proc.txt 68 KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889909192939495969798991001011021031041051061071081091101111121131141151161171181191201211221231241251261271281291301311321331341351361371381391401411421431441451461471481491501511521531541551561571581591601611621631641651661671681691701711721731741751761771781791801811821831841851861871881891901911921931941951961971981992002012022032042052062072082092102112122132142152162172182192202212222232242252262272282292302312322332342352362372382392402412422432442452462472482492502512522532542552562572582592602612622632642652662672682692702712722732742752762772782792802812822832842852862872882892902912922932942952962972982993003013023033043053063073083093103113123133143153163173183193203213223233243253263273283293303313323333343353363373383393403413423433443453463473483493503513523533543553563573583593603613623633643653663673683693703713723733743753763773783793803813823833843853863873883893903913923933943953963973983994004014024034044054064074084094104114124134144154164174184194204214224234244254264274284294304314324334344354364374384394404414424434444454464474484494504514524534544554564574584594604614624634644654664674684694704714724734744754764774784794804814824834844854864874884894904914924934944954964974984995005015025035045055065075085095105115125135145155165175185195205215225235245255265275285295305315325335345355365375385395405415425435445455465475485495505515525535545555565575585595605615625635645655665675685695705715725735745755765775785795805815825835845855865875885895905915925935945955965975985996006016026036046056066076086096106116126136146156166176186196206216226236246256266276286296306316326336346356366376386396406416426436446456466476486496506516526536546556566576586596606616626636646656666676686696706716726736746756766776786796806816826836846856866876886896906916926936946956966976986997007017027037047057067077087097107117127137147157167177187197207217227237247257267277287297307317327337347357367377387397407417427437447457467477487497507517527537547557567577587597607617627637647657667677687697707717727737747757767777787797807817827837847857867877887897907917927937947957967977987998008018028038048058068078088098108118128138148158168178188198208218228238248258268278288298308318328338348358368378388398408418428438448458468478488498508518528538548558568578588598608618628638648658668678688698708718728738748758768778788798808818828838848858868878888898908918928938948958968978988999009019029039049059069079089099109119129139149159169179189199209219229239249259269279289299309319329339349359369379389399409419429439449459469479489499509519529539549559569579589599609619629639649659669679689699709719729739749759769779789799809819829839849859869879889899909919929939949959969979989991000100110021003100410051006100710081009101010111012101310141015101610171018101910201021102210231024102510261027102810291030103110321033103410351036103710381039104010411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105110611071108110911101111111211131114111511161117111811191120112111221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180118111821183118411851186118711881189119011911192119311941195119611971198119912001201120212031204120512061207120812091210121112121213121412151216121712181219122012211222122312241225122612271228122912301231123212331234123512361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357135813591360136113621363136413651366136713681369137013711372137313741375137613771378137913801381138213831384138513861387138813891390139113921393139413951396139713981399140014011402140314041405140614071408140914101411141214131414141514161417141814191420142114221423142414251426142714281429143014311432143314341435143614371438143914401441144214431444144514461447144814491450145114521453145414551456145714581459146014611462146314641465146614671468146914701471147214731474147514761477147814791480148114821483148414851486148714881489149014911492149314941495
  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>
  6. 2.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
  7. move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
  8. ------------------------------------------------------------------------------
  9. Version 1.3 Kernel version 2.2.12
  10. Kernel version 2.4.0-test11-pre4
  11. ------------------------------------------------------------------------------
  12. fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
  13. Table of Contents
  14. -----------------
  15. 0 Preface
  16. 0.1 Introduction/Credits
  17. 0.2 Legal Stuff
  18. 1 Collecting System Information
  19. 1.1 Process-Specific Subdirectories
  20. 1.2 Kernel data
  21. 1.3 IDE devices in /proc/ide
  22. 1.4 Networking info in /proc/net
  23. 1.5 SCSI info
  24. 1.6 Parallel port info in /proc/parport
  25. 1.7 TTY info in /proc/tty
  26. 1.8 Miscellaneous kernel statistics in /proc/stat
  27. 1.9 Ext4 file system parameters
  28. 2 Modifying System Parameters
  29. 3 Per-Process Parameters
  30. 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
  31. score
  32. 3.2 /proc/<pid>/oom_score - Display current oom-killer score
  33. 3.3 /proc/<pid>/io - Display the IO accounting fields
  34. 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
  35. 3.5 /proc/<pid>/mountinfo - Information about mounts
  36. 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
  37. ------------------------------------------------------------------------------
  38. Preface
  39. ------------------------------------------------------------------------------
  40. 0.1 Introduction/Credits
  41. ------------------------
  42. This documentation is part of a soon (or so we hope) to be released book on
  43. the SuSE Linux distribution. As there is no complete documentation for the
  44. /proc file system and we've used many freely available sources to write these
  45. chapters, it seems only fair to give the work back to the Linux community.
  46. This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
  47. afraid it's still far from complete, but we hope it will be useful. As far as
  48. we know, it is the first 'all-in-one' document about the /proc file system. It
  49. is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
  50. SPARC, AXP, etc., features, you probably won't find what you are looking for.
  51. It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
  52. additions and patches are welcome and will be added to this document if you
  53. mail them to Bodo.
  54. We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
  55. other people for help compiling this documentation. We'd also like to extend a
  56. special thank you to Andi Kleen for documentation, which we relied on heavily
  57. to create this document, as well as the additional information he provided.
  58. Thanks to everybody else who contributed source or docs to the Linux kernel
  59. and helped create a great piece of software... :)
  60. If you have any comments, corrections or additions, please don't hesitate to
  61. contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
  62. document.
  63. The latest version of this document is available online at
  64. http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
  65. If the above direction does not works for you, you could try the kernel
  66. mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
  67. comandante@zaralinux.com.
  68. 0.2 Legal Stuff
  69. ---------------
  70. We don't guarantee the correctness of this document, and if you come to us
  71. complaining about how you screwed up your system because of incorrect
  72. documentation, we won't feel responsible...
  73. ------------------------------------------------------------------------------
  74. CHAPTER 1: COLLECTING SYSTEM INFORMATION
  75. ------------------------------------------------------------------------------
  76. ------------------------------------------------------------------------------
  77. In This Chapter
  78. ------------------------------------------------------------------------------
  79. * Investigating the properties of the pseudo file system /proc and its
  80. ability to provide information on the running Linux system
  81. * Examining /proc's structure
  82. * Uncovering various information about the kernel and the processes running
  83. on the system
  84. ------------------------------------------------------------------------------
  85. The proc file system acts as an interface to internal data structures in the
  86. kernel. It can be used to obtain information about the system and to change
  87. certain kernel parameters at runtime (sysctl).
  88. First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
  89. show you how you can use /proc/sys to change settings.
  90. 1.1 Process-Specific Subdirectories
  91. -----------------------------------
  92. The directory /proc contains (among other things) one subdirectory for each
  93. process running on the system, which is named after the process ID (PID).
  94. The link self points to the process reading the file system. Each process
  95. subdirectory has the entries listed in Table 1-1.
  96. Table 1-1: Process specific entries in /proc
  97. ..............................................................................
  98. File Content
  99. clear_refs Clears page referenced bits shown in smaps output
  100. cmdline Command line arguments
  101. cpu Current and last cpu in which it was executed (2.4)(smp)
  102. cwd Link to the current working directory
  103. environ Values of environment variables
  104. exe Link to the executable of this process
  105. fd Directory, which contains all file descriptors
  106. maps Memory maps to executables and library files (2.4)
  107. mem Memory held by this process
  108. root Link to the root directory of this process
  109. stat Process status
  110. statm Process memory status information
  111. status Process status in human readable form
  112. wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
  113. stack Report full stack trace, enable via CONFIG_STACKTRACE
  114. smaps a extension based on maps, showing the memory consumption of
  115. each mapping
  116. ..............................................................................
