1 ------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3 ------------------------------------------------------------------------------
4 /proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
7 2.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8 move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
9 ------------------------------------------------------------------------------
10 Version 1.3 Kernel version 2.2.12
11 Kernel version 2.4.0-test11-pre4
12 ------------------------------------------------------------------------------
13 fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
19 0.1 Introduction/Credits
22 1 Collecting System Information
23 1.1 Process-Specific Subdirectories
25 1.3 IDE devices in /proc/ide
26 1.4 Networking info in /proc/net
28 1.6 Parallel port info in /proc/parport
29 1.7 TTY info in /proc/tty
30 1.8 Miscellaneous kernel statistics in /proc/stat
31 1.9 Ext4 file system parameters
33 2 Modifying System Parameters
35 3 Per-Process Parameters
36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
38 3.2 /proc/<pid>/oom_score - Display current oom-killer score
39 3.3 /proc/<pid>/io - Display the IO accounting fields
40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
41 3.5 /proc/<pid>/mountinfo - Information about mounts
42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
43 3.7 /proc/<pid>/task/<tid>/children - Information about task children
48 ------------------------------------------------------------------------------
50 ------------------------------------------------------------------------------
52 0.1 Introduction/Credits
53 ------------------------
55 This documentation is part of a soon (or so we hope) to be released book on
56 the SuSE Linux distribution. As there is no complete documentation for the
57 /proc file system and we've used many freely available sources to write these
58 chapters, it seems only fair to give the work back to the Linux community.
59 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
60 afraid it's still far from complete, but we hope it will be useful. As far as
61 we know, it is the first 'all-in-one' document about the /proc file system. It
62 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
63 SPARC, AXP, etc., features, you probably won't find what you are looking for.
64 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
65 additions and patches are welcome and will be added to this document if you
68 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
69 other people for help compiling this documentation. We'd also like to extend a
70 special thank you to Andi Kleen for documentation, which we relied on heavily
71 to create this document, as well as the additional information he provided.
72 Thanks to everybody else who contributed source or docs to the Linux kernel
73 and helped create a great piece of software... :)
75 If you have any comments, corrections or additions, please don't hesitate to
76 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
79 The latest version of this document is available online at
80 http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
82 If the above direction does not works for you, you could try the kernel
83 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
84 comandante@zaralinux.com.
89 We don't guarantee the correctness of this document, and if you come to us
90 complaining about how you screwed up your system because of incorrect
91 documentation, we won't feel responsible...
93 ------------------------------------------------------------------------------
94 CHAPTER 1: COLLECTING SYSTEM INFORMATION
95 ------------------------------------------------------------------------------
97 ------------------------------------------------------------------------------
99 ------------------------------------------------------------------------------
100 * Investigating the properties of the pseudo file system /proc and its
101 ability to provide information on the running Linux system
102 * Examining /proc's structure
103 * Uncovering various information about the kernel and the processes running
105 ------------------------------------------------------------------------------
108 The proc file system acts as an interface to internal data structures in the
109 kernel. It can be used to obtain information about the system and to change
110 certain kernel parameters at runtime (sysctl).
112 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
113 show you how you can use /proc/sys to change settings.
115 1.1 Process-Specific Subdirectories
116 -----------------------------------
118 The directory /proc contains (among other things) one subdirectory for each
119 process running on the system, which is named after the process ID (PID).
121 The link self points to the process reading the file system. Each process
122 subdirectory has the entries listed in Table 1-1.
125 Table 1-1: Process specific entries in /proc
126 ..............................................................................
128 clear_refs Clears page referenced bits shown in smaps output
129 cmdline Command line arguments
130 cpu Current and last cpu in which it was executed (2.4)(smp)
131 cwd Link to the current working directory
132 environ Values of environment variables
133 exe Link to the executable of this process
134 fd Directory, which contains all file descriptors
135 maps Memory maps to executables and library files (2.4)
136 mem Memory held by this process
137 root Link to the root directory of this process
139 statm Process memory status information
140 status Process status in human readable form
141 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
143 stack Report full stack trace, enable via CONFIG_STACKTRACE
144 smaps a extension based on maps, showing the memory consumption of
146 ..............................................................................
