| blob | ffead13f9443fd7d11b63e2cd3a6ffb15c0f1994 |
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
43 ------------------------------------------------------------------------------
44 Preface
45 ------------------------------------------------------------------------------
47 0.1 Introduction/Credits
48 ------------------------
50 This documentation is part of a soon (or so we hope) to be released book on
51 the SuSE Linux distribution. As there is no complete documentation for the
52 /proc file system and we've used many freely available sources to write these
53 chapters, it seems only fair to give the work back to the Linux community.
54 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
55 afraid it's still far from complete, but we hope it will be useful. As far as
56 we know, it is the first 'all-in-one' document about the /proc file system. It
57 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
58 SPARC, AXP, etc., features, you probably won't find what you are looking for.
59 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
60 additions and patches are welcome and will be added to this document if you
61 mail them to Bodo.
63 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
64 other people for help compiling this documentation. We'd also like to extend a
65 special thank you to Andi Kleen for documentation, which we relied on heavily
66 to create this document, as well as the additional information he provided.
67 Thanks to everybody else who contributed source or docs to the Linux kernel
68 and helped create a great piece of software... :)
70 If you have any comments, corrections or additions, please don't hesitate to
71 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
72 document.
74 The latest version of this document is available online at
75 http://skaro.nightcrawler.com/~bb/Docs/Proc as HTML version.
77 If the above direction does not works for you, ypu could try the kernel
78 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
79 comandante@zaralinux.com.
81 0.2 Legal Stuff
82 ---------------
84 We don't guarantee the correctness of this document, and if you come to us
85 complaining about how you screwed up your system because of incorrect
86 documentation, we won't feel responsible...
88 ------------------------------------------------------------------------------
89 CHAPTER 1: COLLECTING SYSTEM INFORMATION
90 ------------------------------------------------------------------------------
92 ------------------------------------------------------------------------------
93 In This Chapter
94 ------------------------------------------------------------------------------
95 * Investigating the properties of the pseudo file system /proc and its
96 ability to provide information on the running Linux system
97 * Examining /proc's structure
98 * Uncovering various information about the kernel and the processes running
99 on the system
100 ------------------------------------------------------------------------------
103 The proc file system acts as an interface to internal data structures in the
104 kernel. It can be used to obtain information about the system and to change
105 certain kernel parameters at runtime (sysctl).
107 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
108 show you how you can use /proc/sys to change settings.
110 1.1 Process-Specific Subdirectories
111 -----------------------------------
113 The directory /proc contains (among other things) one subdirectory for each
114 process running on the system, which is named after the process ID (PID).
116 The link self points to the process reading the file system. Each process
117 subdirectory has the entries listed in Table 1-1.
120 Table 1-1: Process specific entries in /proc
121 ..............................................................................
122 File Content
123 clear_refs Clears page referenced bits shown in smaps output
124 cmdline Command line arguments
125 cpu Current and last cpu in which it was executed (2.4)(smp)
126 cwd Link to the current working directory
127 environ Values of environment variables
128 exe Link to the executable of this process
129 fd Directory, which contains all file descriptors
130 maps Memory maps to executables and library files (2.4)
131 mem Memory held by this process
132 root Link to the root directory of this process
133 stat Process status
134 statm Process memory status information
135 status Process status in human readable form
136 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
137 stack Report full stack trace, enable via CONFIG_STACKTRACE
138 smaps a extension based on maps, showing the memory consumption of
139 each mapping
140 ..............................................................................
142 For example, to get the status information of a process, all you have to do is
143 read the file /proc/PID/status:
145 >cat /proc/self/status
146 Name: cat
147 State: R (running)
148 Tgid: 5452
149 Pid: 5452
150 PPid: 743
151 TracerPid: 0 (2.4)
152 Uid: 501 501 501 501
153 Gid: 100 100 100 100
154 FDSize: 256
155 Groups: 100 14 16
156 VmPeak: 5004 kB
157 VmSize: 5004 kB
158 VmLck: 0 kB
159 VmHWM: 476 kB
160 VmRSS: 476 kB
161 VmData: 156 kB
162 VmStk: 88 kB
163 VmExe: 68 kB
164 VmLib: 1412 kB
165 VmPTE: 20 kb
166 Threads: 1
167 SigQ: 0/28578
168 SigPnd: 0000000000000000
169 ShdPnd: 0000000000000000
170 SigBlk: 0000000000000000
171 SigIgn: 0000000000000000
172 SigCgt: 0000000000000000
173 CapInh: 00000000fffffeff
174 CapPrm: 0000000000000000
175 CapEff: 0000000000000000
176 CapBnd: ffffffffffffffff
177 voluntary_ctxt_switches: 0
178 nonvoluntary_ctxt_switches: 1
180 This shows you nearly the same information you would get if you viewed it with
181 the ps command. In fact, ps uses the proc file system to obtain its
182 information. But you get a more detailed view of the process by reading the
183 file /proc/PID/status. It fields are described in table 1-2.
185 The statm file contains more detailed information about the process
186 memory usage. Its seven fields are explained in Table 1-3. The stat file
187 contains details information about the process itself. Its fields are
188 explained in Table 1-4.
190 Table 1-2: Contents of the statm files (as of 2.6.30-rc7)
191 ..............................................................................