  117. For example, to get the status information of a process, all you have to do is
  118. read the file /proc/PID/status:
  119. >cat /proc/self/status
  120. Name: cat
  121. State: R (running)
  122. Tgid: 5452
  123. Pid: 5452
  124. PPid: 743
  125. TracerPid: 0 (2.4)
  126. Uid: 501 501 501 501
  127. Gid: 100 100 100 100
  128. FDSize: 256
  129. Groups: 100 14 16
  130. VmPeak: 5004 kB
  131. VmSize: 5004 kB
  132. VmLck: 0 kB
  133. VmHWM: 476 kB
  134. VmRSS: 476 kB
  135. VmData: 156 kB
  136. VmStk: 88 kB
  137. VmExe: 68 kB
  138. VmLib: 1412 kB
  139. VmPTE: 20 kb
  140. VmSwap: 0 kB
  141. Threads: 1
  142. SigQ: 0/28578
  143. SigPnd: 0000000000000000
  144. ShdPnd: 0000000000000000
  145. SigBlk: 0000000000000000
  146. SigIgn: 0000000000000000
  147. SigCgt: 0000000000000000
  148. CapInh: 00000000fffffeff
  149. CapPrm: 0000000000000000
  150. CapEff: 0000000000000000
  151. CapBnd: ffffffffffffffff
  152. voluntary_ctxt_switches: 0
  153. nonvoluntary_ctxt_switches: 1
  154. This shows you nearly the same information you would get if you viewed it with
  155. the ps command. In fact, ps uses the proc file system to obtain its
  156. information. But you get a more detailed view of the process by reading the
  157. file /proc/PID/status. It fields are described in table 1-2.
  158. The statm file contains more detailed information about the process
  159. memory usage. Its seven fields are explained in Table 1-3. The stat file
  160. contains details information about the process itself. Its fields are
  161. explained in Table 1-4.
  162. (for SMP CONFIG users)
  163. For making accounting scalable, RSS related information are handled in
  164. asynchronous manner and the vaule may not be very precise. To see a precise
  165. snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
  166. It's slow but very precise.
  167. Table 1-2: Contents of the status files (as of 2.6.30-rc7)
  168. ..............................................................................
  169. Field Content
  170. Name filename of the executable
  171. State state (R is running, S is sleeping, D is sleeping
  172. in an uninterruptible wait, Z is zombie,
  173. T is traced or stopped)
  174. Tgid thread group ID
  175. Pid process id
  176. PPid process id of the parent process
  177. TracerPid PID of process tracing this process (0 if not)
  178. Uid Real, effective, saved set, and file system UIDs
  179. Gid Real, effective, saved set, and file system GIDs
  180. FDSize number of file descriptor slots currently allocated
  181. Groups supplementary group list
  182. VmPeak peak virtual memory size
  183. VmSize total program size
  184. VmLck locked memory size
  185. VmHWM peak resident set size ("high water mark")
  186. VmRSS size of memory portions
  187. VmData size of data, stack, and text segments
  188. VmStk size of data, stack, and text segments
  189. VmExe size of text segment
  190. VmLib size of shared library code
  191. VmPTE size of page table entries
  192. VmSwap size of swap usage (the number of referred swapents)
  193. Threads number of threads
  194. SigQ number of signals queued/max. number for queue
  195. SigPnd bitmap of pending signals for the thread
  196. ShdPnd bitmap of shared pending signals for the process
  197. SigBlk bitmap of blocked signals
  198. SigIgn bitmap of ignored signals
  199. SigCgt bitmap of catched signals
  200. CapInh bitmap of inheritable capabilities
  201. CapPrm bitmap of permitted capabilities
  202. CapEff bitmap of effective capabilities
  203. CapBnd bitmap of capabilities bounding set
  204. Cpus_allowed mask of CPUs on which this process may run
  205. Cpus_allowed_list Same as previous, but in "list format"
  206. Mems_allowed mask of memory nodes allowed to this process
  207. Mems_allowed_list Same as previous, but in "list format"
  208. voluntary_ctxt_switches number of voluntary context switches
  209. nonvoluntary_ctxt_switches number of non voluntary context switches
  210. ..............................................................................
  211. Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
  212. ..............................................................................
  213. Field Content
  214. size total program size (pages) (same as VmSize in status)
  215. resident size of memory portions (pages) (same as VmRSS in status)
  216. shared number of pages that are shared (i.e. backed by a file)
  217. trs number of pages that are 'code' (not including libs; broken,
  218. includes data segment)
  219. lrs number of pages of library (always 0 on 2.6)
  220. drs number of pages of data/stack (including libs; broken,
  221. includes library text)
  222. dt number of dirty pages (always 0 on 2.6)
  223. ..............................................................................
  224. Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
  225. ..............................................................................
  226. Field Content
  227. pid process id
  228. tcomm filename of the executable
  229. state state (R is running, S is sleeping, D is sleeping in an
  230. uninterruptible wait, Z is zombie, T is traced or stopped)
  231. ppid process id of the parent process
  232. pgrp pgrp of the process
  233. sid session id
  234. tty_nr tty the process uses
  235. tty_pgrp pgrp of the tty
  236. flags task flags
  237. min_flt number of minor faults
  238. cmin_flt number of minor faults with child's
  239. maj_flt number of major faults
  240. cmaj_flt number of major faults with child's
  241. utime user mode jiffies
  242. stime kernel mode jiffies
  243. cutime user mode jiffies with child's
  244. cstime kernel mode jiffies with child's
  245. priority priority level
  246. nice nice level
  247. num_threads number of threads
  248. it_real_value (obsolete, always 0)
  249. start_time time the process started after system boot
  250. vsize virtual memory size
  251. rss resident set memory size
  252. rsslim current limit in bytes on the rss
  253. start_code address above which program text can run
  254. end_code address below which program text can run
  255. start_stack address of the start of the stack
  256. esp current value of ESP
  257. eip current value of EIP
  258. pending bitmap of pending signals
  259. blocked bitmap of blocked signals
  260. sigign bitmap of ignored signals
  261. sigcatch bitmap of catched signals
  262. wchan address where process went to sleep
  263. 0 (place holder)
  264. 0 (place holder)
  265. exit_signal signal to send to parent thread on exit
  266. task_cpu which CPU the task is scheduled on
  267. rt_priority realtime priority
  268. policy scheduling policy (man sched_setscheduler)
  269. blkio_ticks time spent waiting for block IO
  270. gtime guest time of the task in jiffies
  271. cgtime guest time of the task children in jiffies
  272. ..............................................................................
  273. The /proc/PID/maps file containing the currently mapped memory regions and
  274. their access permissions.