148 For example, to get the status information of a process, all you have to do is
149 read the file /proc/PID/status:
151 >cat /proc/self/status
175 SigPnd: 0000000000000000
176 ShdPnd: 0000000000000000
177 SigBlk: 0000000000000000
178 SigIgn: 0000000000000000
179 SigCgt: 0000000000000000
180 CapInh: 00000000fffffeff
181 CapPrm: 0000000000000000
182 CapEff: 0000000000000000
183 CapBnd: ffffffffffffffff
184 voluntary_ctxt_switches: 0
185 nonvoluntary_ctxt_switches: 1
187 This shows you nearly the same information you would get if you viewed it with
188 the ps command. In fact, ps uses the proc file system to obtain its
189 information. But you get a more detailed view of the process by reading the
190 file /proc/PID/status. It fields are described in table 1-2.
192 The statm file contains more detailed information about the process
193 memory usage. Its seven fields are explained in Table 1-3. The stat file
194 contains details information about the process itself. Its fields are
195 explained in Table 1-4.
197 (for SMP CONFIG users)
198 For making accounting scalable, RSS related information are handled in
199 asynchronous manner and the vaule may not be very precise. To see a precise
200 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
201 It's slow but very precise.
203 Table 1-2: Contents of the status files (as of 2.6.30-rc7)
204 ..............................................................................
206 Name filename of the executable
207 State state (R is running, S is sleeping, D is sleeping
208 in an uninterruptible wait, Z is zombie,
209 T is traced or stopped)
212 PPid process id of the parent process
213 TracerPid PID of process tracing this process (0 if not)
214 Uid Real, effective, saved set, and file system UIDs
215 Gid Real, effective, saved set, and file system GIDs
216 FDSize number of file descriptor slots currently allocated
217 Groups supplementary group list
218 VmPeak peak virtual memory size
219 VmSize total program size
220 VmLck locked memory size
221 VmHWM peak resident set size ("high water mark")
222 VmRSS size of memory portions
223 VmData size of data, stack, and text segments
224 VmStk size of data, stack, and text segments
225 VmExe size of text segment
226 VmLib size of shared library code
227 VmPTE size of page table entries
228 VmSwap size of swap usage (the number of referred swapents)
229 Threads number of threads
230 SigQ number of signals queued/max. number for queue
231 SigPnd bitmap of pending signals for the thread
232 ShdPnd bitmap of shared pending signals for the process
233 SigBlk bitmap of blocked signals
234 SigIgn bitmap of ignored signals
235 SigCgt bitmap of catched signals
236 CapInh bitmap of inheritable capabilities
237 CapPrm bitmap of permitted capabilities
238 CapEff bitmap of effective capabilities
239 CapBnd bitmap of capabilities bounding set
240 Cpus_allowed mask of CPUs on which this process may run
241 Cpus_allowed_list Same as previous, but in "list format"
242 Mems_allowed mask of memory nodes allowed to this process
243 Mems_allowed_list Same as previous, but in "list format"
244 voluntary_ctxt_switches number of voluntary context switches
245 nonvoluntary_ctxt_switches number of non voluntary context switches
246 ..............................................................................
248 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
249 ..............................................................................
251 size total program size (pages) (same as VmSize in status)
252 resident size of memory portions (pages) (same as VmRSS in status)
253 shared number of pages that are shared (i.e. backed by a file)
254 trs number of pages that are 'code' (not including libs; broken,
255 includes data segment)
256 lrs number of pages of library (always 0 on 2.6)
257 drs number of pages of data/stack (including libs; broken,
258 includes library text)
259 dt number of dirty pages (always 0 on 2.6)
260 ..............................................................................
263 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
264 ..............................................................................
