| blob | 220cc6376ef80e0c9bcfec162d45552e729cdf5a |
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
15 Table of Contents
16 -----------------
18 0 Preface
19 0.1 Introduction/Credits
20 0.2 Legal Stuff
22 1 Collecting System Information
23 1.1 Process-Specific Subdirectories
24 1.2 Kernel data
25 1.3 IDE devices in /proc/ide
26 1.4 Networking info in /proc/net
27 1.5 SCSI info
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 - Adjust the oom-killer score
37 3.2 /proc/<pid>/oom_score - Display current oom-killer score
38 3.3 /proc/<pid>/io - Display the IO accounting fields
39 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
40 3.5 /proc/<pid>/mountinfo - Information about mounts
41 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
44 ------------------------------------------------------------------------------
45 Preface
46 ------------------------------------------------------------------------------
48 0.1 Introduction/Credits
49 ------------------------
51 This documentation is part of a soon (or so we hope) to be released book on
52 the SuSE Linux distribution. As there is no complete documentation for the
53 /proc file system and we've used many freely available sources to write these
54 chapters, it seems only fair to give the work back to the Linux community.
55 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
56 afraid it's still far from complete, but we hope it will be useful. As far as
57 we know, it is the first 'all-in-one' document about the /proc file system. It
58 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
59 SPARC, AXP, etc., features, you probably won't find what you are looking for.
60 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
61 additions and patches are welcome and will be added to this document if you
62 mail them to Bodo.
64 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
65 other people for help compiling this documentation. We'd also like to extend a
66 special thank you to Andi Kleen for documentation, which we relied on heavily
67 to create this document, as well as the additional information he provided.
68 Thanks to everybody else who contributed source or docs to the Linux kernel
69 and helped create a great piece of software... :)
71 If you have any comments, corrections or additions, please don't hesitate to
72 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
73 document.
75 The latest version of this document is available online at
76 http://skaro.nightcrawler.com/~bb/Docs/Proc as HTML version.
78 If the above direction does not works for you, ypu could try the kernel
79 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
80 comandante@zaralinux.com.
82 0.2 Legal Stuff
83 ---------------
85 We don't guarantee the correctness of this document, and if you come to us
86 complaining about how you screwed up your system because of incorrect
87 documentation, we won't feel responsible...
89 ------------------------------------------------------------------------------
90 CHAPTER 1: COLLECTING SYSTEM INFORMATION
91 ------------------------------------------------------------------------------
93 ------------------------------------------------------------------------------
94 In This Chapter
95 ------------------------------------------------------------------------------
96 * Investigating the properties of the pseudo file system /proc and its
97 ability to provide information on the running Linux system
98 * Examining /proc's structure
99 * Uncovering various information about the kernel and the processes running
100 on the system
101 ------------------------------------------------------------------------------
104 The proc file system acts as an interface to internal data structures in the
105 kernel. It can be used to obtain information about the system and to change
106 certain kernel parameters at runtime (sysctl).
108 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
109 show you how you can use /proc/sys to change settings.
111 1.1 Process-Specific Subdirectories
112 -----------------------------------
114 The directory /proc contains (among other things) one subdirectory for each
115 process running on the system, which is named after the process ID (PID).
117 The link self points to the process reading the file system. Each process
118 subdirectory has the entries listed in Table 1-1.
121 Table 1-1: Process specific entries in /proc
122 ..............................................................................
123 File Content
124 clear_refs Clears page referenced bits shown in smaps output
125 cmdline Command line arguments
126 cpu Current and last cpu in which it was executed (2.4)(smp)
127 cwd Link to the current working directory
128 environ Values of environment variables
129 exe Link to the executable of this process
130 fd Directory, which contains all file descriptors
131 maps Memory maps to executables and library files (2.4)
132 mem Memory held by this process
133 root Link to the root directory of this process
134 stat Process status
135 statm Process memory status information
136 status Process status in human readable form
137 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
138 stack Report full stack trace, enable via CONFIG_STACKTRACE
139 smaps a extension based on maps, showing the memory consumption of
140 each mapping
141 ..............................................................................
143 For example, to get the status information of a process, all you have to do is
144 read the file /proc/PID/status:
146 >cat /proc/self/status
147 Name: cat
148 State: R (running)
149 Tgid: 5452
150 Pid: 5452
151 PPid: 743
152 TracerPid: 0 (2.4)
153 Uid: 501 501 501 501
154 Gid: 100 100 100 100
155 FDSize: 256
156 Groups: 100 14 16
157 VmPeak: 5004 kB
158 VmSize: 5004 kB
159 VmLck: 0 kB
160 VmHWM: 476 kB
161 VmRSS: 476 kB
162 VmData: 156 kB
163 VmStk: 88 kB
164 VmExe: 68 kB
165 VmLib: 1412 kB
166 VmPTE: 20 kb
167 Threads: 1
168 SigQ: 0/28578
169 SigPnd: 0000000000000000
170 ShdPnd: 0000000000000000
171 SigBlk: 0000000000000000
172 SigIgn: 0000000000000000
173 SigCgt: 0000000000000000
174 CapInh: 00000000fffffeff
175 CapPrm: 0000000000000000
176 CapEff: 0000000000000000
177 CapBnd: ffffffffffffffff
178 voluntary_ctxt_switches: 0
179 nonvoluntary_ctxt_switches: 1
180 Stack usage: 12 kB
182 This shows you nearly the same information you would get if you viewed it with
183 the ps command. In fact, ps uses the proc file system to obtain its
184 information. But you get a more detailed view of the process by reading the
185 file /proc/PID/status. It fields are described in table 1-2.
187 The statm file contains more detailed information about the process
188 memory usage. Its seven fields are explained in Table 1-3. The stat file
189 contains details information about the process itself. Its fields are
190 explained in Table 1-4.
192 Table 1-2: Contents of the statm files (as of 2.6.30-rc7)
193 ..............................................................................