  275. The format is:
  276. address perms offset dev inode pathname
  277. 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
  278. 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
  279. 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
  280. a7cb1000-a7cb2000 ---p 00000000 00:00 0
  281. a7cb2000-a7eb2000 rw-p 00000000 00:00 0
  282. a7eb2000-a7eb3000 ---p 00000000 00:00 0
  283. a7eb3000-a7ed5000 rw-p 00000000 00:00 0
  284. a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
  285. a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
  286. a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
  287. a800b000-a800e000 rw-p 00000000 00:00 0
  288. a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
  289. a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
  290. a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
  291. a8024000-a8027000 rw-p 00000000 00:00 0
  292. a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
  293. a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
  294. a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
  295. aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
  296. ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
  297. where "address" is the address space in the process that it occupies, "perms"
  298. is a set of permissions:
  299. r = read
  300. w = write
  301. x = execute
  302. s = shared
  303. p = private (copy on write)
  304. "offset" is the offset into the mapping, "dev" is the device (major:minor), and
  305. "inode" is the inode on that device. 0 indicates that no inode is associated
  306. with the memory region, as the case would be with BSS (uninitialized data).
  307. The "pathname" shows the name associated file for this mapping. If the mapping
  308. is not associated with a file:
  309. [heap] = the heap of the program
  310. [stack] = the stack of the main process
  311. [vdso] = the "virtual dynamic shared object",
  312. the kernel system call handler
  313. or if empty, the mapping is anonymous.
  314. The /proc/PID/smaps is an extension based on maps, showing the memory
  315. consumption for each of the process's mappings. For each of mappings there
  316. is a series of lines such as the following:
  317. 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
  318. Size: 1084 kB
  319. Rss: 892 kB
  320. Pss: 374 kB
  321. Shared_Clean: 892 kB
  322. Shared_Dirty: 0 kB
  323. Private_Clean: 0 kB
  324. Private_Dirty: 0 kB
  325. Referenced: 892 kB
  326. Swap: 0 kB
  327. KernelPageSize: 4 kB
  328. MMUPageSize: 4 kB
  329. The first of these lines shows the same information as is displayed for the
  330. mapping in /proc/PID/maps. The remaining lines show the size of the mapping,
  331. the amount of the mapping that is currently resident in RAM, the "proportional
  332. set size” (divide each shared page by the number of processes sharing it), the
  333. number of clean and dirty shared pages in the mapping, and the number of clean
  334. and dirty private pages in the mapping. The "Referenced" indicates the amount
  335. of memory currently marked as referenced or accessed.
  336. This file is only present if the CONFIG_MMU kernel configuration option is
  337. enabled.
  338. The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
  339. bits on both physical and virtual pages associated with a process.
  340. To clear the bits for all the pages associated with the process
  341. > echo 1 > /proc/PID/clear_refs
  342. To clear the bits for the anonymous pages associated with the process
  343. > echo 2 > /proc/PID/clear_refs
  344. To clear the bits for the file mapped pages associated with the process
  345. > echo 3 > /proc/PID/clear_refs
  346. Any other value written to /proc/PID/clear_refs will have no effect.
  347. 1.2 Kernel data
  348. ---------------
  349. Similar to the process entries, the kernel data files give information about
  350. the running kernel. The files used to obtain this information are contained in
  351. /proc and are listed in Table 1-5. Not all of these will be present in your
  352. system. It depends on the kernel configuration and the loaded modules, which
  353. files are there, and which are missing.
  354. Table 1-5: Kernel info in /proc
  355. ..............................................................................
  356. File Content
  357. apm Advanced power management info
  358. buddyinfo Kernel memory allocator information (see text) (2.5)
  359. bus Directory containing bus specific information
  360. cmdline Kernel command line
  361. cpuinfo Info about the CPU
  362. devices Available devices (block and character)
  363. dma Used DMS channels
  364. filesystems Supported filesystems
  365. driver Various drivers grouped here, currently rtc (2.4)
  366. execdomains Execdomains, related to security (2.4)
  367. fb Frame Buffer devices (2.4)
  368. fs File system parameters, currently nfs/exports (2.4)
  369. ide Directory containing info about the IDE subsystem
  370. interrupts Interrupt usage
  371. iomem Memory map (2.4)
  372. ioports I/O port usage
  373. irq Masks for irq to cpu affinity (2.4)(smp?)
  374. isapnp ISA PnP (Plug&Play) Info (2.4)
  375. kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
  376. kmsg Kernel messages
  377. ksyms Kernel symbol table
  378. loadavg Load average of last 1, 5 & 15 minutes
  379. locks Kernel locks
  380. meminfo Memory info
  381. misc Miscellaneous
  382. modules List of loaded modules
  383. mounts Mounted filesystems
  384. net Networking info (see text)
  385. pagetypeinfo Additional page allocator information (see text) (2.5)
  386. partitions Table of partitions known to the system
  387. pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
  388. decoupled by lspci (2.4)
  389. rtc Real time clock
  390. scsi SCSI info (see text)
  391. slabinfo Slab pool info
  392. softirqs softirq usage
  393. stat Overall statistics
  394. swaps Swap space utilization
  395. sys See chapter 2
  396. sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
  397. tty Info of tty drivers
  398. uptime System uptime
  399. version Kernel version
  400. video bttv info of video resources (2.4)
  401. vmallocinfo Show vmalloced areas
  402. ..............................................................................
  403. You can, for example, check which interrupts are currently in use and what
  404. they are used for by looking in the file /proc/interrupts:
  405. > cat /proc/interrupts
  406. CPU0
  407. 0: 8728810 XT-PIC timer
  408. 1: 895 XT-PIC keyboard
  409. 2: 0 XT-PIC cascade
  410. 3: 531695 XT-PIC aha152x
  411. 4: 2014133 XT-PIC serial
  412. 5: 44401 XT-PIC pcnet_cs
  413. 8: 2 XT-PIC rtc
  414. 11: 8 XT-PIC i82365
  415. 12: 182918 XT-PIC PS/2 Mouse
  416. 13: 1 XT-PIC fpu
  417. 14: 1232265 XT-PIC ide0
  418. 15: 7 XT-PIC ide1
  419. NMI: 0
  420. In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
  421. output of a SMP machine):
  422. > cat /proc/interrupts
  423. CPU0 CPU1
  424. 0: 1243498 1214548 IO-APIC-edge timer
  425. 1: 8949 8958 IO-APIC-edge keyboard
  426. 2: 0 0 XT-PIC cascade
  427. 5: 11286 10161 IO-APIC-edge soundblaster
  428. 8: 1 0 IO-APIC-edge rtc
  429. 9: 27422 27407 IO-APIC-edge 3c503
  430. 12: 113645 113873 IO-APIC-edge PS/2 Mouse
  431. 13: 0 0 XT-PIC fpu
  432. 14: 22491 24012 IO-APIC-edge ide0
  433. 15: 2183 2415 IO-APIC-edge ide1
  434. 17: 30564 30414 IO-APIC-level eth0
  435. 18: 177 164 IO-APIC-level bttv
  436. NMI: 2457961 2457959
  437. LOC: 2457882 2457881
  438. ERR: 2155
  439. NMI is incremented in this case because every timer interrupt generates a NMI
  440. (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
  441. LOC is the local interrupt counter of the internal APIC of every CPU.
  442. ERR is incremented in the case of errors in the IO-APIC bus (the bus that
  443. connects the CPUs in a SMP system. This means that an error has been detected,
  444. the IO-APIC automatically retry the transmission, so it should not be a big
  445. problem, but you should read the SMP-FAQ.
  446. In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
  447. /proc/interrupts to display every IRQ vector in use by the system, not
  448. just those considered 'most important'. The new vectors are:
  449. THR -- interrupt raised when a machine check threshold counter
  450. (typically counting ECC corrected errors of memory or cache) exceeds
  451. a configurable threshold. Only available on some systems.
  452. TRM -- a thermal event interrupt occurs when a temperature threshold
  453. has been exceeded for the CPU. This interrupt may also be generated
  454. when the temperature drops back to normal.