192 Field Content
193 Name filename of the executable
194 State state (R is running, S is sleeping, D is sleeping
195 in an uninterruptible wait, Z is zombie,
196 T is traced or stopped)
197 Tgid thread group ID
198 Pid process id
199 PPid process id of the parent process
200 TracerPid PID of process tracing this process (0 if not)
201 Uid Real, effective, saved set, and file system UIDs
202 Gid Real, effective, saved set, and file system GIDs
203 FDSize number of file descriptor slots currently allocated
204 Groups supplementary group list
205 VmPeak peak virtual memory size
206 VmSize total program size
207 VmLck locked memory size
208 VmHWM peak resident set size ("high water mark")
209 VmRSS size of memory portions
210 VmData size of data, stack, and text segments
211 VmStk size of data, stack, and text segments
212 VmExe size of text segment
213 VmLib size of shared library code
214 VmPTE size of page table entries
215 Threads number of threads
216 SigQ number of signals queued/max. number for queue
217 SigPnd bitmap of pending signals for the thread
218 ShdPnd bitmap of shared pending signals for the process
219 SigBlk bitmap of blocked signals
220 SigIgn bitmap of ignored signals
221 SigCgt bitmap of catched signals
222 CapInh bitmap of inheritable capabilities
223 CapPrm bitmap of permitted capabilities
224 CapEff bitmap of effective capabilities
225 CapBnd bitmap of capabilities bounding set
226 Cpus_allowed mask of CPUs on which this process may run
227 Cpus_allowed_list Same as previous, but in "list format"
228 Mems_allowed mask of memory nodes allowed to this process
229 Mems_allowed_list Same as previous, but in "list format"
230 voluntary_ctxt_switches number of voluntary context switches
231 nonvoluntary_ctxt_switches number of non voluntary context switches
232 ..............................................................................
234 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
235 ..............................................................................
236 Field Content
237 size total program size (pages) (same as VmSize in status)
238 resident size of memory portions (pages) (same as VmRSS in status)
239 shared number of pages that are shared (i.e. backed by a file)
240 trs number of pages that are 'code' (not including libs; broken,
241 includes data segment)
242 lrs number of pages of library (always 0 on 2.6)
243 drs number of pages of data/stack (including libs; broken,
244 includes library text)
245 dt number of dirty pages (always 0 on 2.6)
246 ..............................................................................
249 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
250 ..............................................................................
251 Field Content
252 pid process id
253 tcomm filename of the executable
254 state state (R is running, S is sleeping, D is sleeping in an
255 uninterruptible wait, Z is zombie, T is traced or stopped)
256 ppid process id of the parent process
257 pgrp pgrp of the process
258 sid session id
259 tty_nr tty the process uses
260 tty_pgrp pgrp of the tty
261 flags task flags
262 min_flt number of minor faults
263 cmin_flt number of minor faults with child's
264 maj_flt number of major faults
265 cmaj_flt number of major faults with child's
266 utime user mode jiffies
267 stime kernel mode jiffies
268 cutime user mode jiffies with child's
269 cstime kernel mode jiffies with child's
270 priority priority level
271 nice nice level
272 num_threads number of threads
273 it_real_value (obsolete, always 0)
274 start_time time the process started after system boot
275 vsize virtual memory size
276 rss resident set memory size
277 rsslim current limit in bytes on the rss
278 start_code address above which program text can run
279 end_code address below which program text can run
280 start_stack address of the start of the stack
281 esp current value of ESP
282 eip current value of EIP
283 pending bitmap of pending signals
284 blocked bitmap of blocked signals
285 sigign bitmap of ignored signals
286 sigcatch bitmap of catched signals
287 wchan address where process went to sleep
288 0 (place holder)
289 0 (place holder)
290 exit_signal signal to send to parent thread on exit
291 task_cpu which CPU the task is scheduled on
292 rt_priority realtime priority
293 policy scheduling policy (man sched_setscheduler)
294 blkio_ticks time spent waiting for block IO
295 gtime guest time of the task in jiffies
296 cgtime guest time of the task children in jiffies
297 ..............................................................................
299 The /proc/PID/map file containing the currently mapped memory regions and
300 their access permissions.
302 The format is:
304 address perms offset dev inode pathname
306 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
307 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
308 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
309 a7cb1000-a7cb2000 ---p 00000000 00:00 0
310 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
311 a7eb2000-a7eb3000 ---p 00000000 00:00 0
312 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
313 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
314 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
315 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
316 a800b000-a800e000 rw-p 00000000 00:00 0
317 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
318 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
319 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
320 a8024000-a8027000 rw-p 00000000 00:00 0
321 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
322 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
323 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
324 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
325 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
327 where "address" is the address space in the process that it occupies, "perms"
328 is a set of permissions:
330 r = read
331 w = write
332 x = execute
333 s = shared
334 p = private (copy on write)
336 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
337 "inode" is the inode on that device. 0 indicates that no inode is associated
338 with the memory region, as the case would be with BSS (uninitialized data).
339 The "pathname" shows the name associated file for this mapping. If the mapping
340 is not associated with a file:
342 [heap] = the heap of the program
343 [stack] = the stack of the main process
344 [vdso] = the "virtual dynamic shared object",
345 the kernel system call handler
347 or if empty, the mapping is anonymous.
350 The /proc/PID/smaps is an extension based on maps, showing the memory
351 consumption for each of the process's mappings. For each of mappings there
352 is a series of lines such as the following:
354 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
355 Size: 1084 kB
356 Rss: 892 kB
357 Pss: 374 kB
358 Shared_Clean: 892 kB
359 Shared_Dirty: 0 kB
360 Private_Clean: 0 kB
361 Private_Dirty: 0 kB
362 Referenced: 892 kB
363 Swap: 0 kB
364 KernelPageSize: 4 kB
365 MMUPageSize: 4 kB
367 The first of these lines shows the same information as is displayed for the
368 mapping in /proc/PID/maps. The remaining lines show the size of the mapping,
369 the amount of the mapping that is currently resident in RAM, the "proportional
370 set size” (divide each shared page by the number of processes sharing it), the
371 number of clean and dirty shared pages in the mapping, and the number of clean
372 and dirty private pages in the mapping. The "Referenced" indicates the amount
373 of memory currently marked as referenced or accessed.
375 This file is only present if the CONFIG_MMU kernel configuration option is
376 enabled.
378 1.2 Kernel data
379 ---------------
381 Similar to the process entries, the kernel data files give information about
382 the running kernel. The files used to obtain this information are contained in
383 /proc and are listed in Table 1-5. Not all of these will be present in your
384 system. It depends on the kernel configuration and the loaded modules, which
385 files are there, and which are missing.
387 Table 1-5: Kernel info in /proc
388 ..............................................................................