267 tcomm filename of the executable
268 state state (R is running, S is sleeping, D is sleeping in an
269 uninterruptible wait, Z is zombie, T is traced or stopped)
270 ppid process id of the parent process
271 pgrp pgrp of the process
273 tty_nr tty the process uses
274 tty_pgrp pgrp of the tty
276 min_flt number of minor faults
277 cmin_flt number of minor faults with child's
278 maj_flt number of major faults
279 cmaj_flt number of major faults with child's
280 utime user mode jiffies
281 stime kernel mode jiffies
282 cutime user mode jiffies with child's
283 cstime kernel mode jiffies with child's
284 priority priority level
286 num_threads number of threads
287 it_real_value (obsolete, always 0)
288 start_time time the process started after system boot
289 vsize virtual memory size
290 rss resident set memory size
291 rsslim current limit in bytes on the rss
292 start_code address above which program text can run
293 end_code address below which program text can run
294 start_stack address of the start of the main process stack
295 esp current value of ESP
296 eip current value of EIP
297 pending bitmap of pending signals
298 blocked bitmap of blocked signals
299 sigign bitmap of ignored signals
300 sigcatch bitmap of catched signals
301 wchan address where process went to sleep
304 exit_signal signal to send to parent thread on exit
305 task_cpu which CPU the task is scheduled on
306 rt_priority realtime priority
307 policy scheduling policy (man sched_setscheduler)
308 blkio_ticks time spent waiting for block IO
309 gtime guest time of the task in jiffies
310 cgtime guest time of the task children in jiffies
311 start_data address above which program data+bss is placed
312 end_data address below which program data+bss is placed
313 start_brk address above which program heap can be expanded with brk()
314 ..............................................................................
316 The /proc/PID/maps file containing the currently mapped memory regions and
317 their access permissions.
321 address perms offset dev inode pathname
323 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
324 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
325 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
326 a7cb1000-a7cb2000 ---p 00000000 00:00 0
327 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
328 a7eb2000-a7eb3000 ---p 00000000 00:00 0
329 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
330 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
331 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
332 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
333 a800b000-a800e000 rw-p 00000000 00:00 0
334 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
335 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
336 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
337 a8024000-a8027000 rw-p 00000000 00:00 0
338 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
339 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
340 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
341 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
342 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
344 where "address" is the address space in the process that it occupies, "perms"
345 is a set of permissions:
351 p = private (copy on write)
353 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
354 "inode" is the inode on that device. 0 indicates that no inode is associated
355 with the memory region, as the case would be with BSS (uninitialized data).
356 The "pathname" shows the name associated file for this mapping. If the mapping
357 is not associated with a file:
359 [heap] = the heap of the program
360 [stack] = the stack of the main process
361 [stack:1001] = the stack of the thread with tid 1001
362 [vdso] = the "virtual dynamic shared object",
363 the kernel system call handler
365 or if empty, the mapping is anonymous.
367 The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
368 of the individual tasks of a process. In this file you will see a mapping marked
369 as [stack] if that task sees it as a stack. This is a key difference from the
370 content of /proc/PID/maps, where you will see all mappings that are being used
371 as stack by all of those tasks. Hence, for the example above, the task-level
372 map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
374 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
375 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
376 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
377 a7cb1000-a7cb2000 ---p 00000000 00:00 0
378 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
379 a7eb2000-a7eb3000 ---p 00000000 00:00 0
380 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
381 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
382 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
383 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
384 a800b000-a800e000 rw-p 00000000 00:00 0
385 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
386 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
387 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
388 a8024000-a8027000 rw-p 00000000 00:00 0
389 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
390 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
391 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
392 aff35000-aff4a000 rw-p 00000000 00:00 0
393 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
395 The /proc/PID/smaps is an extension based on maps, showing the memory
396 consumption for each of the process's mappings. For each of mappings there
397 is a series of lines such as the following:
399 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
414 The first of these lines shows the same information as is displayed for the
415 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
416 (size), the amount of the mapping that is currently resident in RAM (RSS), the
417 process' proportional share of this mapping (PSS), the number of clean and
418 dirty private pages in the mapping. Note that even a page which is part of a
419 MAP_SHARED mapping, but has only a single pte mapped, i.e. is currently used
420 by only one process, is accounted as private and not as shared. "Referenced"
421 indicates the amount of memory currently marked as referenced or accessed.