  455. SPU -- a spurious interrupt is some interrupt that was raised then lowered
  456. by some IO device before it could be fully processed by the APIC. Hence
  457. the APIC sees the interrupt but does not know what device it came from.
  458. For this case the APIC will generate the interrupt with a IRQ vector
  459. of 0xff. This might also be generated by chipset bugs.
  460. RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
  461. sent from one CPU to another per the needs of the OS. Typically,
  462. their statistics are used by kernel developers and interested users to
  463. determine the occurrence of interrupts of the given type.
  464. The above IRQ vectors are displayed only when relevent. For example,
  465. the threshold vector does not exist on x86_64 platforms. Others are
  466. suppressed when the system is a uniprocessor. As of this writing, only
  467. i386 and x86_64 platforms support the new IRQ vector displays.
  468. Of some interest is the introduction of the /proc/irq directory to 2.4.
  469. It could be used to set IRQ to CPU affinity, this means that you can "hook" an
  470. IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
  471. irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
  472. prof_cpu_mask.
  473. For example
  474. > ls /proc/irq/
  475. 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
  476. 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
  477. > ls /proc/irq/0/
  478. smp_affinity
  479. smp_affinity is a bitmask, in which you can specify which CPUs can handle the
  480. IRQ, you can set it by doing:
  481. > echo 1 > /proc/irq/10/smp_affinity
  482. This means that only the first CPU will handle the IRQ, but you can also echo
  483. 5 which means that only the first and fourth CPU can handle the IRQ.
  484. The contents of each smp_affinity file is the same by default:
  485. > cat /proc/irq/0/smp_affinity
  486. ffffffff
  487. The default_smp_affinity mask applies to all non-active IRQs, which are the
  488. IRQs which have not yet been allocated/activated, and hence which lack a
  489. /proc/irq/[0-9]* directory.
  490. The node file on an SMP system shows the node to which the device using the IRQ
  491. reports itself as being attached. This hardware locality information does not
  492. include information about any possible driver locality preference.
  493. prof_cpu_mask specifies which CPUs are to be profiled by the system wide
  494. profiler. Default value is ffffffff (all cpus).
  495. The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
  496. between all the CPUs which are allowed to handle it. As usual the kernel has
  497. more info than you and does a better job than you, so the defaults are the
  498. best choice for almost everyone.
  499. There are three more important subdirectories in /proc: net, scsi, and sys.
  500. The general rule is that the contents, or even the existence of these
  501. directories, depend on your kernel configuration. If SCSI is not enabled, the
  502. directory scsi may not exist. The same is true with the net, which is there
  503. only when networking support is present in the running kernel.
  504. The slabinfo file gives information about memory usage at the slab level.
  505. Linux uses slab pools for memory management above page level in version 2.2.
  506. Commonly used objects have their own slab pool (such as network buffers,
  507. directory cache, and so on).
  508. ..............................................................................
  509. > cat /proc/buddyinfo
  510. Node 0, zone DMA 0 4 5 4 4 3 ...
  511. Node 0, zone Normal 1 0 0 1 101 8 ...
  512. Node 0, zone HighMem 2 0 0 1 1 0 ...
  513. External fragmentation is a problem under some workloads, and buddyinfo is a
  514. useful tool for helping diagnose these problems. Buddyinfo will give you a
  515. clue as to how big an area you can safely allocate, or why a previous
  516. allocation failed.
  517. Each column represents the number of pages of a certain order which are
  518. available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
  519. ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
  520. available in ZONE_NORMAL, etc...
  521. More information relevant to external fragmentation can be found in
  522. pagetypeinfo.
  523. > cat /proc/pagetypeinfo
  524. Page block order: 9
  525. Pages per block: 512
  526. Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
  527. Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
  528. Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
  529. Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
  530. Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
  531. Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
  532. Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
  533. Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
  534. Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
  535. Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
  536. Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
  537. Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
  538. Node 0, zone DMA 2 0 5 1 0
  539. Node 0, zone DMA32 41 6 967 2 0
  540. Fragmentation avoidance in the kernel works by grouping pages of different
  541. migrate types into the same contiguous regions of memory called page blocks.
  542. A page block is typically the size of the default hugepage size e.g. 2MB on
  543. X86-64. By keeping pages grouped based on their ability to move, the kernel
  544. can reclaim pages within a page block to satisfy a high-order allocation.
  545. The pagetypinfo begins with information on the size of a page block. It
  546. then gives the same type of information as buddyinfo except broken down
  547. by migrate-type and finishes with details on how many page blocks of each
  548. type exist.
  549. If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
  550. from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
  551. make an estimate of the likely number of huge pages that can be allocated
  552. at a given point in time. All the "Movable" blocks should be allocatable
  553. unless memory has been mlock()'d. Some of the Reclaimable blocks should
  554. also be allocatable although a lot of filesystem metadata may have to be
  555. reclaimed to achieve this.
  556. ..............................................................................
  557. meminfo:
  558. Provides information about distribution and utilization of memory. This
  559. varies by architecture and compile options. The following is from a
  560. 16GB PIII, which has highmem enabled. You may not have all of these fields.
  561. > cat /proc/meminfo
  562. MemTotal: 16344972 kB
  563. MemFree: 13634064 kB
  564. Buffers: 3656 kB
  565. Cached: 1195708 kB
  566. SwapCached: 0 kB
  567. Active: 891636 kB
  568. Inactive: 1077224 kB
  569. HighTotal: 15597528 kB
  570. HighFree: 13629632 kB
  571. LowTotal: 747444 kB
  572. LowFree: 4432 kB
  573. SwapTotal: 0 kB
  574. SwapFree: 0 kB
  575. Dirty: 968 kB
  576. Writeback: 0 kB
  577. AnonPages: 861800 kB
  578. Mapped: 280372 kB
  579. Slab: 284364 kB
  580. SReclaimable: 159856 kB
  581. SUnreclaim: 124508 kB
  582. PageTables: 24448 kB
  583. NFS_Unstable: 0 kB
  584. Bounce: 0 kB
  585. WritebackTmp: 0 kB
  586. CommitLimit: 7669796 kB
  587. Committed_AS: 100056 kB
  588. VmallocTotal: 112216 kB
  589. VmallocUsed: 428 kB
  590. VmallocChunk: 111088 kB
  591. MemTotal: Total usable ram (i.e. physical ram minus a few reserved
  592. bits and the kernel binary code)
  593. MemFree: The sum of LowFree+HighFree
  594. Buffers: Relatively temporary storage for raw disk blocks
  595. shouldn't get tremendously large (20MB or so)
  596. Cached: in-memory cache for files read from the disk (the
  597. pagecache). Doesn't include SwapCached
  598. SwapCached: Memory that once was swapped out, is swapped back in but
  599. still also is in the swapfile (if memory is needed it
  600. doesn't need to be swapped out AGAIN because it is already
  601. in the swapfile. This saves I/O)
  602. Active: Memory that has been used more recently and usually not
  603. reclaimed unless absolutely necessary.
  604. Inactive: Memory which has been less recently used. It is more
  605. eligible to be reclaimed for other purposes
  606. HighTotal:
  607. HighFree: Highmem is all memory above ~860MB of physical memory
  608. Highmem areas are for use by userspace programs, or
  609. for the pagecache. The kernel must use tricks to access
  610. this memory, making it slower to access than lowmem.
  611. LowTotal:
  612. LowFree: Lowmem is memory which can be used for everything that
  613. highmem can be used for, but it is also available for the
  614. kernel's use for its own data structures. Among many
  615. other things, it is where everything from the Slab is
  616. allocated. Bad things happen when you're out of lowmem.