194 Field Content
195 Name filename of the executable
196 State state (R is running, S is sleeping, D is sleeping
197 in an uninterruptible wait, Z is zombie,
198 T is traced or stopped)
199 Tgid thread group ID
200 Pid process id
201 PPid process id of the parent process
202 TracerPid PID of process tracing this process (0 if not)
203 Uid Real, effective, saved set, and file system UIDs
204 Gid Real, effective, saved set, and file system GIDs
205 FDSize number of file descriptor slots currently allocated
206 Groups supplementary group list
207 VmPeak peak virtual memory size
208 VmSize total program size
209 VmLck locked memory size
210 VmHWM peak resident set size ("high water mark")
211 VmRSS size of memory portions
212 VmData size of data, stack, and text segments
213 VmStk size of data, stack, and text segments
214 VmExe size of text segment
215 VmLib size of shared library code
216 VmPTE size of page table entries
217 Threads number of threads
218 SigQ number of signals queued/max. number for queue
219 SigPnd bitmap of pending signals for the thread
220 ShdPnd bitmap of shared pending signals for the process
221 SigBlk bitmap of blocked signals
222 SigIgn bitmap of ignored signals
223 SigCgt bitmap of catched signals
224 CapInh bitmap of inheritable capabilities
225 CapPrm bitmap of permitted capabilities
226 CapEff bitmap of effective capabilities
227 CapBnd bitmap of capabilities bounding set
228 Cpus_allowed mask of CPUs on which this process may run
229 Cpus_allowed_list Same as previous, but in "list format"
230 Mems_allowed mask of memory nodes allowed to this process
231 Mems_allowed_list Same as previous, but in "list format"
232 voluntary_ctxt_switches number of voluntary context switches
233 nonvoluntary_ctxt_switches number of non voluntary context switches
234 Stack usage: stack usage high water mark (round up to page size)
235 ..............................................................................
237 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
238 ..............................................................................
239 Field Content
240 size total program size (pages) (same as VmSize in status)
241 resident size of memory portions (pages) (same as VmRSS in status)
242 shared number of pages that are shared (i.e. backed by a file)
243 trs number of pages that are 'code' (not including libs; broken,
244 includes data segment)
245 lrs number of pages of library (always 0 on 2.6)
246 drs number of pages of data/stack (including libs; broken,
247 includes library text)
248 dt number of dirty pages (always 0 on 2.6)
249 ..............................................................................
252 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
253 ..............................................................................
254 Field Content
255 pid process id
256 tcomm filename of the executable
257 state state (R is running, S is sleeping, D is sleeping in an
258 uninterruptible wait, Z is zombie, T is traced or stopped)
259 ppid process id of the parent process
260 pgrp pgrp of the process
261 sid session id
262 tty_nr tty the process uses
263 tty_pgrp pgrp of the tty
264 flags task flags
265 min_flt number of minor faults
266 cmin_flt number of minor faults with child's
267 maj_flt number of major faults
268 cmaj_flt number of major faults with child's
269 utime user mode jiffies
270 stime kernel mode jiffies
271 cutime user mode jiffies with child's
272 cstime kernel mode jiffies with child's
273 priority priority level
274 nice nice level
275 num_threads number of threads
276 it_real_value (obsolete, always 0)
277 start_time time the process started after system boot
278 vsize virtual memory size
279 rss resident set memory size
280 rsslim current limit in bytes on the rss
281 start_code address above which program text can run
282 end_code address below which program text can run
283 start_stack address of the start of the stack
284 esp current value of ESP
285 eip current value of EIP
286 pending bitmap of pending signals
287 blocked bitmap of blocked signals
288 sigign bitmap of ignored signals
289 sigcatch bitmap of catched signals
290 wchan address where process went to sleep
291 0 (place holder)
292 0 (place holder)
293 exit_signal signal to send to parent thread on exit
294 task_cpu which CPU the task is scheduled on
295 rt_priority realtime priority
296 policy scheduling policy (man sched_setscheduler)
297 blkio_ticks time spent waiting for block IO
298 gtime guest time of the task in jiffies
299 cgtime guest time of the task children in jiffies
300 ..............................................................................
302 The /proc/PID/map file containing the currently mapped memory regions and
303 their access permissions.
305 The format is:
307 address perms offset dev inode pathname
309 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
310 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
311 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
312 a7cb1000-a7cb2000 ---p 00000000 00:00 0
313 a7cb2000-a7eb2000 rw-p 00000000 00:00 0 [threadstack:001ff4b4]
314 a7eb2000-a7eb3000 ---p 00000000 00:00 0
315 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
316 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
317 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
318 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
319 a800b000-a800e000 rw-p 00000000 00:00 0
320 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
321 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
322 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
323 a8024000-a8027000 rw-p 00000000 00:00 0
324 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
325 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
326 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
327 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
328 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
330 where "address" is the address space in the process that it occupies, "perms"
331 is a set of permissions:
333 r = read
334 w = write
335 x = execute
336 s = shared
337 p = private (copy on write)
339 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
340 "inode" is the inode on that device. 0 indicates that no inode is associated
341 with the memory region, as the case would be with BSS (uninitialized data).
342 The "pathname" shows the name associated file for this mapping. If the mapping
343 is not associated with a file:
345 [heap] = the heap of the program
346 [stack] = the stack of the main process
347 [vdso] = the "virtual dynamic shared object",
348 the kernel system call handler
349 [threadstack:xxxxxxxx] = the stack of the thread, xxxxxxxx is the stack size
351 or if empty, the mapping is anonymous.
354 The /proc/PID/smaps is an extension based on maps, showing the memory
355 consumption for each of the process's mappings. For each of mappings there
356 is a series of lines such as the following:
358 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
359 Size: 1084 kB
360 Rss: 892 kB
361 Pss: 374 kB
362 Shared_Clean: 892 kB
363 Shared_Dirty: 0 kB
364 Private_Clean: 0 kB
365 Private_Dirty: 0 kB
366 Referenced: 892 kB
367 Swap: 0 kB
368 KernelPageSize: 4 kB
369 MMUPageSize: 4 kB
371 The first of these lines shows the same information as is displayed for the
372 mapping in /proc/PID/maps. The remaining lines show the size of the mapping,
373 the amount of the mapping that is currently resident in RAM, the "proportional
374 set size” (divide each shared page by the number of processes sharing it), the
375 number of clean and dirty shared pages in the mapping, and the number of clean
376 and dirty private pages in the mapping. The "Referenced" indicates the amount
377 of memory currently marked as referenced or accessed.