389 File Content
390 apm Advanced power management info
391 buddyinfo Kernel memory allocator information (see text) (2.5)
392 bus Directory containing bus specific information
393 cmdline Kernel command line
394 cpuinfo Info about the CPU
395 devices Available devices (block and character)
396 dma Used DMS channels
397 filesystems Supported filesystems
398 driver Various drivers grouped here, currently rtc (2.4)
399 execdomains Execdomains, related to security (2.4)
400 fb Frame Buffer devices (2.4)
401 fs File system parameters, currently nfs/exports (2.4)
402 ide Directory containing info about the IDE subsystem
403 interrupts Interrupt usage
404 iomem Memory map (2.4)
405 ioports I/O port usage
406 irq Masks for irq to cpu affinity (2.4)(smp?)
407 isapnp ISA PnP (Plug&Play) Info (2.4)
408 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
409 kmsg Kernel messages
410 ksyms Kernel symbol table
411 loadavg Load average of last 1, 5 & 15 minutes
412 locks Kernel locks
413 meminfo Memory info
414 misc Miscellaneous
415 modules List of loaded modules
416 mounts Mounted filesystems
417 net Networking info (see text)
418 partitions Table of partitions known to the system
419 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
420 decoupled by lspci (2.4)
421 rtc Real time clock
422 scsi SCSI info (see text)
423 slabinfo Slab pool info
424 softirqs softirq usage
425 stat Overall statistics
426 swaps Swap space utilization
427 sys See chapter 2
428 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
429 tty Info of tty drivers
430 uptime System uptime
431 version Kernel version
432 video bttv info of video resources (2.4)
433 vmallocinfo Show vmalloced areas
434 ..............................................................................
436 You can, for example, check which interrupts are currently in use and what
437 they are used for by looking in the file /proc/interrupts:
439 > cat /proc/interrupts
440 CPU0
441 0: 8728810 XT-PIC timer
442 1: 895 XT-PIC keyboard
443 2: 0 XT-PIC cascade
444 3: 531695 XT-PIC aha152x
445 4: 2014133 XT-PIC serial
446 5: 44401 XT-PIC pcnet_cs
447 8: 2 XT-PIC rtc
448 11: 8 XT-PIC i82365
449 12: 182918 XT-PIC PS/2 Mouse
450 13: 1 XT-PIC fpu
451 14: 1232265 XT-PIC ide0
452 15: 7 XT-PIC ide1
453 NMI: 0
455 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
456 output of a SMP machine):
458 > cat /proc/interrupts
460 CPU0 CPU1
461 0: 1243498 1214548 IO-APIC-edge timer
462 1: 8949 8958 IO-APIC-edge keyboard
463 2: 0 0 XT-PIC cascade
464 5: 11286 10161 IO-APIC-edge soundblaster
465 8: 1 0 IO-APIC-edge rtc
466 9: 27422 27407 IO-APIC-edge 3c503
467 12: 113645 113873 IO-APIC-edge PS/2 Mouse
468 13: 0 0 XT-PIC fpu
469 14: 22491 24012 IO-APIC-edge ide0
470 15: 2183 2415 IO-APIC-edge ide1
471 17: 30564 30414 IO-APIC-level eth0
472 18: 177 164 IO-APIC-level bttv
473 NMI: 2457961 2457959
474 LOC: 2457882 2457881
475 ERR: 2155
477 NMI is incremented in this case because every timer interrupt generates a NMI
478 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
480 LOC is the local interrupt counter of the internal APIC of every CPU.
482 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
483 connects the CPUs in a SMP system. This means that an error has been detected,
484 the IO-APIC automatically retry the transmission, so it should not be a big
485 problem, but you should read the SMP-FAQ.
487 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
488 /proc/interrupts to display every IRQ vector in use by the system, not
489 just those considered 'most important'. The new vectors are:
491 THR -- interrupt raised when a machine check threshold counter
492 (typically counting ECC corrected errors of memory or cache) exceeds
493 a configurable threshold. Only available on some systems.
495 TRM -- a thermal event interrupt occurs when a temperature threshold
496 has been exceeded for the CPU. This interrupt may also be generated
497 when the temperature drops back to normal.
499 SPU -- a spurious interrupt is some interrupt that was raised then lowered
500 by some IO device before it could be fully processed by the APIC. Hence
501 the APIC sees the interrupt but does not know what device it came from.
502 For this case the APIC will generate the interrupt with a IRQ vector
503 of 0xff. This might also be generated by chipset bugs.
505 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
506 sent from one CPU to another per the needs of the OS. Typically,
507 their statistics are used by kernel developers and interested users to
508 determine the occurrence of interrupts of the given type.
510 The above IRQ vectors are displayed only when relevent. For example,
511 the threshold vector does not exist on x86_64 platforms. Others are
512 suppressed when the system is a uniprocessor. As of this writing, only
513 i386 and x86_64 platforms support the new IRQ vector displays.
515 Of some interest is the introduction of the /proc/irq directory to 2.4.
516 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
517 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
518 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
519 prof_cpu_mask.
521 For example
522 > ls /proc/irq/
523 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
524 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
525 > ls /proc/irq/0/
526 smp_affinity
528 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
529 IRQ, you can set it by doing:
531 > echo 1 > /proc/irq/10/smp_affinity
533 This means that only the first CPU will handle the IRQ, but you can also echo
534 5 which means that only the first and fourth CPU can handle the IRQ.
536 The contents of each smp_affinity file is the same by default:
538 > cat /proc/irq/0/smp_affinity
539 ffffffff
541 The default_smp_affinity mask applies to all non-active IRQs, which are the
542 IRQs which have not yet been allocated/activated, and hence which lack a
543 /proc/irq/[0-9]* directory.
545 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
546 profiler. Default value is ffffffff (all cpus).
548 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
549 between all the CPUs which are allowed to handle it. As usual the kernel has
550 more info than you and does a better job than you, so the defaults are the
551 best choice for almost everyone.
