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
902 Mapped: files which have been mmaped, such as libraries
903 Slab: in-kernel data structures cache
904 SReclaimable: Part of Slab, that might be reclaimed, such as caches
905 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
906 PageTables: amount of memory dedicated to the lowest level of page
908 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
910 Bounce: Memory used for block device "bounce buffers"
911 WritebackTmp: Memory used by FUSE for temporary writeback buffers
912 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
913 this is the total amount of memory currently available to
914 be allocated on the system. This limit is only adhered to
915 if strict overcommit accounting is enabled (mode 2 in
916 'vm.overcommit_memory').
917 The CommitLimit is calculated with the following formula:
918 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
919 overcommit_ratio / 100 + [total swap pages]
920 For example, on a system with 1G of physical RAM and 7G
921 of swap with a `vm.overcommit_ratio` of 30 it would
922 yield a CommitLimit of 7.3G.
923 For more details, see the memory overcommit documentation
924 in vm/overcommit-accounting.
925 Committed_AS: The amount of memory presently allocated on the system.
926 The committed memory is a sum of all of the memory which
927 has been allocated by processes, even if it has not been
928 "used" by them as of yet. A process which malloc()'s 1G
929 of memory, but only touches 300M of it will show up as
930 using 1G. This 1G is memory which has been "committed" to
931 by the VM and can be used at any time by the allocating
932 application. With strict overcommit enabled on the system
933 (mode 2 in 'vm.overcommit_memory'),allocations which would
934 exceed the CommitLimit (detailed above) will not be permitted.
935 This is useful if one needs to guarantee that processes will
936 not fail due to lack of memory once that memory has been
937 successfully allocated.
938 VmallocTotal: total size of vmalloc memory area
939 VmallocUsed: amount of vmalloc area which is used
940 VmallocChunk: largest contiguous block of vmalloc area which is free
942 ..............................................................................
946 Provides information about vmalloced/vmaped areas. One line per area,
947 containing the virtual address range of the area, size in bytes,
948 caller information of the creator, and optional information depending
949 on the kind of area :
951 pages=nr number of pages
952 phys=addr if a physical address was specified
953 ioremap I/O mapping (ioremap() and friends)
954 vmalloc vmalloc() area
957 vpages buffer for pages pointers was vmalloced (huge area)
958 N<node>=nr (Only on NUMA kernels)
959 Number of pages allocated on memory node <node>
961 > cat /proc/vmallocinfo
962 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
963 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
964 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
965 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
966 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
967 phys=7fee8000 ioremap
968 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
969 phys=7fee7000 ioremap
970 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
971 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
972 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
973 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
975 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
976 /0x130 [x_tables] pages=4 vmalloc N0=4
977 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
978 pages=14 vmalloc N2=14
979 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
981 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
983 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
984 pages=10 vmalloc N0=10
986 ..............................................................................
990 Provides counts of softirq handlers serviced since boot time, for each cpu.
995 TIMER: 27166 27120 27097 27034
1000 SCHED: 27035 26983 26971 26746
1002 RCU: 1678 1769 2178 2250
1005 1.3 IDE devices in /proc/ide
1006 ----------------------------
1008 The subdirectory /proc/ide contains information about all IDE devices of which
1009 the kernel is aware. There is one subdirectory for each IDE controller, the
1010 file drivers and a link for each IDE device, pointing to the device directory
1011 in the controller specific subtree.
1013 The file drivers contains general information about the drivers used for the
1016 > cat /proc/ide/drivers
1017 ide-cdrom version 4.53
1018 ide-disk version 1.08
1020 More detailed information can be found in the controller specific
1021 subdirectories. These are named ide0, ide1 and so on. Each of these
1022 directories contains the files shown in table 1-6.
1025 Table 1-6: IDE controller info in /proc/ide/ide?
1026 ..............................................................................
1028 channel IDE channel (0 or 1)
1029 config Configuration (only for PCI/IDE bridge)
1031 model Type/Chipset of IDE controller
1032 ..............................................................................
1034 Each device connected to a controller has a separate subdirectory in the
1035 controllers directory. The files listed in table 1-7 are contained in these
1039 Table 1-7: IDE device information
1040 ..............................................................................