379 This file is only present if the CONFIG_MMU kernel configuration option is
380 enabled.
382 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
383 bits on both physical and virtual pages associated with a process.
384 To clear the bits for all the pages associated with the process
385 > echo 1 > /proc/PID/clear_refs
387 To clear the bits for the anonymous pages associated with the process
388 > echo 2 > /proc/PID/clear_refs
390 To clear the bits for the file mapped pages associated with the process
391 > echo 3 > /proc/PID/clear_refs
392 Any other value written to /proc/PID/clear_refs will have no effect.
395 1.2 Kernel data
396 ---------------
398 Similar to the process entries, the kernel data files give information about
399 the running kernel. The files used to obtain this information are contained in
400 /proc and are listed in Table 1-5. Not all of these will be present in your
401 system. It depends on the kernel configuration and the loaded modules, which
402 files are there, and which are missing.
404 Table 1-5: Kernel info in /proc
405 ..............................................................................
406 File Content
407 apm Advanced power management info
408 buddyinfo Kernel memory allocator information (see text) (2.5)
409 bus Directory containing bus specific information
410 cmdline Kernel command line
411 cpuinfo Info about the CPU
412 devices Available devices (block and character)
413 dma Used DMS channels
414 filesystems Supported filesystems
415 driver Various drivers grouped here, currently rtc (2.4)
416 execdomains Execdomains, related to security (2.4)
417 fb Frame Buffer devices (2.4)
418 fs File system parameters, currently nfs/exports (2.4)
419 ide Directory containing info about the IDE subsystem
420 interrupts Interrupt usage
421 iomem Memory map (2.4)
422 ioports I/O port usage
423 irq Masks for irq to cpu affinity (2.4)(smp?)
424 isapnp ISA PnP (Plug&Play) Info (2.4)
425 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
426 kmsg Kernel messages
427 ksyms Kernel symbol table
428 loadavg Load average of last 1, 5 & 15 minutes
429 locks Kernel locks
430 meminfo Memory info
431 misc Miscellaneous
432 modules List of loaded modules
433 mounts Mounted filesystems
434 net Networking info (see text)
435 partitions Table of partitions known to the system
436 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
437 decoupled by lspci (2.4)
438 rtc Real time clock
439 scsi SCSI info (see text)
440 slabinfo Slab pool info
441 softirqs softirq usage
442 stat Overall statistics
443 swaps Swap space utilization
444 sys See chapter 2
445 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
446 tty Info of tty drivers
447 uptime System uptime
448 version Kernel version
449 video bttv info of video resources (2.4)
450 vmallocinfo Show vmalloced areas
451 ..............................................................................
453 You can, for example, check which interrupts are currently in use and what
454 they are used for by looking in the file /proc/interrupts:
456 > cat /proc/interrupts
457 CPU0
458 0: 8728810 XT-PIC timer
459 1: 895 XT-PIC keyboard
460 2: 0 XT-PIC cascade
461 3: 531695 XT-PIC aha152x
462 4: 2014133 XT-PIC serial
463 5: 44401 XT-PIC pcnet_cs
464 8: 2 XT-PIC rtc
465 11: 8 XT-PIC i82365
466 12: 182918 XT-PIC PS/2 Mouse
467 13: 1 XT-PIC fpu
468 14: 1232265 XT-PIC ide0
469 15: 7 XT-PIC ide1
470 NMI: 0
472 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
473 output of a SMP machine):
475 > cat /proc/interrupts
477 CPU0 CPU1
478 0: 1243498 1214548 IO-APIC-edge timer
479 1: 8949 8958 IO-APIC-edge keyboard
480 2: 0 0 XT-PIC cascade
481 5: 11286 10161 IO-APIC-edge soundblaster
482 8: 1 0 IO-APIC-edge rtc
483 9: 27422 27407 IO-APIC-edge 3c503
484 12: 113645 113873 IO-APIC-edge PS/2 Mouse
485 13: 0 0 XT-PIC fpu
486 14: 22491 24012 IO-APIC-edge ide0
487 15: 2183 2415 IO-APIC-edge ide1
488 17: 30564 30414 IO-APIC-level eth0
489 18: 177 164 IO-APIC-level bttv
490 NMI: 2457961 2457959
491 LOC: 2457882 2457881
492 ERR: 2155
494 NMI is incremented in this case because every timer interrupt generates a NMI
495 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
497 LOC is the local interrupt counter of the internal APIC of every CPU.
499 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
500 connects the CPUs in a SMP system. This means that an error has been detected,
501 the IO-APIC automatically retry the transmission, so it should not be a big
502 problem, but you should read the SMP-FAQ.
504 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
505 /proc/interrupts to display every IRQ vector in use by the system, not
506 just those considered 'most important'. The new vectors are:
508 THR -- interrupt raised when a machine check threshold counter
509 (typically counting ECC corrected errors of memory or cache) exceeds
510 a configurable threshold. Only available on some systems.
512 TRM -- a thermal event interrupt occurs when a temperature threshold
513 has been exceeded for the CPU. This interrupt may also be generated
514 when the temperature drops back to normal.
516 SPU -- a spurious interrupt is some interrupt that was raised then lowered
517 by some IO device before it could be fully processed by the APIC. Hence
518 the APIC sees the interrupt but does not know what device it came from.
519 For this case the APIC will generate the interrupt with a IRQ vector
520 of 0xff. This might also be generated by chipset bugs.
522 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
523 sent from one CPU to another per the needs of the OS. Typically,
524 their statistics are used by kernel developers and interested users to
525 determine the occurrence of interrupts of the given type.
527 The above IRQ vectors are displayed only when relevent. For example,
528 the threshold vector does not exist on x86_64 platforms. Others are
529 suppressed when the system is a uniprocessor. As of this writing, only
530 i386 and x86_64 platforms support the new IRQ vector displays.
532 Of some interest is the introduction of the /proc/irq directory to 2.4.
533 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
534 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
535 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
536 prof_cpu_mask.
538 For example
539 > ls /proc/irq/
540 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
541 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
542 > ls /proc/irq/0/
543 smp_affinity
545 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
546 IRQ, you can set it by doing:
548 > echo 1 > /proc/irq/10/smp_affinity
550 This means that only the first CPU will handle the IRQ, but you can also echo
551 5 which means that only the first and fourth CPU can handle the IRQ.