  617. SwapTotal: total amount of swap space available
  618. SwapFree: Memory which has been evicted from RAM, and is temporarily
  619. on the disk
  620. Dirty: Memory which is waiting to get written back to the disk
  621. Writeback: Memory which is actively being written back to the disk
  622. AnonPages: Non-file backed pages mapped into userspace page tables
  623. Mapped: files which have been mmaped, such as libraries
  624. Slab: in-kernel data structures cache
  625. SReclaimable: Part of Slab, that might be reclaimed, such as caches
  626. SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
  627. PageTables: amount of memory dedicated to the lowest level of page
  628. tables.
  629. NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
  630. storage
  631. Bounce: Memory used for block device "bounce buffers"
  632. WritebackTmp: Memory used by FUSE for temporary writeback buffers
  633. CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
  634. this is the total amount of memory currently available to
  635. be allocated on the system. This limit is only adhered to
  636. if strict overcommit accounting is enabled (mode 2 in
  637. 'vm.overcommit_memory').
  638. The CommitLimit is calculated with the following formula:
  639. CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
  640. For example, on a system with 1G of physical RAM and 7G
  641. of swap with a `vm.overcommit_ratio` of 30 it would
  642. yield a CommitLimit of 7.3G.
  643. For more details, see the memory overcommit documentation
  644. in vm/overcommit-accounting.
  645. Committed_AS: The amount of memory presently allocated on the system.
  646. The committed memory is a sum of all of the memory which
  647. has been allocated by processes, even if it has not been
  648. "used" by them as of yet. A process which malloc()'s 1G
  649. of memory, but only touches 300M of it will only show up
  650. as using 300M of memory even if it has the address space
  651. allocated for the entire 1G. This 1G is memory which has
  652. been "committed" to by the VM and can be used at any time
  653. by the allocating application. With strict overcommit
  654. enabled on the system (mode 2 in 'vm.overcommit_memory'),
  655. allocations which would exceed the CommitLimit (detailed
  656. above) will not be permitted. This is useful if one needs
  657. to guarantee that processes will not fail due to lack of
  658. memory once that memory has been successfully allocated.
  659. VmallocTotal: total size of vmalloc memory area
  660. VmallocUsed: amount of vmalloc area which is used
  661. VmallocChunk: largest contiguous block of vmalloc area which is free
  662. ..............................................................................
  663. vmallocinfo:
  664. Provides information about vmalloced/vmaped areas. One line per area,
  665. containing the virtual address range of the area, size in bytes,
  666. caller information of the creator, and optional information depending
  667. on the kind of area :
  668. pages=nr number of pages
  669. phys=addr if a physical address was specified
  670. ioremap I/O mapping (ioremap() and friends)
  671. vmalloc vmalloc() area
  672. vmap vmap()ed pages
  673. user VM_USERMAP area
  674. vpages buffer for pages pointers was vmalloced (huge area)
  675. N<node>=nr (Only on NUMA kernels)
  676. Number of pages allocated on memory node <node>
  677. > cat /proc/vmallocinfo
  678. 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
  679. /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
  680. 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
  681. /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
  682. 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
  683. phys=7fee8000 ioremap
  684. 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
  685. phys=7fee7000 ioremap
  686. 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
  687. 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
  688. /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
  689. 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
  690. pages=2 vmalloc N1=2
  691. 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
  692. /0x130 [x_tables] pages=4 vmalloc N0=4
  693. 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
  694. pages=14 vmalloc N2=14
  695. 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
  696. pages=4 vmalloc N1=4
  697. 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
  698. pages=2 vmalloc N1=2
  699. 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
  700. pages=10 vmalloc N0=10
  701. ..............................................................................
  702. softirqs:
  703. Provides counts of softirq handlers serviced since boot time, for each cpu.
  704. > cat /proc/softirqs
  705. CPU0 CPU1 CPU2 CPU3
  706. HI: 0 0 0 0
  707. TIMER: 27166 27120 27097 27034
  708. NET_TX: 0 0 0 17
  709. NET_RX: 42 0 0 39
  710. BLOCK: 0 0 107 1121
  711. TASKLET: 0 0 0 290
  712. SCHED: 27035 26983 26971 26746
  713. HRTIMER: 0 0 0 0
  714. RCU: 1678 1769 2178 2250
  715. 1.3 IDE devices in /proc/ide
  716. ----------------------------
  717. The subdirectory /proc/ide contains information about all IDE devices of which
  718. the kernel is aware. There is one subdirectory for each IDE controller, the
  719. file drivers and a link for each IDE device, pointing to the device directory
  720. in the controller specific subtree.
  721. The file drivers contains general information about the drivers used for the
  722. IDE devices:
  723. > cat /proc/ide/drivers
  724. ide-cdrom version 4.53
  725. ide-disk version 1.08
  726. More detailed information can be found in the controller specific
  727. subdirectories. These are named ide0, ide1 and so on. Each of these
  728. directories contains the files shown in table 1-6.
  729. Table 1-6: IDE controller info in /proc/ide/ide?
  730. ..............................................................................
  731. File Content
  732. channel IDE channel (0 or 1)
  733. config Configuration (only for PCI/IDE bridge)
  734. mate Mate name
  735. model Type/Chipset of IDE controller
  736. ..............................................................................
  737. Each device connected to a controller has a separate subdirectory in the
  738. controllers directory. The files listed in table 1-7 are contained in these
  739. directories.
  740. Table 1-7: IDE device information
  741. ..............................................................................
  742. File Content
  743. cache The cache
  744. capacity Capacity of the medium (in 512Byte blocks)
  745. driver driver and version
  746. geometry physical and logical geometry
  747. identify device identify block
  748. media media type
  749. model device identifier
  750. settings device setup
  751. smart_thresholds IDE disk management thresholds
  752. smart_values IDE disk management values
  753. ..............................................................................
  754. The most interesting file is settings. This file contains a nice overview of
  755. the drive parameters:
  756. # cat /proc/ide/ide0/hda/settings
  757. name value min max mode
  758. ---- ----- --- --- ----
  759. bios_cyl 526 0 65535 rw
  760. bios_head 255 0 255 rw
  761. bios_sect 63 0 63 rw
  762. breada_readahead 4 0 127 rw
  763. bswap 0 0 1 r
  764. file_readahead 72 0 2097151 rw
  765. io_32bit 0 0 3 rw
  766. keepsettings 0 0 1 rw
  767. max_kb_per_request 122 1 127 rw
  768. multcount 0 0 8 rw
  769. nice1 1 0 1 rw
  770. nowerr 0 0 1 rw
  771. pio_mode write-only 0 255 w
  772. slow 0 0 1 rw
  773. unmaskirq 0 0 1 rw
  774. using_dma 0 0 1 rw
  775. 1.4 Networking info in /proc/net
  776. --------------------------------
  777. The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
  778. additional values you get for IP version 6 if you configure the kernel to
  779. support this. Table 1-9 lists the files and their meaning.
  780. Table 1-8: IPv6 info in /proc/net
  781. ..............................................................................
  782. File Content
  783. udp6 UDP sockets (IPv6)
  784. tcp6 TCP sockets (IPv6)
  785. raw6 Raw device statistics (IPv6)
  786. igmp6 IP multicast addresses, which this host joined (IPv6)
  787. if_inet6 List of IPv6 interface addresses
  788. ipv6_route Kernel routing table for IPv6
  789. rt6_stats Global IPv6 routing tables statistics
  790. sockstat6 Socket statistics (IPv6)
  791. snmp6 Snmp data (IPv6)
  792. ..............................................................................