553 There are three more important subdirectories in /proc: net, scsi, and sys.
554 The general rule is that the contents, or even the existence of these
555 directories, depend on your kernel configuration. If SCSI is not enabled, the
556 directory scsi may not exist. The same is true with the net, which is there
557 only when networking support is present in the running kernel.
559 The slabinfo file gives information about memory usage at the slab level.
560 Linux uses slab pools for memory management above page level in version 2.2.
561 Commonly used objects have their own slab pool (such as network buffers,
562 directory cache, and so on).
564 ..............................................................................
566 > cat /proc/buddyinfo
568 Node 0, zone DMA 0 4 5 4 4 3 ...
569 Node 0, zone Normal 1 0 0 1 101 8 ...
570 Node 0, zone HighMem 2 0 0 1 1 0 ...
572 Memory fragmentation is a problem under some workloads, and buddyinfo is a
573 useful tool for helping diagnose these problems. Buddyinfo will give you a
574 clue as to how big an area you can safely allocate, or why a previous
575 allocation failed.
577 Each column represents the number of pages of a certain order which are
578 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
579 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
580 available in ZONE_NORMAL, etc...
582 ..............................................................................
584 meminfo:
586 Provides information about distribution and utilization of memory. This
587 varies by architecture and compile options. The following is from a
588 16GB PIII, which has highmem enabled. You may not have all of these fields.
590 > cat /proc/meminfo
593 MemTotal: 16344972 kB
594 MemFree: 13634064 kB
595 Buffers: 3656 kB
596 Cached: 1195708 kB
597 SwapCached: 0 kB
598 Active: 891636 kB
599 Inactive: 1077224 kB
600 HighTotal: 15597528 kB
601 HighFree: 13629632 kB
602 LowTotal: 747444 kB
603 LowFree: 4432 kB
604 SwapTotal: 0 kB
605 SwapFree: 0 kB
606 Dirty: 968 kB
607 Writeback: 0 kB
608 AnonPages: 861800 kB
609 Mapped: 280372 kB
610 Slab: 284364 kB
611 SReclaimable: 159856 kB
612 SUnreclaim: 124508 kB
613 PageTables: 24448 kB
614 NFS_Unstable: 0 kB
615 Bounce: 0 kB
616 WritebackTmp: 0 kB
617 CommitLimit: 7669796 kB
618 Committed_AS: 100056 kB
619 VmallocTotal: 112216 kB
620 VmallocUsed: 428 kB
621 VmallocChunk: 111088 kB
623 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
624 bits and the kernel binary code)
625 MemFree: The sum of LowFree+HighFree
626 Buffers: Relatively temporary storage for raw disk blocks
627 shouldn't get tremendously large (20MB or so)
628 Cached: in-memory cache for files read from the disk (the
629 pagecache). Doesn't include SwapCached
630 SwapCached: Memory that once was swapped out, is swapped back in but
631 still also is in the swapfile (if memory is needed it
632 doesn't need to be swapped out AGAIN because it is already
633 in the swapfile. This saves I/O)
634 Active: Memory that has been used more recently and usually not
635 reclaimed unless absolutely necessary.
636 Inactive: Memory which has been less recently used. It is more
637 eligible to be reclaimed for other purposes
638 HighTotal:
639 HighFree: Highmem is all memory above ~860MB of physical memory
640 Highmem areas are for use by userspace programs, or
641 for the pagecache. The kernel must use tricks to access
642 this memory, making it slower to access than lowmem.
643 LowTotal:
644 LowFree: Lowmem is memory which can be used for everything that
645 highmem can be used for, but it is also available for the
646 kernel's use for its own data structures. Among many
647 other things, it is where everything from the Slab is
648 allocated. Bad things happen when you're out of lowmem.
649 SwapTotal: total amount of swap space available
650 SwapFree: Memory which has been evicted from RAM, and is temporarily
651 on the disk
652 Dirty: Memory which is waiting to get written back to the disk
653 Writeback: Memory which is actively being written back to the disk
654 AnonPages: Non-file backed pages mapped into userspace page tables
655 Mapped: files which have been mmaped, such as libraries
656 Slab: in-kernel data structures cache
657 SReclaimable: Part of Slab, that might be reclaimed, such as caches
658 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
659 PageTables: amount of memory dedicated to the lowest level of page
660 tables.
661 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
662 storage
663 Bounce: Memory used for block device "bounce buffers"
664 WritebackTmp: Memory used by FUSE for temporary writeback buffers
665 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
666 this is the total amount of memory currently available to
667 be allocated on the system. This limit is only adhered to
668 if strict overcommit accounting is enabled (mode 2 in
669 'vm.overcommit_memory').
670 The CommitLimit is calculated with the following formula:
671 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
672 For example, on a system with 1G of physical RAM and 7G
673 of swap with a `vm.overcommit_ratio` of 30 it would
674 yield a CommitLimit of 7.3G.
675 For more details, see the memory overcommit documentation
676 in vm/overcommit-accounting.
677 Committed_AS: The amount of memory presently allocated on the system.
678 The committed memory is a sum of all of the memory which
679 has been allocated by processes, even if it has not been
680 "used" by them as of yet. A process which malloc()'s 1G
681 of memory, but only touches 300M of it will only show up
682 as using 300M of memory even if it has the address space
683 allocated for the entire 1G. This 1G is memory which has
684 been "committed" to by the VM and can be used at any time
685 by the allocating application. With strict overcommit
686 enabled on the system (mode 2 in 'vm.overcommit_memory'),
687 allocations which would exceed the CommitLimit (detailed
688 above) will not be permitted. This is useful if one needs
689 to guarantee that processes will not fail due to lack of
690 memory once that memory has been successfully allocated.
691 VmallocTotal: total size of vmalloc memory area
692 VmallocUsed: amount of vmalloc area which is used
693 VmallocChunk: largest contiguous block of vmalloc area which is free
695 ..............................................................................