1043 capacity Capacity of the medium (in 512Byte blocks)
1044 driver driver and version
1045 geometry physical and logical geometry
1046 identify device identify block
1048 model device identifier
1049 settings device setup
1050 smart_thresholds IDE disk management thresholds
1051 smart_values IDE disk management values
1052 ..............................................................................
1054 The most interesting file is settings. This file contains a nice overview of
1055 the drive parameters:
1057 # cat /proc/ide/ide0/hda/settings
1058 name value min max mode
1059 ---- ----- --- --- ----
1060 bios_cyl 526 0 65535 rw
1061 bios_head 255 0 255 rw
1062 bios_sect 63 0 63 rw
1063 breada_readahead 4 0 127 rw
1065 file_readahead 72 0 2097151 rw
1067 keepsettings 0 0 1 rw
1068 max_kb_per_request 122 1 127 rw
1072 pio_mode write-only 0 255 w
1078 1.4 Networking info in /proc/net
1079 --------------------------------
1081 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1082 additional values you get for IP version 6 if you configure the kernel to
1083 support this. Table 1-9 lists the files and their meaning.
1086 Table 1-8: IPv6 info in /proc/net
1087 ..............................................................................
1089 udp6 UDP sockets (IPv6)
1090 tcp6 TCP sockets (IPv6)
1091 raw6 Raw device statistics (IPv6)
1092 igmp6 IP multicast addresses, which this host joined (IPv6)
1093 if_inet6 List of IPv6 interface addresses
1094 ipv6_route Kernel routing table for IPv6
1095 rt6_stats Global IPv6 routing tables statistics
1096 sockstat6 Socket statistics (IPv6)
1097 snmp6 Snmp data (IPv6)
1098 ..............................................................................
1101 Table 1-9: Network info in /proc/net
1102 ..............................................................................
1104 arp Kernel ARP table
1105 dev network devices with statistics
1106 dev_mcast the Layer2 multicast groups a device is listening too
1107 (interface index, label, number of references, number of bound
1109 dev_stat network device status
1110 ip_fwchains Firewall chain linkage
1111 ip_fwnames Firewall chain names
1112 ip_masq Directory containing the masquerading tables
1113 ip_masquerade Major masquerading table
1114 netstat Network statistics
1115 raw raw device statistics
1116 route Kernel routing table
1117 rpc Directory containing rpc info
1118 rt_cache Routing cache
1120 sockstat Socket statistics
1123 unix UNIX domain sockets
1124 wireless Wireless interface data (Wavelan etc)
1125 igmp IP multicast addresses, which this host joined
1126 psched Global packet scheduler parameters.
1127 netlink List of PF_NETLINK sockets
1128 ip_mr_vifs List of multicast virtual interfaces
1129 ip_mr_cache List of multicast routing cache
1130 ..............................................................................
1132 You can use this information to see which network devices are available in
1133 your system and how much traffic was routed over those devices:
1136 Inter-|Receive |[...
1137 face |bytes packets errs drop fifo frame compressed multicast|[...
1138 lo: 908188 5596 0 0 0 0 0 0 [...
1139 ppp0:15475140 20721 410 0 0 410 0 0 [...
1140 eth0: 614530 7085 0 0 0 0 0 1 [...
1143 ...] bytes packets errs drop fifo colls carrier compressed
1144 ...] 908188 5596 0 0 0 0 0 0
1145 ...] 1375103 17405 0 0 0 0 0 0
1146 ...] 1703981 5535 0 0 0 3 0 0
1148 In addition, each Channel Bond interface has its own directory. For
1149 example, the bond0 device will have a directory called /proc/net/bond0/.
1150 It will contain information that is specific to that bond, such as the
1151 current slaves of the bond, the link status of the slaves, and how
1152 many times the slaves link has failed.