  793. Table 1-9: Network info in /proc/net
  794. ..............................................................................
  795. File Content
  796. arp Kernel ARP table
  797. dev network devices with statistics
  798. dev_mcast the Layer2 multicast groups a device is listening too
  799. (interface index, label, number of references, number of bound
  800. addresses).
  801. dev_stat network device status
  802. ip_fwchains Firewall chain linkage
  803. ip_fwnames Firewall chain names
  804. ip_masq Directory containing the masquerading tables
  805. ip_masquerade Major masquerading table
  806. netstat Network statistics
  807. raw raw device statistics
  808. route Kernel routing table
  809. rpc Directory containing rpc info
  810. rt_cache Routing cache
  811. snmp SNMP data
  812. sockstat Socket statistics
  813. tcp TCP sockets
  814. tr_rif Token ring RIF routing table
  815. udp UDP sockets
  816. unix UNIX domain sockets
  817. wireless Wireless interface data (Wavelan etc)
  818. igmp IP multicast addresses, which this host joined
  819. psched Global packet scheduler parameters.
  820. netlink List of PF_NETLINK sockets
  821. ip_mr_vifs List of multicast virtual interfaces
  822. ip_mr_cache List of multicast routing cache
  823. ..............................................................................
  824. You can use this information to see which network devices are available in
  825. your system and how much traffic was routed over those devices:
  826. > cat /proc/net/dev
  827. Inter-|Receive |[...
  828. face |bytes packets errs drop fifo frame compressed multicast|[...
  829. lo: 908188 5596 0 0 0 0 0 0 [...
  830. ppp0:15475140 20721 410 0 0 410 0 0 [...
  831. eth0: 614530 7085 0 0 0 0 0 1 [...
  832. ...] Transmit
  833. ...] bytes packets errs drop fifo colls carrier compressed
  834. ...] 908188 5596 0 0 0 0 0 0
  835. ...] 1375103 17405 0 0 0 0 0 0
  836. ...] 1703981 5535 0 0 0 3 0 0
  837. In addition, each Channel Bond interface has its own directory. For
  838. example, the bond0 device will have a directory called /proc/net/bond0/.
  839. It will contain information that is specific to that bond, such as the
  840. current slaves of the bond, the link status of the slaves, and how
  841. many times the slaves link has failed.
  842. 1.5 SCSI info
  843. -------------
  844. If you have a SCSI host adapter in your system, you'll find a subdirectory
  845. named after the driver for this adapter in /proc/scsi. You'll also see a list
  846. of all recognized SCSI devices in /proc/scsi:
  847. >cat /proc/scsi/scsi
  848. Attached devices:
  849. Host: scsi0 Channel: 00 Id: 00 Lun: 00
  850. Vendor: IBM Model: DGHS09U Rev: 03E0
  851. Type: Direct-Access ANSI SCSI revision: 03
  852. Host: scsi0 Channel: 00 Id: 06 Lun: 00
  853. Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
  854. Type: CD-ROM ANSI SCSI revision: 02
  855. The directory named after the driver has one file for each adapter found in
  856. the system. These files contain information about the controller, including
  857. the used IRQ and the IO address range. The amount of information shown is
  858. dependent on the adapter you use. The example shows the output for an Adaptec
  859. AHA-2940 SCSI adapter:
  860. > cat /proc/scsi/aic7xxx/0
  861. Adaptec AIC7xxx driver version: 5.1.19/3.2.4
  862. Compile Options:
  863. TCQ Enabled By Default : Disabled
  864. AIC7XXX_PROC_STATS : Disabled
  865. AIC7XXX_RESET_DELAY : 5
  866. Adapter Configuration:
  867. SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
  868. Ultra Wide Controller
  869. PCI MMAPed I/O Base: 0xeb001000
  870. Adapter SEEPROM Config: SEEPROM found and used.
  871. Adaptec SCSI BIOS: Enabled
  872. IRQ: 10
  873. SCBs: Active 0, Max Active 2,
  874. Allocated 15, HW 16, Page 255
  875. Interrupts: 160328
  876. BIOS Control Word: 0x18b6
  877. Adapter Control Word: 0x005b
  878. Extended Translation: Enabled
  879. Disconnect Enable Flags: 0xffff
  880. Ultra Enable Flags: 0x0001
  881. Tag Queue Enable Flags: 0x0000
  882. Ordered Queue Tag Flags: 0x0000
  883. Default Tag Queue Depth: 8
  884. Tagged Queue By Device array for aic7xxx host instance 0:
  885. {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
  886. Actual queue depth per device for aic7xxx host instance 0:
  887. {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
  888. Statistics:
  889. (scsi0:0:0:0)
  890. Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
  891. Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
  892. Total transfers 160151 (74577 reads and 85574 writes)
  893. (scsi0:0:6:0)
  894. Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
  895. Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
  896. Total transfers 0 (0 reads and 0 writes)
  897. 1.6 Parallel port info in /proc/parport
  898. ---------------------------------------
  899. The directory /proc/parport contains information about the parallel ports of
  900. your system. It has one subdirectory for each port, named after the port
  901. number (0,1,2,...).
  902. These directories contain the four files shown in Table 1-10.
  903. Table 1-10: Files in /proc/parport
  904. ..............................................................................
  905. File Content
  906. autoprobe Any IEEE-1284 device ID information that has been acquired.
  907. devices list of the device drivers using that port. A + will appear by the
  908. name of the device currently using the port (it might not appear
  909. against any).
  910. hardware Parallel port's base address, IRQ line and DMA channel.
  911. irq IRQ that parport is using for that port. This is in a separate
  912. file to allow you to alter it by writing a new value in (IRQ
  913. number or none).
  914. ..............................................................................
  915. 1.7 TTY info in /proc/tty
  916. -------------------------
  917. Information about the available and actually used tty's can be found in the
  918. directory /proc/tty.You'll find entries for drivers and line disciplines in
  919. this directory, as shown in Table 1-11.
  920. Table 1-11: Files in /proc/tty
  921. ..............................................................................
  922. File Content
  923. drivers list of drivers and their usage
  924. ldiscs registered line disciplines
  925. driver/serial usage statistic and status of single tty lines
  926. ..............................................................................