422 "Anonymous" shows the amount of memory that does not belong to any file. Even
423 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
424 and a page is modified, the file page is replaced by a private anonymous copy.
425 "Swap" shows how much would-be-anonymous memory is also used, but out on
428 This file is only present if the CONFIG_MMU kernel configuration option is
431 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
432 bits on both physical and virtual pages associated with a process.
433 To clear the bits for all the pages associated with the process
434 > echo 1 > /proc/PID/clear_refs
436 To clear the bits for the anonymous pages associated with the process
437 > echo 2 > /proc/PID/clear_refs
439 To clear the bits for the file mapped pages associated with the process
440 > echo 3 > /proc/PID/clear_refs
441 Any other value written to /proc/PID/clear_refs will have no effect.
443 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
444 using /proc/kpageflags and number of times a page is mapped using
445 /proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
450 Similar to the process entries, the kernel data files give information about
451 the running kernel. The files used to obtain this information are contained in
452 /proc and are listed in Table 1-5. Not all of these will be present in your
453 system. It depends on the kernel configuration and the loaded modules, which
454 files are there, and which are missing.
456 Table 1-5: Kernel info in /proc
457 ..............................................................................
459 apm Advanced power management info
460 buddyinfo Kernel memory allocator information (see text) (2.5)
461 bus Directory containing bus specific information
462 cmdline Kernel command line
463 cpuinfo Info about the CPU
464 devices Available devices (block and character)
465 dma Used DMS channels
466 filesystems Supported filesystems
467 driver Various drivers grouped here, currently rtc (2.4)
468 execdomains Execdomains, related to security (2.4)
469 fb Frame Buffer devices (2.4)
470 fs File system parameters, currently nfs/exports (2.4)
471 ide Directory containing info about the IDE subsystem
472 interrupts Interrupt usage
473 iomem Memory map (2.4)
474 ioports I/O port usage
475 irq Masks for irq to cpu affinity (2.4)(smp?)
476 isapnp ISA PnP (Plug&Play) Info (2.4)
477 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
479 ksyms Kernel symbol table
480 loadavg Load average of last 1, 5 & 15 minutes
484 modules List of loaded modules
485 mounts Mounted filesystems
486 net Networking info (see text)
487 pagetypeinfo Additional page allocator information (see text) (2.5)
488 partitions Table of partitions known to the system
489 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
490 decoupled by lspci (2.4)
492 scsi SCSI info (see text)
493 slabinfo Slab pool info
494 softirqs softirq usage
495 stat Overall statistics
496 swaps Swap space utilization
498 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
499 tty Info of tty drivers
501 version Kernel version
502 video bttv info of video resources (2.4)
503 vmallocinfo Show vmalloced areas
504 ..............................................................................
506 You can, for example, check which interrupts are currently in use and what
507 they are used for by looking in the file /proc/interrupts:
509 > cat /proc/interrupts
511 0: 8728810 XT-PIC timer
512 1: 895 XT-PIC keyboard
514 3: 531695 XT-PIC aha152x
515 4: 2014133 XT-PIC serial
516 5: 44401 XT-PIC pcnet_cs
519 12: 182918 XT-PIC PS/2 Mouse
521 14: 1232265 XT-PIC ide0
525 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
526 output of a SMP machine):