553 The contents of each smp_affinity file is the same by default:
555 > cat /proc/irq/0/smp_affinity
556 ffffffff
558 The default_smp_affinity mask applies to all non-active IRQs, which are the
559 IRQs which have not yet been allocated/activated, and hence which lack a
560 /proc/irq/[0-9]* directory.
562 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
563 profiler. Default value is ffffffff (all cpus).
565 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
566 between all the CPUs which are allowed to handle it. As usual the kernel has
567 more info than you and does a better job than you, so the defaults are the
568 best choice for almost everyone.
570 There are three more important subdirectories in /proc: net, scsi, and sys.
571 The general rule is that the contents, or even the existence of these
572 directories, depend on your kernel configuration. If SCSI is not enabled, the
573 directory scsi may not exist. The same is true with the net, which is there
574 only when networking support is present in the running kernel.
576 The slabinfo file gives information about memory usage at the slab level.
577 Linux uses slab pools for memory management above page level in version 2.2.
578 Commonly used objects have their own slab pool (such as network buffers,
579 directory cache, and so on).
581 ..............................................................................
583 > cat /proc/buddyinfo
585 Node 0, zone DMA 0 4 5 4 4 3 ...
586 Node 0, zone Normal 1 0 0 1 101 8 ...
587 Node 0, zone HighMem 2 0 0 1 1 0 ...
589 Memory fragmentation is a problem under some workloads, and buddyinfo is a
590 useful tool for helping diagnose these problems. Buddyinfo will give you a
591 clue as to how big an area you can safely allocate, or why a previous
592 allocation failed.
594 Each column represents the number of pages of a certain order which are
595 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
596 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
597 available in ZONE_NORMAL, etc...
599 ..............................................................................
601 meminfo:
603 Provides information about distribution and utilization of memory. This
604 varies by architecture and compile options. The following is from a
605 16GB PIII, which has highmem enabled. You may not have all of these fields.
607 > cat /proc/meminfo
610 MemTotal: 16344972 kB
611 MemFree: 13634064 kB
612 Buffers: 3656 kB
613 Cached: 1195708 kB
614 SwapCached: 0 kB
615 Active: 891636 kB
616 Inactive: 1077224 kB
617 HighTotal: 15597528 kB
618 HighFree: 13629632 kB
619 LowTotal: 747444 kB
620 LowFree: 4432 kB
621 SwapTotal: 0 kB
622 SwapFree: 0 kB
623 Dirty: 968 kB
624 Writeback: 0 kB
625 AnonPages: 861800 kB
626 Mapped: 280372 kB
627 Slab: 284364 kB
628 SReclaimable: 159856 kB
629 SUnreclaim: 124508 kB
630 PageTables: 24448 kB
631 NFS_Unstable: 0 kB
632 Bounce: 0 kB
633 WritebackTmp: 0 kB
634 CommitLimit: 7669796 kB
635 Committed_AS: 100056 kB
636 VmallocTotal: 112216 kB
637 VmallocUsed: 428 kB
638 VmallocChunk: 111088 kB
640 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
641 bits and the kernel binary code)
642 MemFree: The sum of LowFree+HighFree
643 Buffers: Relatively temporary storage for raw disk blocks
644 shouldn't get tremendously large (20MB or so)
645 Cached: in-memory cache for files read from the disk (the
646 pagecache). Doesn't include SwapCached
647 SwapCached: Memory that once was swapped out, is swapped back in but
648 still also is in the swapfile (if memory is needed it
649 doesn't need to be swapped out AGAIN because it is already
650 in the swapfile. This saves I/O)
651 Active: Memory that has been used more recently and usually not
652 reclaimed unless absolutely necessary.
653 Inactive: Memory which has been less recently used. It is more
654 eligible to be reclaimed for other purposes
655 HighTotal:
656 HighFree: Highmem is all memory above ~860MB of physical memory
657 Highmem areas are for use by userspace programs, or
658 for the pagecache. The kernel must use tricks to access
659 this memory, making it slower to access than lowmem.
660 LowTotal:
661 LowFree: Lowmem is memory which can be used for everything that
662 highmem can be used for, but it is also available for the
663 kernel's use for its own data structures. Among many
664 other things, it is where everything from the Slab is
665 allocated. Bad things happen when you're out of lowmem.
666 SwapTotal: total amount of swap space available
667 SwapFree: Memory which has been evicted from RAM, and is temporarily
668 on the disk
669 Dirty: Memory which is waiting to get written back to the disk
670 Writeback: Memory which is actively being written back to the disk
671 AnonPages: Non-file backed pages mapped into userspace page tables
672 Mapped: files which have been mmaped, such as libraries
673 Slab: in-kernel data structures cache
674 SReclaimable: Part of Slab, that might be reclaimed, such as caches
675 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
676 PageTables: amount of memory dedicated to the lowest level of page
677 tables.
678 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
679 storage
680 Bounce: Memory used for block device "bounce buffers"
681 WritebackTmp: Memory used by FUSE for temporary writeback buffers
682 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
683 this is the total amount of memory currently available to
684 be allocated on the system. This limit is only adhered to
685 if strict overcommit accounting is enabled (mode 2 in
686 'vm.overcommit_memory').
687 The CommitLimit is calculated with the following formula:
688 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
689 For example, on a system with 1G of physical RAM and 7G
690 of swap with a `vm.overcommit_ratio` of 30 it would
691 yield a CommitLimit of 7.3G.
692 For more details, see the memory overcommit documentation
693 in vm/overcommit-accounting.
694 Committed_AS: The amount of memory presently allocated on the system.
695 The committed memory is a sum of all of the memory which
696 has been allocated by processes, even if it has not been
697 "used" by them as of yet. A process which malloc()'s 1G
698 of memory, but only touches 300M of it will only show up
699 as using 300M of memory even if it has the address space
700 allocated for the entire 1G. This 1G is memory which has
701 been "committed" to by the VM and can be used at any time
702 by the allocating application. With strict overcommit
703 enabled on the system (mode 2 in 'vm.overcommit_memory'),
704 allocations which would exceed the CommitLimit (detailed
705 above) will not be permitted. This is useful if one needs
706 to guarantee that processes will not fail due to lack of
707 memory once that memory has been successfully allocated.