  927. To see which tty's are currently in use, you can simply look into the file
  928. /proc/tty/drivers:
  929. > cat /proc/tty/drivers
  930. pty_slave /dev/pts 136 0-255 pty:slave
  931. pty_master /dev/ptm 128 0-255 pty:master
  932. pty_slave /dev/ttyp 3 0-255 pty:slave
  933. pty_master /dev/pty 2 0-255 pty:master
  934. serial /dev/cua 5 64-67 serial:callout
  935. serial /dev/ttyS 4 64-67 serial
  936. /dev/tty0 /dev/tty0 4 0 system:vtmaster
  937. /dev/ptmx /dev/ptmx 5 2 system
  938. /dev/console /dev/console 5 1 system:console
  939. /dev/tty /dev/tty 5 0 system:/dev/tty
  940. unknown /dev/tty 4 1-63 console
  941. 1.8 Miscellaneous kernel statistics in /proc/stat
  942. -------------------------------------------------
  943. Various pieces of information about kernel activity are available in the
  944. /proc/stat file. All of the numbers reported in this file are aggregates
  945. since the system first booted. For a quick look, simply cat the file:
  946. > cat /proc/stat
  947. cpu 2255 34 2290 22625563 6290 127 456 0 0
  948. cpu0 1132 34 1441 11311718 3675 127 438 0 0
  949. cpu1 1123 0 849 11313845 2614 0 18 0 0
  950. intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
  951. ctxt 1990473
  952. btime 1062191376
  953. processes 2915
  954. procs_running 1
  955. procs_blocked 0
  956. softirq 183433 0 21755 12 39 1137 231 21459 2263
  957. The very first "cpu" line aggregates the numbers in all of the other "cpuN"
  958. lines. These numbers identify the amount of time the CPU has spent performing
  959. different kinds of work. Time units are in USER_HZ (typically hundredths of a
  960. second). The meanings of the columns are as follows, from left to right:
  961. - user: normal processes executing in user mode
  962. - nice: niced processes executing in user mode
  963. - system: processes executing in kernel mode
  964. - idle: twiddling thumbs
  965. - iowait: waiting for I/O to complete
  966. - irq: servicing interrupts
  967. - softirq: servicing softirqs
  968. - steal: involuntary wait
  969. - guest: running a normal guest
  970. - guest_nice: running a niced guest
  971. The "intr" line gives counts of interrupts serviced since boot time, for each
  972. of the possible system interrupts. The first column is the total of all
  973. interrupts serviced; each subsequent column is the total for that particular
  974. interrupt.
  975. The "ctxt" line gives the total number of context switches across all CPUs.
  976. The "btime" line gives the time at which the system booted, in seconds since
  977. the Unix epoch.
  978. The "processes" line gives the number of processes and threads created, which
  979. includes (but is not limited to) those created by calls to the fork() and
  980. clone() system calls.
  981. The "procs_running" line gives the total number of threads that are
  982. running or ready to run (i.e., the total number of runnable threads).
  983. The "procs_blocked" line gives the number of processes currently blocked,
  984. waiting for I/O to complete.
  985. The "softirq" line gives counts of softirqs serviced since boot time, for each
  986. of the possible system softirqs. The first column is the total of all
  987. softirqs serviced; each subsequent column is the total for that particular
  988. softirq.
  989. 1.9 Ext4 file system parameters
  990. ------------------------------
  991. Information about mounted ext4 file systems can be found in
  992. /proc/fs/ext4. Each mounted filesystem will have a directory in
  993. /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
  994. /proc/fs/ext4/dm-0). The files in each per-device directory are shown
  995. in Table 1-12, below.
  996. Table 1-12: Files in /proc/fs/ext4/<devname>
  997. ..............................................................................
  998. File Content
  999. mb_groups details of multiblock allocator buddy cache of free blocks
  1000. ..............................................................................
  1001. ------------------------------------------------------------------------------
  1002. Summary
  1003. ------------------------------------------------------------------------------
  1004. The /proc file system serves information about the running system. It not only
  1005. allows access to process data but also allows you to request the kernel status
  1006. by reading files in the hierarchy.
  1007. The directory structure of /proc reflects the types of information and makes
  1008. it easy, if not obvious, where to look for specific data.
  1009. ------------------------------------------------------------------------------
  1010. ------------------------------------------------------------------------------
  1011. CHAPTER 2: MODIFYING SYSTEM PARAMETERS
  1012. ------------------------------------------------------------------------------
  1013. ------------------------------------------------------------------------------
  1014. In This Chapter
  1015. ------------------------------------------------------------------------------
  1016. * Modifying kernel parameters by writing into files found in /proc/sys
  1017. * Exploring the files which modify certain parameters
  1018. * Review of the /proc/sys file tree
  1019. ------------------------------------------------------------------------------
  1020. A very interesting part of /proc is the directory /proc/sys. This is not only
  1021. a source of information, it also allows you to change parameters within the
  1022. kernel. Be very careful when attempting this. You can optimize your system,
  1023. but you can also cause it to crash. Never alter kernel parameters on a
  1024. production system. Set up a development machine and test to make sure that
  1025. everything works the way you want it to. You may have no alternative but to
  1026. reboot the machine once an error has been made.
  1027. To change a value, simply echo the new value into the file. An example is
  1028. given below in the section on the file system data. You need to be root to do
  1029. this. You can create your own boot script to perform this every time your
  1030. system boots.
  1031. The files in /proc/sys can be used to fine tune and monitor miscellaneous and
  1032. general things in the operation of the Linux kernel. Since some of the files
  1033. can inadvertently disrupt your system, it is advisable to read both
  1034. documentation and source before actually making adjustments. In any case, be
  1035. very careful when writing to any of these files. The entries in /proc may
  1036. change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
  1037. review the kernel documentation in the directory /usr/src/linux/Documentation.
  1038. This chapter is heavily based on the documentation included in the pre 2.2
  1039. kernels, and became part of it in version 2.2.1 of the Linux kernel.
  1040. Please see: Documentation/sysctls/ directory for descriptions of these
  1041. entries.
  1042. ------------------------------------------------------------------------------
  1043. Summary
  1044. ------------------------------------------------------------------------------
  1045. Certain aspects of kernel behavior can be modified at runtime, without the
  1046. need to recompile the kernel, or even to reboot the system. The files in the
  1047. /proc/sys tree can not only be read, but also modified. You can use the echo
  1048. command to write value into these files, thereby changing the default settings
  1049. of the kernel.
  1050. ------------------------------------------------------------------------------
  1051. ------------------------------------------------------------------------------
  1052. CHAPTER 3: PER-PROCESS PARAMETERS
  1053. ------------------------------------------------------------------------------
  1054. 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
  1055. --------------------------------------------------------------------------------
  1056. These file can be used to adjust the badness heuristic used to select which
  1057. process gets killed in out of memory conditions.
  1058. The badness heuristic assigns a value to each candidate task ranging from 0
  1059. (never kill) to 1000 (always kill) to determine which process is targeted. The
  1060. units are roughly a proportion along that range of allowed memory the process
  1061. may allocate from based on an estimation of its current memory and swap use.
  1062. For example, if a task is using all allowed memory, its badness score will be
  1063. 1000. If it is using half of its allowed memory, its score will be 500.
  1064. There is an additional factor included in the badness score: root
  1065. processes are given 3% extra memory over other tasks.
  1066. The amount of "allowed" memory depends on the context in which the oom killer
  1067. was called. If it is due to the memory assigned to the allocating task's cpuset
  1068. being exhausted, the allowed memory represents the set of mems assigned to that
  1069. cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
  1070. memory represents the set of mempolicy nodes. If it is due to a memory
  1071. limit (or swap limit) being reached, the allowed memory is that configured
  1072. limit. Finally, if it is due to the entire system being out of memory, the
  1073. allowed memory represents all allocatable resources.
  1074. The value of /proc/<pid>/oom_score_adj is added to the badness score before it
  1075. is used to determine which task to kill. Acceptable values range from -1000
  1076. (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
  1077. polarize the preference for oom killing either by always preferring a certain
  1078. task or completely disabling it. The lowest possible value, -1000, is
  1079. equivalent to disabling oom killing entirely for that task since it will always
  1080. report a badness score of 0.
  1081. Consequently, it is very simple for userspace to define the amount of memory to
  1082. consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
  1083. example, is roughly equivalent to allowing the remainder of tasks sharing the
  1084. same system, cpuset, mempolicy, or memory controller resources to use at least
  1085. 50% more memory. A value of -500, on the other hand, would be roughly
  1086. equivalent to discounting 50% of the task's allowed memory from being considered
  1087. as scoring against the task.