528 > cat /proc/interrupts
531 0: 1243498 1214548 IO-APIC-edge timer
532 1: 8949 8958 IO-APIC-edge keyboard
533 2: 0 0 XT-PIC cascade
534 5: 11286 10161 IO-APIC-edge soundblaster
535 8: 1 0 IO-APIC-edge rtc
536 9: 27422 27407 IO-APIC-edge 3c503
537 12: 113645 113873 IO-APIC-edge PS/2 Mouse
539 14: 22491 24012 IO-APIC-edge ide0
540 15: 2183 2415 IO-APIC-edge ide1
541 17: 30564 30414 IO-APIC-level eth0
542 18: 177 164 IO-APIC-level bttv
547 NMI is incremented in this case because every timer interrupt generates a NMI
548 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
550 LOC is the local interrupt counter of the internal APIC of every CPU.
552 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
553 connects the CPUs in a SMP system. This means that an error has been detected,
554 the IO-APIC automatically retry the transmission, so it should not be a big
555 problem, but you should read the SMP-FAQ.
557 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
558 /proc/interrupts to display every IRQ vector in use by the system, not
559 just those considered 'most important'. The new vectors are:
561 THR -- interrupt raised when a machine check threshold counter
562 (typically counting ECC corrected errors of memory or cache) exceeds
563 a configurable threshold. Only available on some systems.
565 TRM -- a thermal event interrupt occurs when a temperature threshold
566 has been exceeded for the CPU. This interrupt may also be generated
567 when the temperature drops back to normal.
569 SPU -- a spurious interrupt is some interrupt that was raised then lowered
570 by some IO device before it could be fully processed by the APIC. Hence
571 the APIC sees the interrupt but does not know what device it came from.
572 For this case the APIC will generate the interrupt with a IRQ vector
573 of 0xff. This might also be generated by chipset bugs.
575 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
576 sent from one CPU to another per the needs of the OS. Typically,
577 their statistics are used by kernel developers and interested users to
578 determine the occurrence of interrupts of the given type.
580 The above IRQ vectors are displayed only when relevant. For example,
581 the threshold vector does not exist on x86_64 platforms. Others are
582 suppressed when the system is a uniprocessor. As of this writing, only
583 i386 and x86_64 platforms support the new IRQ vector displays.
585 Of some interest is the introduction of the /proc/irq directory to 2.4.
586 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
587 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
588 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
593 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
594 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
598 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
599 IRQ, you can set it by doing:
601 > echo 1 > /proc/irq/10/smp_affinity
603 This means that only the first CPU will handle the IRQ, but you can also echo
604 5 which means that only the first and fourth CPU can handle the IRQ.
606 The contents of each smp_affinity file is the same by default:
608 > cat /proc/irq/0/smp_affinity
611 There is an alternate interface, smp_affinity_list which allows specifying
612 a cpu range instead of a bitmask:
614 > cat /proc/irq/0/smp_affinity_list
617 The default_smp_affinity mask applies to all non-active IRQs, which are the
618 IRQs which have not yet been allocated/activated, and hence which lack a
619 /proc/irq/[0-9]* directory.
621 The node file on an SMP system shows the node to which the device using the IRQ
622 reports itself as being attached. This hardware locality information does not
623 include information about any possible driver locality preference.
625 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
626 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
628 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
629 between all the CPUs which are allowed to handle it. As usual the kernel has
630 more info than you and does a better job than you, so the defaults are the
631 best choice for almost everyone. [Note this applies only to those IO-APIC's
632 that support "Round Robin" interrupt distribution.]
634 There are three more important subdirectories in /proc: net, scsi, and sys.
635 The general rule is that the contents, or even the existence of these
636 directories, depend on your kernel configuration. If SCSI is not enabled, the
637 directory scsi may not exist. The same is true with the net, which is there
638 only when networking support is present in the running kernel.
640 The slabinfo file gives information about memory usage at the slab level.
641 Linux uses slab pools for memory management above page level in version 2.2.
642 Commonly used objects have their own slab pool (such as network buffers,
643 directory cache, and so on).
645 ..............................................................................
647 > cat /proc/buddyinfo
649 Node 0, zone DMA 0 4 5 4 4 3 ...
650 Node 0, zone Normal 1 0 0 1 101 8 ...
651 Node 0, zone HighMem 2 0 0 1 1 0 ...
653 External fragmentation is a problem under some workloads, and buddyinfo is a
654 useful tool for helping diagnose these problems. Buddyinfo will give you a
655 clue as to how big an area you can safely allocate, or why a previous
658 Each column represents the number of pages of a certain order which are
659 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
660 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
661 available in ZONE_NORMAL, etc...