708 VmallocTotal: total size of vmalloc memory area
709 VmallocUsed: amount of vmalloc area which is used
710 VmallocChunk: largest contiguous block of vmalloc area which is free
712 ..............................................................................
714 vmallocinfo:
716 Provides information about vmalloced/vmaped areas. One line per area,
717 containing the virtual address range of the area, size in bytes,
718 caller information of the creator, and optional information depending
719 on the kind of area :
721 pages=nr number of pages
722 phys=addr if a physical address was specified
723 ioremap I/O mapping (ioremap() and friends)
724 vmalloc vmalloc() area
725 vmap vmap()ed pages
726 user VM_USERMAP area
727 vpages buffer for pages pointers was vmalloced (huge area)
728 N<node>=nr (Only on NUMA kernels)
729 Number of pages allocated on memory node <node>
731 > cat /proc/vmallocinfo
732 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
733 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
734 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
735 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
736 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
737 phys=7fee8000 ioremap
738 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
739 phys=7fee7000 ioremap
740 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
741 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
742 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
743 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
744 pages=2 vmalloc N1=2
745 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
746 /0x130 [x_tables] pages=4 vmalloc N0=4
747 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
748 pages=14 vmalloc N2=14
749 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
750 pages=4 vmalloc N1=4
751 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
752 pages=2 vmalloc N1=2
753 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
754 pages=10 vmalloc N0=10
756 ..............................................................................
758 softirqs:
760 Provides counts of softirq handlers serviced since boot time, for each cpu.
762 > cat /proc/softirqs
763 CPU0 CPU1 CPU2 CPU3
764 HI: 0 0 0 0
765 TIMER: 27166 27120 27097 27034
766 NET_TX: 0 0 0 17
767 NET_RX: 42 0 0 39
768 BLOCK: 0 0 107 1121
769 TASKLET: 0 0 0 290
770 SCHED: 27035 26983 26971 26746
771 HRTIMER: 0 0 0 0
772 RCU: 1678 1769 2178 2250
775 1.3 IDE devices in /proc/ide
776 ----------------------------
778 The subdirectory /proc/ide contains information about all IDE devices of which
779 the kernel is aware. There is one subdirectory for each IDE controller, the
780 file drivers and a link for each IDE device, pointing to the device directory
781 in the controller specific subtree.
783 The file drivers contains general information about the drivers used for the
784 IDE devices:
786 > cat /proc/ide/drivers
787 ide-cdrom version 4.53
788 ide-disk version 1.08
790 More detailed information can be found in the controller specific
791 subdirectories. These are named ide0, ide1 and so on. Each of these
792 directories contains the files shown in table 1-6.
795 Table 1-6: IDE controller info in /proc/ide/ide?
796 ..............................................................................
797 File Content
798 channel IDE channel (0 or 1)
799 config Configuration (only for PCI/IDE bridge)
800 mate Mate name
801 model Type/Chipset of IDE controller
802 ..............................................................................
804 Each device connected to a controller has a separate subdirectory in the
805 controllers directory. The files listed in table 1-7 are contained in these
806 directories.
809 Table 1-7: IDE device information
810 ..............................................................................
811 File Content
812 cache The cache
813 capacity Capacity of the medium (in 512Byte blocks)
814 driver driver and version
815 geometry physical and logical geometry
816 identify device identify block
817 media media type
818 model device identifier
819 settings device setup
820 smart_thresholds IDE disk management thresholds
821 smart_values IDE disk management values
822 ..............................................................................
824 The most interesting file is settings. This file contains a nice overview of
825 the drive parameters:
827 # cat /proc/ide/ide0/hda/settings
828 name value min max mode
829 ---- ----- --- --- ----
830 bios_cyl 526 0 65535 rw
831 bios_head 255 0 255 rw
832 bios_sect 63 0 63 rw
833 breada_readahead 4 0 127 rw
834 bswap 0 0 1 r
835 file_readahead 72 0 2097151 rw
836 io_32bit 0 0 3 rw
837 keepsettings 0 0 1 rw
838 max_kb_per_request 122 1 127 rw
839 multcount 0 0 8 rw
840 nice1 1 0 1 rw
841 nowerr 0 0 1 rw
842 pio_mode write-only 0 255 w
843 slow 0 0 1 rw
844 unmaskirq 0 0 1 rw
845 using_dma 0 0 1 rw
848 1.4 Networking info in /proc/net
849 --------------------------------
851 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
852 additional values you get for IP version 6 if you configure the kernel to
853 support this. Table 1-9 lists the files and their meaning.
856 Table 1-8: IPv6 info in /proc/net
857 ..............................................................................
858 File Content
859 udp6 UDP sockets (IPv6)
860 tcp6 TCP sockets (IPv6)
861 raw6 Raw device statistics (IPv6)
862 igmp6 IP multicast addresses, which this host joined (IPv6)
863 if_inet6 List of IPv6 interface addresses
864 ipv6_route Kernel routing table for IPv6
865 rt6_stats Global IPv6 routing tables statistics
866 sockstat6 Socket statistics (IPv6)
867 snmp6 Snmp data (IPv6)
868 ..............................................................................
871 Table 1-9: Network info in /proc/net
872 ..............................................................................
873 File Content
874 arp Kernel ARP table
875 dev network devices with statistics
876 dev_mcast the Layer2 multicast groups a device is listening too
877 (interface index, label, number of references, number of bound
878 addresses).
879 dev_stat network device status
880 ip_fwchains Firewall chain linkage
881 ip_fwnames Firewall chain names
882 ip_masq Directory containing the masquerading tables
883 ip_masquerade Major masquerading table
884 netstat Network statistics
885 raw raw device statistics
886 route Kernel routing table
887 rpc Directory containing rpc info
888 rt_cache Routing cache
889 snmp SNMP data
890 sockstat Socket statistics
891 tcp TCP sockets
892 tr_rif Token ring RIF routing table
893 udp UDP sockets
894 unix UNIX domain sockets
895 wireless Wireless interface data (Wavelan etc)
896 igmp IP multicast addresses, which this host joined
897 psched Global packet scheduler parameters.