1157 If you have a SCSI host adapter in your system, you'll find a subdirectory
1158 named after the driver for this adapter in /proc/scsi. You'll also see a list
1159 of all recognized SCSI devices in /proc/scsi:
1161 >cat /proc/scsi/scsi
1163 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1164 Vendor: IBM Model: DGHS09U Rev: 03E0
1165 Type: Direct-Access ANSI SCSI revision: 03
1166 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1167 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1168 Type: CD-ROM ANSI SCSI revision: 02
1171 The directory named after the driver has one file for each adapter found in
1172 the system. These files contain information about the controller, including
1173 the used IRQ and the IO address range. The amount of information shown is
1174 dependent on the adapter you use. The example shows the output for an Adaptec
1175 AHA-2940 SCSI adapter:
1177 > cat /proc/scsi/aic7xxx/0
1179 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1181 TCQ Enabled By Default : Disabled
1182 AIC7XXX_PROC_STATS : Disabled
1183 AIC7XXX_RESET_DELAY : 5
1184 Adapter Configuration:
1185 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1186 Ultra Wide Controller
1187 PCI MMAPed I/O Base: 0xeb001000
1188 Adapter SEEPROM Config: SEEPROM found and used.
1189 Adaptec SCSI BIOS: Enabled
1191 SCBs: Active 0, Max Active 2,
1192 Allocated 15, HW 16, Page 255
1194 BIOS Control Word: 0x18b6
1195 Adapter Control Word: 0x005b
1196 Extended Translation: Enabled
1197 Disconnect Enable Flags: 0xffff
1198 Ultra Enable Flags: 0x0001
1199 Tag Queue Enable Flags: 0x0000
1200 Ordered Queue Tag Flags: 0x0000
1201 Default Tag Queue Depth: 8
1202 Tagged Queue By Device array for aic7xxx host instance 0:
1203 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1204 Actual queue depth per device for aic7xxx host instance 0:
1205 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1208 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1209 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1210 Total transfers 160151 (74577 reads and 85574 writes)
1212 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1213 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1214 Total transfers 0 (0 reads and 0 writes)
1217 1.6 Parallel port info in /proc/parport
1218 ---------------------------------------
1220 The directory /proc/parport contains information about the parallel ports of
1221 your system. It has one subdirectory for each port, named after the port
1224 These directories contain the four files shown in Table 1-10.
1227 Table 1-10: Files in /proc/parport
1228 ..............................................................................
1230 autoprobe Any IEEE-1284 device ID information that has been acquired.
1231 devices list of the device drivers using that port. A + will appear by the
1232 name of the device currently using the port (it might not appear
1234 hardware Parallel port's base address, IRQ line and DMA channel.
1235 irq IRQ that parport is using for that port. This is in a separate
1236 file to allow you to alter it by writing a new value in (IRQ
1238 ..............................................................................
1240 1.7 TTY info in /proc/tty
1241 -------------------------
1243 Information about the available and actually used tty's can be found in the
1244 directory /proc/tty.You'll find entries for drivers and line disciplines in
1245 this directory, as shown in Table 1-11.
1248 Table 1-11: Files in /proc/tty
1249 ..............................................................................
1251 drivers list of drivers and their usage
1252 ldiscs registered line disciplines
1253 driver/serial usage statistic and status of single tty lines
1254 ..............................................................................
1256 To see which tty's are currently in use, you can simply look into the file
1259 > cat /proc/tty/drivers
1260 pty_slave /dev/pts 136 0-255 pty:slave
1261 pty_master /dev/ptm 128 0-255 pty:master
1262 pty_slave /dev/ttyp 3 0-255 pty:slave
1263 pty_master /dev/pty 2 0-255 pty:master
1264 serial /dev/cua 5 64-67 serial:callout
1265 serial /dev/ttyS 4 64-67 serial
1266 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1267 /dev/ptmx /dev/ptmx 5 2 system
1268 /dev/console /dev/console 5 1 system:console
1269 /dev/tty /dev/tty 5 0 system:/dev/tty
1270 unknown /dev/tty 4 1-63 console
1273 1.8 Miscellaneous kernel statistics in /proc/stat
1274 -------------------------------------------------
1276 Various pieces of information about kernel activity are available in the
1277 /proc/stat file. All of the numbers reported in this file are aggregates
1278 since the system first booted. For a quick look, simply cat the file:
1281 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1282 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1283 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1284 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1290 softirq 183433 0 21755 12 39 1137 231 21459 2263
1292 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1293 lines. These numbers identify the amount of time the CPU has spent performing
1294 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1295 second). The meanings of the columns are as follows, from left to right:
1297 - user: normal processes executing in user mode
1298 - nice: niced processes executing in user mode
1299 - system: processes executing in kernel mode
1300 - idle: twiddling thumbs
1301 - iowait: waiting for I/O to complete
1302 - irq: servicing interrupts
1303 - softirq: servicing softirqs
1304 - steal: involuntary wait
1305 - guest: running a normal guest
1306 - guest_nice: running a niced guest
1308 The "intr" line gives counts of interrupts serviced since boot time, for each
1309 of the possible system interrupts. The first column is the total of all
1310 interrupts serviced including unnumbered architecture specific interrupts;
1311 each subsequent column is the total for that particular numbered interrupt.