697 vmallocinfo:
699 Provides information about vmalloced/vmaped areas. One line per area,
700 containing the virtual address range of the area, size in bytes,
701 caller information of the creator, and optional information depending
702 on the kind of area :
704 pages=nr number of pages
705 phys=addr if a physical address was specified
706 ioremap I/O mapping (ioremap() and friends)
707 vmalloc vmalloc() area
708 vmap vmap()ed pages
709 user VM_USERMAP area
710 vpages buffer for pages pointers was vmalloced (huge area)
711 N<node>=nr (Only on NUMA kernels)
712 Number of pages allocated on memory node <node>
714 > cat /proc/vmallocinfo
715 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
716 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
717 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
718 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
719 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
720 phys=7fee8000 ioremap
721 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
722 phys=7fee7000 ioremap
723 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
724 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
725 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
726 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
727 pages=2 vmalloc N1=2
728 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
729 /0x130 [x_tables] pages=4 vmalloc N0=4
730 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
731 pages=14 vmalloc N2=14
732 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
733 pages=4 vmalloc N1=4
734 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
735 pages=2 vmalloc N1=2
736 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
737 pages=10 vmalloc N0=10
739 ..............................................................................
741 softirqs:
743 Provides counts of softirq handlers serviced since boot time, for each cpu.
745 > cat /proc/softirqs
746 CPU0 CPU1 CPU2 CPU3
747 HI: 0 0 0 0
748 TIMER: 27166 27120 27097 27034
749 NET_TX: 0 0 0 17
750 NET_RX: 42 0 0 39
751 BLOCK: 0 0 107 1121
752 TASKLET: 0 0 0 290
753 SCHED: 27035 26983 26971 26746
754 HRTIMER: 0 0 0 0
755 RCU: 1678 1769 2178 2250
758 1.3 IDE devices in /proc/ide
759 ----------------------------
761 The subdirectory /proc/ide contains information about all IDE devices of which
762 the kernel is aware. There is one subdirectory for each IDE controller, the
763 file drivers and a link for each IDE device, pointing to the device directory
764 in the controller specific subtree.
766 The file drivers contains general information about the drivers used for the
767 IDE devices:
769 > cat /proc/ide/drivers
770 ide-cdrom version 4.53
771 ide-disk version 1.08
773 More detailed information can be found in the controller specific
774 subdirectories. These are named ide0, ide1 and so on. Each of these
775 directories contains the files shown in table 1-6.
778 Table 1-6: IDE controller info in /proc/ide/ide?
779 ..............................................................................
780 File Content
781 channel IDE channel (0 or 1)
782 config Configuration (only for PCI/IDE bridge)
783 mate Mate name
784 model Type/Chipset of IDE controller
785 ..............................................................................
787 Each device connected to a controller has a separate subdirectory in the
788 controllers directory. The files listed in table 1-7 are contained in these
789 directories.
792 Table 1-7: IDE device information
793 ..............................................................................
794 File Content
795 cache The cache
796 capacity Capacity of the medium (in 512Byte blocks)
797 driver driver and version
798 geometry physical and logical geometry
799 identify device identify block
800 media media type
801 model device identifier
802 settings device setup
803 smart_thresholds IDE disk management thresholds
804 smart_values IDE disk management values
805 ..............................................................................
807 The most interesting file is settings. This file contains a nice overview of
808 the drive parameters:
810 # cat /proc/ide/ide0/hda/settings
811 name value min max mode
812 ---- ----- --- --- ----
813 bios_cyl 526 0 65535 rw
814 bios_head 255 0 255 rw
815 bios_sect 63 0 63 rw
816 breada_readahead 4 0 127 rw
817 bswap 0 0 1 r
818 file_readahead 72 0 2097151 rw
819 io_32bit 0 0 3 rw
820 keepsettings 0 0 1 rw
821 max_kb_per_request 122 1 127 rw
822 multcount 0 0 8 rw
823 nice1 1 0 1 rw
824 nowerr 0 0 1 rw
825 pio_mode write-only 0 255 w
826 slow 0 0 1 rw
827 unmaskirq 0 0 1 rw
828 using_dma 0 0 1 rw
831 1.4 Networking info in /proc/net
832 --------------------------------
834 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
835 additional values you get for IP version 6 if you configure the kernel to
836 support this. Table 1-9 lists the files and their meaning.
839 Table 1-8: IPv6 info in /proc/net
840 ..............................................................................
841 File Content
842 udp6 UDP sockets (IPv6)
843 tcp6 TCP sockets (IPv6)
844 raw6 Raw device statistics (IPv6)
845 igmp6 IP multicast addresses, which this host joined (IPv6)
846 if_inet6 List of IPv6 interface addresses
847 ipv6_route Kernel routing table for IPv6
848 rt6_stats Global IPv6 routing tables statistics
849 sockstat6 Socket statistics (IPv6)
850 snmp6 Snmp data (IPv6)
851 ..............................................................................
854 Table 1-9: Network info in /proc/net
855 ..............................................................................
856 File Content
857 arp Kernel ARP table
858 dev network devices with statistics
859 dev_mcast the Layer2 multicast groups a device is listening too
860 (interface index, label, number of references, number of bound
861 addresses).
862 dev_stat network device status
863 ip_fwchains Firewall chain linkage
864 ip_fwnames Firewall chain names
865 ip_masq Directory containing the masquerading tables
866 ip_masquerade Major masquerading table
867 netstat Network statistics
868 raw raw device statistics
869 route Kernel routing table
870 rpc Directory containing rpc info
871 rt_cache Routing cache
872 snmp SNMP data
873 sockstat Socket statistics
874 tcp TCP sockets
875 tr_rif Token ring RIF routing table
876 udp UDP sockets
877 unix UNIX domain sockets
878 wireless Wireless interface data (Wavelan etc)
879 igmp IP multicast addresses, which this host joined
880 psched Global packet scheduler parameters.
881 netlink List of PF_NETLINK sockets
882 ip_mr_vifs List of multicast virtual interfaces
883 ip_mr_cache List of multicast routing cache
884 ..............................................................................
886 You can use this information to see which network devices are available in
887 your system and how much traffic was routed over those devices:
889 > cat /proc/net/dev
890 Inter-|Receive |[...
891 face |bytes packets errs drop fifo frame compressed multicast|[...