898 netlink List of PF_NETLINK sockets
899 ip_mr_vifs List of multicast virtual interfaces
900 ip_mr_cache List of multicast routing cache
901 ..............................................................................
903 You can use this information to see which network devices are available in
904 your system and how much traffic was routed over those devices:
906 > cat /proc/net/dev
907 Inter-|Receive |[...
908 face |bytes packets errs drop fifo frame compressed multicast|[...
909 lo: 908188 5596 0 0 0 0 0 0 [...
910 ppp0:15475140 20721 410 0 0 410 0 0 [...
911 eth0: 614530 7085 0 0 0 0 0 1 [...
913 ...] Transmit
914 ...] bytes packets errs drop fifo colls carrier compressed
915 ...] 908188 5596 0 0 0 0 0 0
916 ...] 1375103 17405 0 0 0 0 0 0
917 ...] 1703981 5535 0 0 0 3 0 0
919 In addition, each Channel Bond interface has it's own directory. For
920 example, the bond0 device will have a directory called /proc/net/bond0/.
921 It will contain information that is specific to that bond, such as the
922 current slaves of the bond, the link status of the slaves, and how
923 many times the slaves link has failed.
925 1.5 SCSI info
926 -------------
928 If you have a SCSI host adapter in your system, you'll find a subdirectory
929 named after the driver for this adapter in /proc/scsi. You'll also see a list
930 of all recognized SCSI devices in /proc/scsi:
932 >cat /proc/scsi/scsi
933 Attached devices:
934 Host: scsi0 Channel: 00 Id: 00 Lun: 00
935 Vendor: IBM Model: DGHS09U Rev: 03E0
936 Type: Direct-Access ANSI SCSI revision: 03
937 Host: scsi0 Channel: 00 Id: 06 Lun: 00
938 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
939 Type: CD-ROM ANSI SCSI revision: 02
942 The directory named after the driver has one file for each adapter found in
943 the system. These files contain information about the controller, including
944 the used IRQ and the IO address range. The amount of information shown is
945 dependent on the adapter you use. The example shows the output for an Adaptec
946 AHA-2940 SCSI adapter:
948 > cat /proc/scsi/aic7xxx/0
950 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
951 Compile Options:
952 TCQ Enabled By Default : Disabled
953 AIC7XXX_PROC_STATS : Disabled
954 AIC7XXX_RESET_DELAY : 5
955 Adapter Configuration:
956 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
957 Ultra Wide Controller
958 PCI MMAPed I/O Base: 0xeb001000
959 Adapter SEEPROM Config: SEEPROM found and used.
960 Adaptec SCSI BIOS: Enabled
961 IRQ: 10
962 SCBs: Active 0, Max Active 2,
963 Allocated 15, HW 16, Page 255
964 Interrupts: 160328
965 BIOS Control Word: 0x18b6
966 Adapter Control Word: 0x005b
967 Extended Translation: Enabled
968 Disconnect Enable Flags: 0xffff
969 Ultra Enable Flags: 0x0001
970 Tag Queue Enable Flags: 0x0000
971 Ordered Queue Tag Flags: 0x0000
972 Default Tag Queue Depth: 8
973 Tagged Queue By Device array for aic7xxx host instance 0:
974 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
975 Actual queue depth per device for aic7xxx host instance 0:
976 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
977 Statistics:
978 (scsi0:0:0:0)
979 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
980 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
981 Total transfers 160151 (74577 reads and 85574 writes)
982 (scsi0:0:6:0)
983 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
984 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
985 Total transfers 0 (0 reads and 0 writes)
988 1.6 Parallel port info in /proc/parport
989 ---------------------------------------
991 The directory /proc/parport contains information about the parallel ports of
992 your system. It has one subdirectory for each port, named after the port
993 number (0,1,2,...).
995 These directories contain the four files shown in Table 1-10.
998 Table 1-10: Files in /proc/parport
999 ..............................................................................
1000 File Content
1001 autoprobe Any IEEE-1284 device ID information that has been acquired.
1002 devices list of the device drivers using that port. A + will appear by the
1003 name of the device currently using the port (it might not appear
1004 against any).
1005 hardware Parallel port's base address, IRQ line and DMA channel.
1006 irq IRQ that parport is using for that port. This is in a separate
1007 file to allow you to alter it by writing a new value in (IRQ
1008 number or none).
1009 ..............................................................................
1011 1.7 TTY info in /proc/tty
1012 -------------------------
1014 Information about the available and actually used tty's can be found in the
1015 directory /proc/tty.You'll find entries for drivers and line disciplines in
1016 this directory, as shown in Table 1-11.
1019 Table 1-11: Files in /proc/tty
1020 ..............................................................................
1021 File Content
1022 drivers list of drivers and their usage
1023 ldiscs registered line disciplines
1024 driver/serial usage statistic and status of single tty lines
1025 ..............................................................................
1027 To see which tty's are currently in use, you can simply look into the file
1028 /proc/tty/drivers:
1030 > cat /proc/tty/drivers
1031 pty_slave /dev/pts 136 0-255 pty:slave
1032 pty_master /dev/ptm 128 0-255 pty:master
1033 pty_slave /dev/ttyp 3 0-255 pty:slave
1034 pty_master /dev/pty 2 0-255 pty:master
1035 serial /dev/cua 5 64-67 serial:callout
1036 serial /dev/ttyS 4 64-67 serial
1037 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1038 /dev/ptmx /dev/ptmx 5 2 system
1039 /dev/console /dev/console 5 1 system:console
1040 /dev/tty /dev/tty 5 0 system:/dev/tty
1041 unknown /dev/tty 4 1-63 console
1044 1.8 Miscellaneous kernel statistics in /proc/stat
1045 -------------------------------------------------
1047 Various pieces of information about kernel activity are available in the
1048 /proc/stat file. All of the numbers reported in this file are aggregates
1049 since the system first booted. For a quick look, simply cat the file:
1051 > cat /proc/stat
1052 cpu 2255 34 2290 22625563 6290 127 456 0 0
1053 cpu0 1132 34 1441 11311718 3675 127 438 0 0
1054 cpu1 1123 0 849 11313845 2614 0 18 0 0
1055 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1056 ctxt 1990473
1057 btime 1062191376
1058 processes 2915
1059 procs_running 1
1060 procs_blocked 0
1061 softirq 183433 0 21755 12 39 1137 231 21459 2263
1063 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1064 lines. These numbers identify the amount of time the CPU has spent performing
1065 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1066 second). The meanings of the columns are as follows, from left to right:
1068 - user: normal processes executing in user mode
1069 - nice: niced processes executing in user mode
1070 - system: processes executing in kernel mode
1071 - idle: twiddling thumbs
1072 - iowait: waiting for I/O to complete
1073 - irq: servicing interrupts
1074 - softirq: servicing softirqs
1075 - steal: involuntary wait
1076 - guest: running a normal guest
1077 - guest_nice: running a niced guest
1079 The "intr" line gives counts of interrupts serviced since boot time, for each
1080 of the possible system interrupts. The first column is the total of all
1081 interrupts serviced; each subsequent column is the total for that particular
1082 interrupt.