1312 Unnumbered interrupts are not shown, only summed into the total.
1314 The "ctxt" line gives the total number of context switches across all CPUs.
1316 The "btime" line gives the time at which the system booted, in seconds since
1319 The "processes" line gives the number of processes and threads created, which
1320 includes (but is not limited to) those created by calls to the fork() and
1321 clone() system calls.
1323 The "procs_running" line gives the total number of threads that are
1324 running or ready to run (i.e., the total number of runnable threads).
1326 The "procs_blocked" line gives the number of processes currently blocked,
1327 waiting for I/O to complete.
1329 The "softirq" line gives counts of softirqs serviced since boot time, for each
1330 of the possible system softirqs. The first column is the total of all
1331 softirqs serviced; each subsequent column is the total for that particular
1335 1.9 Ext4 file system parameters
1336 -------------------------------
1338 Information about mounted ext4 file systems can be found in
1339 /proc/fs/ext4. Each mounted filesystem will have a directory in
1340 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1341 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1342 in Table 1-12, below.
1344 Table 1-12: Files in /proc/fs/ext4/<devname>
1345 ..............................................................................
1347 mb_groups details of multiblock allocator buddy cache of free blocks
1348 ..............................................................................
1352 Shows registered system console lines.
1354 To see which character device lines are currently used for the system console
1355 /dev/console, you may simply look into the file /proc/consoles:
1357 > cat /proc/consoles
1363 device name of the device
1364 operations R = can do read operations
1365 W = can do write operations
1367 flags E = it is enabled
1368 C = it is preferred console
1369 B = it is primary boot console
1370 p = it is used for printk buffer
1371 b = it is not a TTY but a Braille device
1372 a = it is safe to use when cpu is offline
1373 major:minor major and minor number of the device separated by a colon
1375 ------------------------------------------------------------------------------
1377 ------------------------------------------------------------------------------
1378 The /proc file system serves information about the running system. It not only
1379 allows access to process data but also allows you to request the kernel status
1380 by reading files in the hierarchy.
1382 The directory structure of /proc reflects the types of information and makes
1383 it easy, if not obvious, where to look for specific data.
1384 ------------------------------------------------------------------------------
1386 ------------------------------------------------------------------------------
1387 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1388 ------------------------------------------------------------------------------
1390 ------------------------------------------------------------------------------
1392 ------------------------------------------------------------------------------
1393 * Modifying kernel parameters by writing into files found in /proc/sys
1394 * Exploring the files which modify certain parameters
1395 * Review of the /proc/sys file tree
1396 ------------------------------------------------------------------------------
1399 A very interesting part of /proc is the directory /proc/sys. This is not only
1400 a source of information, it also allows you to change parameters within the
1401 kernel. Be very careful when attempting this. You can optimize your system,
1402 but you can also cause it to crash. Never alter kernel parameters on a
1403 production system. Set up a development machine and test to make sure that
1404 everything works the way you want it to. You may have no alternative but to
1405 reboot the machine once an error has been made.
1407 To change a value, simply echo the new value into the file. An example is
1408 given below in the section on the file system data. You need to be root to do
1409 this. You can create your own boot script to perform this every time your
1412 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1413 general things in the operation of the Linux kernel. Since some of the files
1414 can inadvertently disrupt your system, it is advisable to read both
1415 documentation and source before actually making adjustments. In any case, be
1416 very careful when writing to any of these files. The entries in /proc may
1417 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1418 review the kernel documentation in the directory /usr/src/linux/Documentation.