892 lo: 908188 5596 0 0 0 0 0 0 [...
893 ppp0:15475140 20721 410 0 0 410 0 0 [...
894 eth0: 614530 7085 0 0 0 0 0 1 [...
896 ...] Transmit
897 ...] bytes packets errs drop fifo colls carrier compressed
898 ...] 908188 5596 0 0 0 0 0 0
899 ...] 1375103 17405 0 0 0 0 0 0
900 ...] 1703981 5535 0 0 0 3 0 0
902 In addition, each Channel Bond interface has it's own directory. For
903 example, the bond0 device will have a directory called /proc/net/bond0/.
904 It will contain information that is specific to that bond, such as the
905 current slaves of the bond, the link status of the slaves, and how
906 many times the slaves link has failed.
908 1.5 SCSI info
909 -------------
911 If you have a SCSI host adapter in your system, you'll find a subdirectory
912 named after the driver for this adapter in /proc/scsi. You'll also see a list
913 of all recognized SCSI devices in /proc/scsi:
915 >cat /proc/scsi/scsi
916 Attached devices:
917 Host: scsi0 Channel: 00 Id: 00 Lun: 00
918 Vendor: IBM Model: DGHS09U Rev: 03E0
919 Type: Direct-Access ANSI SCSI revision: 03
920 Host: scsi0 Channel: 00 Id: 06 Lun: 00
921 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
922 Type: CD-ROM ANSI SCSI revision: 02
925 The directory named after the driver has one file for each adapter found in
926 the system. These files contain information about the controller, including
927 the used IRQ and the IO address range. The amount of information shown is
928 dependent on the adapter you use. The example shows the output for an Adaptec
929 AHA-2940 SCSI adapter:
931 > cat /proc/scsi/aic7xxx/0
933 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
934 Compile Options:
935 TCQ Enabled By Default : Disabled
936 AIC7XXX_PROC_STATS : Disabled
937 AIC7XXX_RESET_DELAY : 5
938 Adapter Configuration:
939 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
940 Ultra Wide Controller
941 PCI MMAPed I/O Base: 0xeb001000
942 Adapter SEEPROM Config: SEEPROM found and used.
943 Adaptec SCSI BIOS: Enabled
944 IRQ: 10
945 SCBs: Active 0, Max Active 2,
946 Allocated 15, HW 16, Page 255
947 Interrupts: 160328
948 BIOS Control Word: 0x18b6
949 Adapter Control Word: 0x005b
950 Extended Translation: Enabled
951 Disconnect Enable Flags: 0xffff
952 Ultra Enable Flags: 0x0001
953 Tag Queue Enable Flags: 0x0000
954 Ordered Queue Tag Flags: 0x0000
955 Default Tag Queue Depth: 8
956 Tagged Queue By Device array for aic7xxx host instance 0:
957 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
958 Actual queue depth per device for aic7xxx host instance 0:
959 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
960 Statistics:
961 (scsi0:0:0:0)
962 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
963 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
964 Total transfers 160151 (74577 reads and 85574 writes)
965 (scsi0:0:6:0)
966 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
967 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
968 Total transfers 0 (0 reads and 0 writes)
971 1.6 Parallel port info in /proc/parport
972 ---------------------------------------
974 The directory /proc/parport contains information about the parallel ports of
975 your system. It has one subdirectory for each port, named after the port
976 number (0,1,2,...).
978 These directories contain the four files shown in Table 1-10.
981 Table 1-10: Files in /proc/parport
982 ..............................................................................
983 File Content
984 autoprobe Any IEEE-1284 device ID information that has been acquired.
985 devices list of the device drivers using that port. A + will appear by the
986 name of the device currently using the port (it might not appear
987 against any).
988 hardware Parallel port's base address, IRQ line and DMA channel.
989 irq IRQ that parport is using for that port. This is in a separate
990 file to allow you to alter it by writing a new value in (IRQ
991 number or none).
992 ..............................................................................
994 1.7 TTY info in /proc/tty
995 -------------------------
997 Information about the available and actually used tty's can be found in the
998 directory /proc/tty.You'll find entries for drivers and line disciplines in
999 this directory, as shown in Table 1-11.
1002 Table 1-11: Files in /proc/tty
1003 ..............................................................................
1004 File Content
1005 drivers list of drivers and their usage
1006 ldiscs registered line disciplines
1007 driver/serial usage statistic and status of single tty lines
1008 ..............................................................................
1010 To see which tty's are currently in use, you can simply look into the file
1011 /proc/tty/drivers:
1013 > cat /proc/tty/drivers
1014 pty_slave /dev/pts 136 0-255 pty:slave
1015 pty_master /dev/ptm 128 0-255 pty:master
1016 pty_slave /dev/ttyp 3 0-255 pty:slave
1017 pty_master /dev/pty 2 0-255 pty:master
1018 serial /dev/cua 5 64-67 serial:callout
1019 serial /dev/ttyS 4 64-67 serial
1020 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1021 /dev/ptmx /dev/ptmx 5 2 system
1022 /dev/console /dev/console 5 1 system:console
1023 /dev/tty /dev/tty 5 0 system:/dev/tty
1024 unknown /dev/tty 4 1-63 console
1027 1.8 Miscellaneous kernel statistics in /proc/stat
1028 -------------------------------------------------
1030 Various pieces of information about kernel activity are available in the
1031 /proc/stat file. All of the numbers reported in this file are aggregates
1032 since the system first booted. For a quick look, simply cat the file:
1034 > cat /proc/stat
1035 cpu 2255 34 2290 22625563 6290 127 456 0
1036 cpu0 1132 34 1441 11311718 3675 127 438 0
1037 cpu1 1123 0 849 11313845 2614 0 18 0
1038 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1039 ctxt 1990473
1040 btime 1062191376
1041 processes 2915
1042 procs_running 1
1043 procs_blocked 0
1044 softirq 183433 0 21755 12 39 1137 231 21459 2263
1046 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1047 lines. These numbers identify the amount of time the CPU has spent performing
1048 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1049 second). The meanings of the columns are as follows, from left to right:
1051 - user: normal processes executing in user mode
1052 - nice: niced processes executing in user mode
1053 - system: processes executing in kernel mode
1054 - idle: twiddling thumbs
1055 - iowait: waiting for I/O to complete
1056 - irq: servicing interrupts
1057 - softirq: servicing softirqs
1058 - steal: involuntary wait
1060 The "intr" line gives counts of interrupts serviced since boot time, for each
1061 of the possible system interrupts. The first column is the total of all
1062 interrupts serviced; each subsequent column is the total for that particular
1063 interrupt.