1084 The "ctxt" line gives the total number of context switches across all CPUs.
1086 The "btime" line gives the time at which the system booted, in seconds since
1087 the Unix epoch.
1089 The "processes" line gives the number of processes and threads created, which
1090 includes (but is not limited to) those created by calls to the fork() and
1091 clone() system calls.
1093 The "procs_running" line gives the total number of threads that are
1094 running or ready to run (i.e., the total number of runnable threads).
1096 The "procs_blocked" line gives the number of processes currently blocked,
1097 waiting for I/O to complete.
1099 The "softirq" line gives counts of softirqs serviced since boot time, for each
1100 of the possible system softirqs. The first column is the total of all
1101 softirqs serviced; each subsequent column is the total for that particular
1102 softirq.
1105 1.9 Ext4 file system parameters
1106 ------------------------------
1108 Information about mounted ext4 file systems can be found in
1109 /proc/fs/ext4. Each mounted filesystem will have a directory in
1110 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1111 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1112 in Table 1-12, below.
1114 Table 1-12: Files in /proc/fs/ext4/<devname>
1115 ..............................................................................
1116 File Content
1117 mb_groups details of multiblock allocator buddy cache of free blocks
1118 ..............................................................................
1121 ------------------------------------------------------------------------------
1122 Summary
1123 ------------------------------------------------------------------------------
1124 The /proc file system serves information about the running system. It not only
1125 allows access to process data but also allows you to request the kernel status
1126 by reading files in the hierarchy.
1128 The directory structure of /proc reflects the types of information and makes
1129 it easy, if not obvious, where to look for specific data.
1130 ------------------------------------------------------------------------------
1132 ------------------------------------------------------------------------------
1133 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1134 ------------------------------------------------------------------------------
1136 ------------------------------------------------------------------------------
1137 In This Chapter
1138 ------------------------------------------------------------------------------
1139 * Modifying kernel parameters by writing into files found in /proc/sys
1140 * Exploring the files which modify certain parameters
1141 * Review of the /proc/sys file tree
1142 ------------------------------------------------------------------------------
1145 A very interesting part of /proc is the directory /proc/sys. This is not only
1146 a source of information, it also allows you to change parameters within the
1147 kernel. Be very careful when attempting this. You can optimize your system,
1148 but you can also cause it to crash. Never alter kernel parameters on a
1149 production system. Set up a development machine and test to make sure that
1150 everything works the way you want it to. You may have no alternative but to
1151 reboot the machine once an error has been made.
1153 To change a value, simply echo the new value into the file. An example is
1154 given below in the section on the file system data. You need to be root to do
1155 this. You can create your own boot script to perform this every time your
1156 system boots.
1158 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1159 general things in the operation of the Linux kernel. Since some of the files
1160 can inadvertently disrupt your system, it is advisable to read both
1161 documentation and source before actually making adjustments. In any case, be
1162 very careful when writing to any of these files. The entries in /proc may
1163 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1164 review the kernel documentation in the directory /usr/src/linux/Documentation.
1165 This chapter is heavily based on the documentation included in the pre 2.2
1166 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1168 Please see: Documentation/sysctls/ directory for descriptions of these
1169 entries.
1171 ------------------------------------------------------------------------------
1172 Summary
1173 ------------------------------------------------------------------------------
1174 Certain aspects of kernel behavior can be modified at runtime, without the
1175 need to recompile the kernel, or even to reboot the system. The files in the
1176 /proc/sys tree can not only be read, but also modified. You can use the echo
1177 command to write value into these files, thereby changing the default settings
1178 of the kernel.
1179 ------------------------------------------------------------------------------
1181 ------------------------------------------------------------------------------
1182 CHAPTER 3: PER-PROCESS PARAMETERS
1183 ------------------------------------------------------------------------------
1185 3.1 /proc/<pid>/oom_adj - Adjust the oom-killer score
1186 ------------------------------------------------------
1188 This file can be used to adjust the score used to select which processes
1189 should be killed in an out-of-memory situation. Giving it a high score will
1190 increase the likelihood of this process being killed by the oom-killer. Valid
1191 values are in the range -16 to +15, plus the special value -17, which disables
1192 oom-killing altogether for this process.
1194 The process to be killed in an out-of-memory situation is selected among all others
1195 based on its badness score. This value equals the original memory size of the process
1196 and is then updated according to its CPU time (utime + stime) and the
1197 run time (uptime - start time). The longer it runs the smaller is the score.
1198 Badness score is divided by the square root of the CPU time and then by
1199 the double square root of the run time.
1201 Swapped out tasks are killed first. Half of each child's memory size is added to
1202 the parent's score if they do not share the same memory. Thus forking servers
1203 are the prime candidates to be killed. Having only one 'hungry' child will make
1204 parent less preferable than the child.
1206 /proc/<pid>/oom_score shows process' current badness score.
1208 The following heuristics are then applied:
1209 * if the task was reniced, its score doubles
1210 * superuser or direct hardware access tasks (CAP_SYS_ADMIN, CAP_SYS_RESOURCE
1211 or CAP_SYS_RAWIO) have their score divided by 4
1212 * if oom condition happened in one cpuset and checked process does not belong
1213 to it, its score is divided by 8
1214 * the resulting score is multiplied by two to the power of oom_adj, i.e.