1419 This chapter is heavily based on the documentation included in the pre 2.2
1420 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1422 Please see: Documentation/sysctl/ directory for descriptions of these
1425 ------------------------------------------------------------------------------
1427 ------------------------------------------------------------------------------
1428 Certain aspects of kernel behavior can be modified at runtime, without the
1429 need to recompile the kernel, or even to reboot the system. The files in the
1430 /proc/sys tree can not only be read, but also modified. You can use the echo
1431 command to write value into these files, thereby changing the default settings
1433 ------------------------------------------------------------------------------
1435 ------------------------------------------------------------------------------
1436 CHAPTER 3: PER-PROCESS PARAMETERS
1437 ------------------------------------------------------------------------------
1439 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1440 --------------------------------------------------------------------------------
1442 These file can be used to adjust the badness heuristic used to select which
1443 process gets killed in out of memory conditions.
1445 The badness heuristic assigns a value to each candidate task ranging from 0
1446 (never kill) to 1000 (always kill) to determine which process is targeted. The
1447 units are roughly a proportion along that range of allowed memory the process
1448 may allocate from based on an estimation of its current memory and swap use.
1449 For example, if a task is using all allowed memory, its badness score will be
1450 1000. If it is using half of its allowed memory, its score will be 500.
1452 There is an additional factor included in the badness score: the current memory
1453 and swap usage is discounted by 3% for root processes.
1455 The amount of "allowed" memory depends on the context in which the oom killer
1456 was called. If it is due to the memory assigned to the allocating task's cpuset
1457 being exhausted, the allowed memory represents the set of mems assigned to that
1458 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1459 memory represents the set of mempolicy nodes. If it is due to a memory
1460 limit (or swap limit) being reached, the allowed memory is that configured
1461 limit. Finally, if it is due to the entire system being out of memory, the
1462 allowed memory represents all allocatable resources.
1464 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1465 is used to determine which task to kill. Acceptable values range from -1000
1466 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1467 polarize the preference for oom killing either by always preferring a certain
1468 task or completely disabling it. The lowest possible value, -1000, is
1469 equivalent to disabling oom killing entirely for that task since it will always
1470 report a badness score of 0.
1472 Consequently, it is very simple for userspace to define the amount of memory to
1473 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1474 example, is roughly equivalent to allowing the remainder of tasks sharing the
1475 same system, cpuset, mempolicy, or memory controller resources to use at least
1476 50% more memory. A value of -500, on the other hand, would be roughly
1477 equivalent to discounting 50% of the task's allowed memory from being considered
1478 as scoring against the task.
1480 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1481 be used to tune the badness score. Its acceptable values range from -16
1482 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1483 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1484 scaled linearly with /proc/<pid>/oom_score_adj.
1486 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1487 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1488 requires CAP_SYS_RESOURCE.
1490 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1491 generation children with separate address spaces instead, if possible. This
1492 avoids servers and important system daemons from being killed and loses the
1493 minimal amount of work.
1496 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1497 -------------------------------------------------------------
1499 This file can be used to check the current score used by the oom-killer is for
1500 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1501 process should be killed in an out-of-memory situation.
1504 3.3 /proc/<pid>/io - Display the IO accounting fields
1505 -------------------------------------------------------
1507 This file contains IO statistics for each running process
1512 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1515 test:/tmp # cat /proc/3828/io
1521 write_bytes: 323932160
1522 cancelled_write_bytes: 0
1531 I/O counter: chars read
1532 The number of bytes which this task has caused to be read from storage. This
1533 is simply the sum of bytes which this process passed to read() and pread().
1534 It includes things like tty IO and it is unaffected by whether or not actual
1535 physical disk IO was required (the read might have been satisfied from
1542 I/O counter: chars written
1543 The number of bytes which this task has caused, or shall cause to be written
1544 to disk. Similar caveats apply here as with rchar.
1550 I/O counter: read syscalls
1551 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1558 I/O counter: write syscalls
1559 Attempt to count the number of write I/O operations, i.e. syscalls like
1560 write() and pwrite().
1566 I/O counter: bytes read
1567 Attempt to count the number of bytes which this process really did cause to
1568 be fetched from the storage layer. Done at the submit_bio() level, so it is
1569 accurate for block-backed filesystems. <please add status regarding NFS and
1570 CIFS at a later time>
1576 I/O counter: bytes written
1577 Attempt to count the number of bytes which this process caused to be sent to
1578 the storage layer. This is done at page-dirtying time.