  1088. For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
  1089. be used to tune the badness score. Its acceptable values range from -16
  1090. (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
  1091. (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
  1092. scaled linearly with /proc/<pid>/oom_score_adj.
  1093. Writing to /proc/<pid>/oom_score_adj or /proc/<pid>/oom_adj will change the
  1094. other with its scaled value.
  1095. NOTICE: /proc/<pid>/oom_adj is deprecated and will be removed, please see
  1096. Documentation/feature-removal-schedule.txt.
  1097. Caveat: when a parent task is selected, the oom killer will sacrifice any first
  1098. generation children with seperate address spaces instead, if possible. This
  1099. avoids servers and important system daemons from being killed and loses the
  1100. minimal amount of work.
  1101. 3.2 /proc/<pid>/oom_score - Display current oom-killer score
  1102. -------------------------------------------------------------
  1103. This file can be used to check the current score used by the oom-killer is for
  1104. any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
  1105. process should be killed in an out-of-memory situation.
  1106. 3.3 /proc/<pid>/io - Display the IO accounting fields
  1107. -------------------------------------------------------
  1108. This file contains IO statistics for each running process
  1109. Example
  1110. -------
  1111. test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
  1112. [1] 3828
  1113. test:/tmp # cat /proc/3828/io
  1114. rchar: 323934931
  1115. wchar: 323929600
  1116. syscr: 632687
  1117. syscw: 632675
  1118. read_bytes: 0
  1119. write_bytes: 323932160
  1120. cancelled_write_bytes: 0
  1121. Description
  1122. -----------
  1123. rchar
  1124. -----
  1125. I/O counter: chars read
  1126. The number of bytes which this task has caused to be read from storage. This
  1127. is simply the sum of bytes which this process passed to read() and pread().
  1128. It includes things like tty IO and it is unaffected by whether or not actual
  1129. physical disk IO was required (the read might have been satisfied from
  1130. pagecache)
  1131. wchar
  1132. -----
  1133. I/O counter: chars written
  1134. The number of bytes which this task has caused, or shall cause to be written
  1135. to disk. Similar caveats apply here as with rchar.
  1136. syscr
  1137. -----
  1138. I/O counter: read syscalls
  1139. Attempt to count the number of read I/O operations, i.e. syscalls like read()
  1140. and pread().
  1141. syscw
  1142. -----
  1143. I/O counter: write syscalls
  1144. Attempt to count the number of write I/O operations, i.e. syscalls like
  1145. write() and pwrite().
  1146. read_bytes
  1147. ----------
  1148. I/O counter: bytes read
  1149. Attempt to count the number of bytes which this process really did cause to
  1150. be fetched from the storage layer. Done at the submit_bio() level, so it is
  1151. accurate for block-backed filesystems. <please add status regarding NFS and
  1152. CIFS at a later time>
  1153. write_bytes
  1154. -----------
  1155. I/O counter: bytes written
  1156. Attempt to count the number of bytes which this process caused to be sent to
  1157. the storage layer. This is done at page-dirtying time.
  1158. cancelled_write_bytes
  1159. ---------------------
  1160. The big inaccuracy here is truncate. If a process writes 1MB to a file and
  1161. then deletes the file, it will in fact perform no writeout. But it will have
  1162. been accounted as having caused 1MB of write.
  1163. In other words: The number of bytes which this process caused to not happen,
  1164. by truncating pagecache. A task can cause "negative" IO too. If this task
  1165. truncates some dirty pagecache, some IO which another task has been accounted
  1166. for (in its write_bytes) will not be happening. We _could_ just subtract that
  1167. from the truncating task's write_bytes, but there is information loss in doing
  1168. that.
  1169. Note
  1170. ----
  1171. At its current implementation state, this is a bit racy on 32-bit machines: if
  1172. process A reads process B's /proc/pid/io while process B is updating one of
  1173. those 64-bit counters, process A could see an intermediate result.
  1174. More information about this can be found within the taskstats documentation in
  1175. Documentation/accounting.
  1176. 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
  1177. ---------------------------------------------------------------
  1178. When a process is dumped, all anonymous memory is written to a core file as
  1179. long as the size of the core file isn't limited. But sometimes we don't want
  1180. to dump some memory segments, for example, huge shared memory. Conversely,
  1181. sometimes we want to save file-backed memory segments into a core file, not
  1182. only the individual files.
  1183. /proc/<pid>/coredump_filter allows you to customize which memory segments
  1184. will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
  1185. of memory types. If a bit of the bitmask is set, memory segments of the
  1186. corresponding memory type are dumped, otherwise they are not dumped.
  1187. The following 7 memory types are supported:
  1188. - (bit 0) anonymous private memory
  1189. - (bit 1) anonymous shared memory
  1190. - (bit 2) file-backed private memory
  1191. - (bit 3) file-backed shared memory
  1192. - (bit 4) ELF header pages in file-backed private memory areas (it is
  1193. effective only if the bit 2 is cleared)
  1194. - (bit 5) hugetlb private memory
  1195. - (bit 6) hugetlb shared memory
  1196. Note that MMIO pages such as frame buffer are never dumped and vDSO pages
  1197. are always dumped regardless of the bitmask status.
  1198. Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
  1199. effected by bit 5-6.
  1200. Default value of coredump_filter is 0x23; this means all anonymous memory
  1201. segments and hugetlb private memory are dumped.
  1202. If you don't want to dump all shared memory segments attached to pid 1234,
  1203. write 0x21 to the process's proc file.
  1204. $ echo 0x21 > /proc/1234/coredump_filter
  1205. When a new process is created, the process inherits the bitmask status from its
  1206. parent. It is useful to set up coredump_filter before the program runs.
  1207. For example:
  1208. $ echo 0x7 > /proc/self/coredump_filter
  1209. $ ./some_program
  1210. 3.5 /proc/<pid>/mountinfo - Information about mounts
  1211. --------------------------------------------------------
  1212. This file contains lines of the form:
  1213. 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
  1214. (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
  1215. (1) mount ID: unique identifier of the mount (may be reused after umount)
  1216. (2) parent ID: ID of parent (or of self for the top of the mount tree)
  1217. (3) major:minor: value of st_dev for files on filesystem
  1218. (4) root: root of the mount within the filesystem
  1219. (5) mount point: mount point relative to the process's root
  1220. (6) mount options: per mount options
  1221. (7) optional fields: zero or more fields of the form "tag[:value]"
  1222. (8) separator: marks the end of the optional fields
  1223. (9) filesystem type: name of filesystem of the form "type[.subtype]"
  1224. (10) mount source: filesystem specific information or "none"
  1225. (11) super options: per super block options
  1226. Parsers should ignore all unrecognised optional fields. Currently the
  1227. possible optional fields are:
  1228. shared:X mount is shared in peer group X
  1229. master:X mount is slave to peer group X
  1230. propagate_from:X mount is slave and receives propagation from peer group X (*)
  1231. unbindable mount is unbindable
  1232. (*) X is the closest dominant peer group under the process's root. If
  1233. X is the immediate master of the mount, or if there's no dominant peer
  1234. group under the same root, then only the "master:X" field is present
  1235. and not the "propagate_from:X" field.
  1236. For more information on mount propagation see:
  1237. Documentation/filesystems/sharedsubtree.txt
  1238. 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
  1239. --------------------------------------------------------
  1240. These files provide a method to access a tasks comm value. It also allows for
  1241. a task to set its own or one of its thread siblings comm value. The comm value
  1242. is limited in size compared to the cmdline value, so writing anything longer
  1243. then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
  1244. comm value.