1065 The "ctxt" line gives the total number of context switches across all CPUs.
1067 The "btime" line gives the time at which the system booted, in seconds since
1068 the Unix epoch.
1070 The "processes" line gives the number of processes and threads created, which
1071 includes (but is not limited to) those created by calls to the fork() and
1072 clone() system calls.
1074 The "procs_running" line gives the number of processes currently running on
1075 CPUs.
1077 The "procs_blocked" line gives the number of processes currently blocked,
1078 waiting for I/O to complete.
1080 The "softirq" line gives counts of softirqs serviced since boot time, for each
1081 of the possible system softirqs. The first column is the total of all
1082 softirqs serviced; each subsequent column is the total for that particular
1083 softirq.
1086 1.9 Ext4 file system parameters
1087 ------------------------------
1089 Information about mounted ext4 file systems can be found in
1090 /proc/fs/ext4. Each mounted filesystem will have a directory in
1091 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1092 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1093 in Table 1-12, below.
1095 Table 1-12: Files in /proc/fs/ext4/<devname>
1096 ..............................................................................
1097 File Content
1098 mb_groups details of multiblock allocator buddy cache of free blocks
1099 mb_history multiblock allocation history
1100 ..............................................................................
1103 ------------------------------------------------------------------------------
1104 Summary
1105 ------------------------------------------------------------------------------
1106 The /proc file system serves information about the running system. It not only
1107 allows access to process data but also allows you to request the kernel status
1108 by reading files in the hierarchy.
1110 The directory structure of /proc reflects the types of information and makes
1111 it easy, if not obvious, where to look for specific data.
1112 ------------------------------------------------------------------------------
1114 ------------------------------------------------------------------------------
1115 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1116 ------------------------------------------------------------------------------
1118 ------------------------------------------------------------------------------
1119 In This Chapter
1120 ------------------------------------------------------------------------------
1121 * Modifying kernel parameters by writing into files found in /proc/sys
1122 * Exploring the files which modify certain parameters
1123 * Review of the /proc/sys file tree
1124 ------------------------------------------------------------------------------
1127 A very interesting part of /proc is the directory /proc/sys. This is not only
1128 a source of information, it also allows you to change parameters within the
1129 kernel. Be very careful when attempting this. You can optimize your system,
1130 but you can also cause it to crash. Never alter kernel parameters on a
1131 production system. Set up a development machine and test to make sure that
1132 everything works the way you want it to. You may have no alternative but to
1133 reboot the machine once an error has been made.
1135 To change a value, simply echo the new value into the file. An example is
1136 given below in the section on the file system data. You need to be root to do
1137 this. You can create your own boot script to perform this every time your
1138 system boots.
1140 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1141 general things in the operation of the Linux kernel. Since some of the files
1142 can inadvertently disrupt your system, it is advisable to read both
1143 documentation and source before actually making adjustments. In any case, be
1144 very careful when writing to any of these files. The entries in /proc may
1145 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1146 review the kernel documentation in the directory /usr/src/linux/Documentation.
1147 This chapter is heavily based on the documentation included in the pre 2.2
1148 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1150 Please see: Documentation/sysctls/ directory for descriptions of these
1151 entries.
1153 ------------------------------------------------------------------------------
1154 Summary
1155 ------------------------------------------------------------------------------
1156 Certain aspects of kernel behavior can be modified at runtime, without the
1157 need to recompile the kernel, or even to reboot the system. The files in the
1158 /proc/sys tree can not only be read, but also modified. You can use the echo
1159 command to write value into these files, thereby changing the default settings
1160 of the kernel.
1161 ------------------------------------------------------------------------------
1163 ------------------------------------------------------------------------------
1164 CHAPTER 3: PER-PROCESS PARAMETERS
1165 ------------------------------------------------------------------------------
1167 3.1 /proc/<pid>/oom_adj - Adjust the oom-killer score
1168 ------------------------------------------------------
1170 This file can be used to adjust the score used to select which processes
1171 should be killed in an out-of-memory situation. Giving it a high score will
1172 increase the likelihood of this process being killed by the oom-killer. Valid
1173 values are in the range -16 to +15, plus the special value -17, which disables
1174 oom-killing altogether for this process.
1176 The process to be killed in an out-of-memory situation is selected among all others
1177 based on its badness score. This value equals the original memory size of the process
1178 and is then updated according to its CPU time (utime + stime) and the
1179 run time (uptime - start time). The longer it runs the smaller is the score.
1180 Badness score is divided by the square root of the CPU time and then by
1181 the double square root of the run time.
1183 Swapped out tasks are killed first. Half of each child's memory size is added to
1184 the parent's score if they do not share the same memory. Thus forking servers
1185 are the prime candidates to be killed. Having only one 'hungry' child will make
1186 parent less preferable than the child.
1188 /proc/<pid>/oom_score shows process' current badness score.
1190 The following heuristics are then applied:
1191 * if the task was reniced, its score doubles
1192 * superuser or direct hardware access tasks (CAP_SYS_ADMIN, CAP_SYS_RESOURCE
1193 or CAP_SYS_RAWIO) have their score divided by 4
1194 * if oom condition happened in one cpuset and checked task does not belong
1195 to it, its score is divided by 8
1196 * the resulting score is multiplied by two to the power of oom_adj, i.e.