1581 cancelled_write_bytes
1582 ---------------------
1584 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1585 then deletes the file, it will in fact perform no writeout. But it will have
1586 been accounted as having caused 1MB of write.
1587 In other words: The number of bytes which this process caused to not happen,
1588 by truncating pagecache. A task can cause "negative" IO too. If this task
1589 truncates some dirty pagecache, some IO which another task has been accounted
1590 for (in its write_bytes) will not be happening. We _could_ just subtract that
1591 from the truncating task's write_bytes, but there is information loss in doing
1598 At its current implementation state, this is a bit racy on 32-bit machines: if
1599 process A reads process B's /proc/pid/io while process B is updating one of
1600 those 64-bit counters, process A could see an intermediate result.
1603 More information about this can be found within the taskstats documentation in
1604 Documentation/accounting.
1606 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1607 ---------------------------------------------------------------
1608 When a process is dumped, all anonymous memory is written to a core file as
1609 long as the size of the core file isn't limited. But sometimes we don't want
1610 to dump some memory segments, for example, huge shared memory. Conversely,
1611 sometimes we want to save file-backed memory segments into a core file, not
1612 only the individual files.
1614 /proc/<pid>/coredump_filter allows you to customize which memory segments
1615 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1616 of memory types. If a bit of the bitmask is set, memory segments of the
1617 corresponding memory type are dumped, otherwise they are not dumped.
1619 The following 7 memory types are supported:
1620 - (bit 0) anonymous private memory
1621 - (bit 1) anonymous shared memory
1622 - (bit 2) file-backed private memory
1623 - (bit 3) file-backed shared memory
1624 - (bit 4) ELF header pages in file-backed private memory areas (it is
1625 effective only if the bit 2 is cleared)
1626 - (bit 5) hugetlb private memory
1627 - (bit 6) hugetlb shared memory
1629 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1630 are always dumped regardless of the bitmask status.
1632 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1633 effected by bit 5-6.
1635 Default value of coredump_filter is 0x23; this means all anonymous memory
1636 segments and hugetlb private memory are dumped.
1638 If you don't want to dump all shared memory segments attached to pid 1234,
1639 write 0x21 to the process's proc file.
1641 $ echo 0x21 > /proc/1234/coredump_filter
1643 When a new process is created, the process inherits the bitmask status from its
1644 parent. It is useful to set up coredump_filter before the program runs.
1647 $ echo 0x7 > /proc/self/coredump_filter
1650 3.5 /proc/<pid>/mountinfo - Information about mounts
1651 --------------------------------------------------------
1653 This file contains lines of the form:
1655 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1656 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1658 (1) mount ID: unique identifier of the mount (may be reused after umount)
1659 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1660 (3) major:minor: value of st_dev for files on filesystem
1661 (4) root: root of the mount within the filesystem
1662 (5) mount point: mount point relative to the process's root
1663 (6) mount options: per mount options
1664 (7) optional fields: zero or more fields of the form "tag[:value]"
1665 (8) separator: marks the end of the optional fields
1666 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1667 (10) mount source: filesystem specific information or "none"
1668 (11) super options: per super block options
1670 Parsers should ignore all unrecognised optional fields. Currently the
1671 possible optional fields are:
1673 shared:X mount is shared in peer group X
1674 master:X mount is slave to peer group X
1675 propagate_from:X mount is slave and receives propagation from peer group X (*)
1676 unbindable mount is unbindable
1678 (*) X is the closest dominant peer group under the process's root. If
1679 X is the immediate master of the mount, or if there's no dominant peer
1680 group under the same root, then only the "master:X" field is present
1681 and not the "propagate_from:X" field.
1683 For more information on mount propagation see:
1685 Documentation/filesystems/sharedsubtree.txt
1688 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1689 --------------------------------------------------------
1690 These files provide a method to access a tasks comm value. It also allows for
1691 a task to set its own or one of its thread siblings comm value. The comm value
1692 is limited in size compared to the cmdline value, so writing anything longer
1693 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1697 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1698 -------------------------------------------------------------------------
1699 This file provides a fast way to retrieve first level children pids
1700 of a task pointed by <pid>/<tid> pair. The format is a space separated
1703 Note the "first level" here -- if a child has own children they will
1704 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1705 to obtain the descendants.