1215 points <<= oom_adj when it is positive and
1216 points >>= -(oom_adj) otherwise
1218 The task with the highest badness score is then selected and its children
1219 are killed, process itself will be killed in an OOM situation when it does
1220 not have children or some of them disabled oom like described above.
1222 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1223 -------------------------------------------------------------
1225 This file can be used to check the current score used by the oom-killer is for
1226 any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
1227 process should be killed in an out-of-memory situation.
1230 3.3 /proc/<pid>/io - Display the IO accounting fields
1231 -------------------------------------------------------
1233 This file contains IO statistics for each running process
1235 Example
1236 -------
1238 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1239 [1] 3828
1241 test:/tmp # cat /proc/3828/io
1242 rchar: 323934931
1243 wchar: 323929600
1244 syscr: 632687
1245 syscw: 632675
1246 read_bytes: 0
1247 write_bytes: 323932160
1248 cancelled_write_bytes: 0
1251 Description
1252 -----------
1254 rchar
1255 -----
1257 I/O counter: chars read
1258 The number of bytes which this task has caused to be read from storage. This
1259 is simply the sum of bytes which this process passed to read() and pread().
1260 It includes things like tty IO and it is unaffected by whether or not actual
1261 physical disk IO was required (the read might have been satisfied from
1262 pagecache)
1265 wchar
1266 -----
1268 I/O counter: chars written
1269 The number of bytes which this task has caused, or shall cause to be written
1270 to disk. Similar caveats apply here as with rchar.
1273 syscr
1274 -----
1276 I/O counter: read syscalls
1277 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1278 and pread().
1281 syscw
1282 -----
1284 I/O counter: write syscalls
1285 Attempt to count the number of write I/O operations, i.e. syscalls like
1286 write() and pwrite().
1289 read_bytes
1290 ----------
1292 I/O counter: bytes read
1293 Attempt to count the number of bytes which this process really did cause to
1294 be fetched from the storage layer. Done at the submit_bio() level, so it is
1295 accurate for block-backed filesystems. <please add status regarding NFS and
1296 CIFS at a later time>
1299 write_bytes
1300 -----------
1302 I/O counter: bytes written
1303 Attempt to count the number of bytes which this process caused to be sent to
1304 the storage layer. This is done at page-dirtying time.
1307 cancelled_write_bytes
1308 ---------------------
1310 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1311 then deletes the file, it will in fact perform no writeout. But it will have
1312 been accounted as having caused 1MB of write.
1313 In other words: The number of bytes which this process caused to not happen,
1314 by truncating pagecache. A task can cause "negative" IO too. If this task
1315 truncates some dirty pagecache, some IO which another task has been accounted
1316 for (in it's write_bytes) will not be happening. We _could_ just subtract that
1317 from the truncating task's write_bytes, but there is information loss in doing
1318 that.
1321 Note
1322 ----
1324 At its current implementation state, this is a bit racy on 32-bit machines: if
1325 process A reads process B's /proc/pid/io while process B is updating one of
1326 those 64-bit counters, process A could see an intermediate result.
1329 More information about this can be found within the taskstats documentation in
1330 Documentation/accounting.
1332 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1333 ---------------------------------------------------------------
1334 When a process is dumped, all anonymous memory is written to a core file as
1335 long as the size of the core file isn't limited. But sometimes we don't want
1336 to dump some memory segments, for example, huge shared memory. Conversely,
1337 sometimes we want to save file-backed memory segments into a core file, not
1338 only the individual files.
1340 /proc/<pid>/coredump_filter allows you to customize which memory segments
1341 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1342 of memory types. If a bit of the bitmask is set, memory segments of the
1343 corresponding memory type are dumped, otherwise they are not dumped.
1345 The following 7 memory types are supported:
1346 - (bit 0) anonymous private memory
1347 - (bit 1) anonymous shared memory
1348 - (bit 2) file-backed private memory
1349 - (bit 3) file-backed shared memory
1350 - (bit 4) ELF header pages in file-backed private memory areas (it is
1351 effective only if the bit 2 is cleared)
1352 - (bit 5) hugetlb private memory
1353 - (bit 6) hugetlb shared memory
1355 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1356 are always dumped regardless of the bitmask status.
1358 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1359 effected by bit 5-6.
1361 Default value of coredump_filter is 0x23; this means all anonymous memory
1362 segments and hugetlb private memory are dumped.
1364 If you don't want to dump all shared memory segments attached to pid 1234,
1365 write 0x21 to the process's proc file.
1367 $ echo 0x21 > /proc/1234/coredump_filter
1369 When a new process is created, the process inherits the bitmask status from its
1370 parent. It is useful to set up coredump_filter before the program runs.
1371 For example:
1373 $ echo 0x7 > /proc/self/coredump_filter
1374 $ ./some_program
1376 3.5 /proc/<pid>/mountinfo - Information about mounts
1377 --------------------------------------------------------
1379 This file contains lines of the form:
1381 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1382 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1384 (1) mount ID: unique identifier of the mount (may be reused after umount)
1385 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1386 (3) major:minor: value of st_dev for files on filesystem
1387 (4) root: root of the mount within the filesystem
1388 (5) mount point: mount point relative to the process's root
1389 (6) mount options: per mount options
1390 (7) optional fields: zero or more fields of the form "tag[:value]"
1391 (8) separator: marks the end of the optional fields
1392 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1393 (10) mount source: filesystem specific information or "none"
1394 (11) super options: per super block options
1396 Parsers should ignore all unrecognised optional fields. Currently the
1397 possible optional fields are:
1399 shared:X mount is shared in peer group X
1400 master:X mount is slave to peer group X
1401 propagate_from:X mount is slave and receives propagation from peer group X (*)
1402 unbindable mount is unbindable
1404 (*) X is the closest dominant peer group under the process's root. If
1405 X is the immediate master of the mount, or if there's no dominant peer
1406 group under the same root, then only the "master:X" field is present
1407 and not the "propagate_from:X" field.
1409 For more information on mount propagation see:
1411 Documentation/filesystems/sharedsubtree.txt
1414 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1415 --------------------------------------------------------
1416 These files provide a method to access a tasks comm value. It also allows for
1417 a task to set its own or one of its thread siblings comm value. The comm value
1418 is limited in size compared to the cmdline value, so writing anything longer
1419 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1420 comm value.