1197 points <<= oom_adj when it is positive and
1198 points >>= -(oom_adj) otherwise
1200 The task with the highest badness score is then selected and its children
1201 are killed, process itself will be killed in an OOM situation when it does
1202 not have children or some of them disabled oom like described above.
1204 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1205 -------------------------------------------------------------
1207 This file can be used to check the current score used by the oom-killer is for
1208 any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
1209 process should be killed in an out-of-memory situation.
1212 3.3 /proc/<pid>/io - Display the IO accounting fields
1213 -------------------------------------------------------
1215 This file contains IO statistics for each running process
1217 Example
1218 -------
1220 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1221 [1] 3828
1223 test:/tmp # cat /proc/3828/io
1224 rchar: 323934931
1225 wchar: 323929600
1226 syscr: 632687
1227 syscw: 632675
1228 read_bytes: 0
1229 write_bytes: 323932160
1230 cancelled_write_bytes: 0
1233 Description
1234 -----------
1236 rchar
1237 -----
1239 I/O counter: chars read
1240 The number of bytes which this task has caused to be read from storage. This
1241 is simply the sum of bytes which this process passed to read() and pread().
1242 It includes things like tty IO and it is unaffected by whether or not actual
1243 physical disk IO was required (the read might have been satisfied from
1244 pagecache)
1247 wchar
1248 -----
1250 I/O counter: chars written
1251 The number of bytes which this task has caused, or shall cause to be written
1252 to disk. Similar caveats apply here as with rchar.
1255 syscr
1256 -----
1258 I/O counter: read syscalls
1259 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1260 and pread().
1263 syscw
1264 -----
1266 I/O counter: write syscalls
1267 Attempt to count the number of write I/O operations, i.e. syscalls like
1268 write() and pwrite().
1271 read_bytes
1272 ----------
1274 I/O counter: bytes read
1275 Attempt to count the number of bytes which this process really did cause to
1276 be fetched from the storage layer. Done at the submit_bio() level, so it is
1277 accurate for block-backed filesystems. <please add status regarding NFS and
1278 CIFS at a later time>
1281 write_bytes
1282 -----------
1284 I/O counter: bytes written
1285 Attempt to count the number of bytes which this process caused to be sent to
1286 the storage layer. This is done at page-dirtying time.
1289 cancelled_write_bytes
1290 ---------------------
1292 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1293 then deletes the file, it will in fact perform no writeout. But it will have
1294 been accounted as having caused 1MB of write.
1295 In other words: The number of bytes which this process caused to not happen,
1296 by truncating pagecache. A task can cause "negative" IO too. If this task
1297 truncates some dirty pagecache, some IO which another task has been accounted
1298 for (in it's write_bytes) will not be happening. We _could_ just subtract that
1299 from the truncating task's write_bytes, but there is information loss in doing
1300 that.
1303 Note
1304 ----
1306 At its current implementation state, this is a bit racy on 32-bit machines: if
1307 process A reads process B's /proc/pid/io while process B is updating one of
1308 those 64-bit counters, process A could see an intermediate result.
1311 More information about this can be found within the taskstats documentation in
1312 Documentation/accounting.
1314 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1315 ---------------------------------------------------------------
1316 When a process is dumped, all anonymous memory is written to a core file as
1317 long as the size of the core file isn't limited. But sometimes we don't want
1318 to dump some memory segments, for example, huge shared memory. Conversely,
1319 sometimes we want to save file-backed memory segments into a core file, not
1320 only the individual files.
1322 /proc/<pid>/coredump_filter allows you to customize which memory segments
1323 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1324 of memory types. If a bit of the bitmask is set, memory segments of the
1325 corresponding memory type are dumped, otherwise they are not dumped.
1327 The following 7 memory types are supported:
1328 - (bit 0) anonymous private memory
1329 - (bit 1) anonymous shared memory
1330 - (bit 2) file-backed private memory
1331 - (bit 3) file-backed shared memory
1332 - (bit 4) ELF header pages in file-backed private memory areas (it is
1333 effective only if the bit 2 is cleared)
1334 - (bit 5) hugetlb private memory
1335 - (bit 6) hugetlb shared memory
1337 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1338 are always dumped regardless of the bitmask status.
1340 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1341 effected by bit 5-6.
1343 Default value of coredump_filter is 0x23; this means all anonymous memory
1344 segments and hugetlb private memory are dumped.
1346 If you don't want to dump all shared memory segments attached to pid 1234,
1347 write 0x21 to the process's proc file.
1349 $ echo 0x21 > /proc/1234/coredump_filter
1351 When a new process is created, the process inherits the bitmask status from its
1352 parent. It is useful to set up coredump_filter before the program runs.
1353 For example:
1355 $ echo 0x7 > /proc/self/coredump_filter
1356 $ ./some_program
1358 3.5 /proc/<pid>/mountinfo - Information about mounts
1359 --------------------------------------------------------
1361 This file contains lines of the form:
1363 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1364 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1366 (1) mount ID: unique identifier of the mount (may be reused after umount)
1367 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1368 (3) major:minor: value of st_dev for files on filesystem
1369 (4) root: root of the mount within the filesystem
1370 (5) mount point: mount point relative to the process's root
1371 (6) mount options: per mount options
1372 (7) optional fields: zero or more fields of the form "tag[:value]"
1373 (8) separator: marks the end of the optional fields
1374 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1375 (10) mount source: filesystem specific information or "none"
1376 (11) super options: per super block options
1378 Parsers should ignore all unrecognised optional fields. Currently the
1379 possible optional fields are:
1381 shared:X mount is shared in peer group X
1382 master:X mount is slave to peer group X
1383 propagate_from:X mount is slave and receives propagation from peer group X (*)
1384 unbindable mount is unbindable
1386 (*) X is the closest dominant peer group under the process's root. If
1387 X is the immediate master of the mount, or if there's no dominant peer
1388 group under the same root, then only the "master:X" field is present
1389 and not the "propagate_from:X" field.
1391 For more information on mount propagation see:
1393 Documentation/filesystems/sharedsubtree.txt