1707 Since this interface is intended to be fast and cheap it doesn't
1708 guarantee to provide precise results and some children might be
1709 skipped, especially if they've exited right after we printed their
1710 pids, so one need to either stop or freeze processes being inspected
1711 if precise results are needed.
1714 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1715 ---------------------------------------------------------------
1716 This file provides information associated with an opened file. The regular
1717 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1718 represents the current offset of the opened file in decimal form [see lseek(2)
1719 for details], 'flags' denotes the octal O_xxx mask the file has been
1720 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1721 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1730 All locks associated with a file descriptor are shown in its fdinfo too.
1732 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1734 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1735 pair provide additional information particular to the objects they represent.
1744 where 'eventfd-count' is hex value of a counter.
1751 sigmask: 0000000000000200
1753 where 'sigmask' is hex value of the signal mask associated
1761 tfd: 5 events: 1d data: ffffffffffffffff
1763 where 'tfd' is a target file descriptor number in decimal form,
1764 'events' is events mask being watched and the 'data' is data
1765 associated with a target [see epoll(7) for more details].
1769 For inotify files the format is the following
1773 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1775 where 'wd' is a watch descriptor in decimal form, ie a target file
1776 descriptor number, 'ino' and 'sdev' are inode and device where the
1777 target file resides and the 'mask' is the mask of events, all in hex
1778 form [see inotify(7) for more details].
1780 If the kernel was built with exportfs support, the path to the target
1781 file is encoded as a file handle. The file handle is provided by three
1782 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1785 If the kernel is built without exportfs support the file handle won't be
1788 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1790 For fanotify files the format is
1795 fanotify flags:10 event-flags:0
1796 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1797 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1799 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1800 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1801 flags associated with mark which are tracked separately from events
1802 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1803 mask and 'ignored_mask' is the mask of events which are to be ignored.
1804 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1805 does provide information about flags and mask used in fanotify_mark
1806 call [see fsnotify manpage for details].
1808 While the first three lines are mandatory and always printed, the rest is
1809 optional and may be omitted if no marks created yet.
1820 it_value: (0, 49406829)
1823 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1824 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1825 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1826 details]. 'it_value' is remaining time until the timer exiration.
1827 'it_interval' is the interval for the timer. Note the timer might be set up
1828 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1829 still exhibits timer's remaining time.
1831 3.9 /proc/<pid>/map_files - Information about memory mapped files
1832 ---------------------------------------------------------------------
1833 This directory contains symbolic links which represent memory mapped files
1834 the process is maintaining. Example output:
1836 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1837 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1838 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1840 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1841 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1843 The name of a link represents the virtual memory bounds of a mapping, i.e.
1844 vm_area_struct::vm_start-vm_area_struct::vm_end.
1846 The main purpose of the map_files is to retrieve a set of memory mapped
1847 files in a fast way instead of parsing /proc/<pid>/maps or
1848 /proc/<pid>/smaps, both of which contain many more records. At the same
1849 time one can open(2) mappings from the listings of two processes and
1850 comparing their inode numbers to figure out which anonymous memory areas
1851 are actually shared.
1853 ------------------------------------------------------------------------------
1855 ------------------------------------------------------------------------------
1858 ---------------------
1860 The following mount options are supported:
1862 hidepid= Set /proc/<pid>/ access mode.
1863 gid= Set the group authorized to learn processes information.
1865 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1868 hidepid=1 means users may not access any /proc/<pid>/ directories but their
1869 own. Sensitive files like cmdline, sched*, status are now protected against
1870 other users. This makes it impossible to learn whether any user runs
1871 specific program (given the program doesn't reveal itself by its behaviour).
1872 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1873 poorly written programs passing sensitive information via program arguments are
1874 now protected against local eavesdroppers.
1876 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1877 users. It doesn't mean that it hides a fact whether a process with a specific
1878 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1879 but it hides process' uid and gid, which may be learned by stat()'ing
1880 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1881 information about running processes, whether some daemon runs with elevated
1882 privileges, whether other user runs some sensitive program, whether other users
1883 run any program at all, etc.
1885 gid= defines a group authorized to learn processes information otherwise
1886 prohibited by hidepid=. If you use some daemon like identd which needs to learn
1887 information about processes information, just add identd to this group.
projects / linux-beck.git / blob