663 More information relevant to external fragmentation can be found in
666 > cat /proc/pagetypeinfo
670 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
671 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
672 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
673 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
674 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
675 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
676 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
677 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
678 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
679 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
680 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
682 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
683 Node 0, zone DMA 2 0 5 1 0
684 Node 0, zone DMA32 41 6 967 2 0
686 Fragmentation avoidance in the kernel works by grouping pages of different
687 migrate types into the same contiguous regions of memory called page blocks.
688 A page block is typically the size of the default hugepage size e.g. 2MB on
689 X86-64. By keeping pages grouped based on their ability to move, the kernel
690 can reclaim pages within a page block to satisfy a high-order allocation.
692 The pagetypinfo begins with information on the size of a page block. It
693 then gives the same type of information as buddyinfo except broken down
694 by migrate-type and finishes with details on how many page blocks of each
697 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
698 from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
699 make an estimate of the likely number of huge pages that can be allocated
700 at a given point in time. All the "Movable" blocks should be allocatable
701 unless memory has been mlock()'d. Some of the Reclaimable blocks should
702 also be allocatable although a lot of filesystem metadata may have to be
703 reclaimed to achieve this.
705 ..............................................................................
709 Provides information about distribution and utilization of memory. This
710 varies by architecture and compile options. The following is from a
711 16GB PIII, which has highmem enabled. You may not have all of these fields.
715 The "Locked" indicates whether the mapping is locked in memory or not.
718 MemTotal: 16344972 kB
725 HighTotal: 15597528 kB
726 HighFree: 13629632 kB
736 SReclaimable: 159856 kB
737 SUnreclaim: 124508 kB
742 CommitLimit: 7669796 kB
743 Committed_AS: 100056 kB
744 VmallocTotal: 112216 kB
746 VmallocChunk: 111088 kB
747 AnonHugePages: 49152 kB
749 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
750 bits and the kernel binary code)
751 MemFree: The sum of LowFree+HighFree
752 Buffers: Relatively temporary storage for raw disk blocks
753 shouldn't get tremendously large (20MB or so)
754 Cached: in-memory cache for files read from the disk (the
755 pagecache). Doesn't include SwapCached
756 SwapCached: Memory that once was swapped out, is swapped back in but
757 still also is in the swapfile (if memory is needed it
758 doesn't need to be swapped out AGAIN because it is already
759 in the swapfile. This saves I/O)
760 Active: Memory that has been used more recently and usually not
761 reclaimed unless absolutely necessary.
762 Inactive: Memory which has been less recently used. It is more
763 eligible to be reclaimed for other purposes
765 HighFree: Highmem is all memory above ~860MB of physical memory
766 Highmem areas are for use by userspace programs, or
767 for the pagecache. The kernel must use tricks to access
768 this memory, making it slower to access than lowmem.
770 LowFree: Lowmem is memory which can be used for everything that
771 highmem can be used for, but it is also available for the
772 kernel's use for its own data structures. Among many
773 other things, it is where everything from the Slab is
774 allocated. Bad things happen when you're out of lowmem.
775 SwapTotal: total amount of swap space available
776 SwapFree: Memory which has been evicted from RAM, and is temporarily
778 Dirty: Memory which is waiting to get written back to the disk
779 Writeback: Memory which is actively being written back to the disk
780 AnonPages: Non-file backed pages mapped into userspace page tables
781 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
782 Mapped: files which have been mmaped, such as libraries
783 Slab: in-kernel data structures cache
784 SReclaimable: Part of Slab, that might be reclaimed, such as caches
785 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
786 PageTables: amount of memory dedicated to the lowest level of page
788 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
790 Bounce: Memory used for block device "bounce buffers"
791 WritebackTmp: Memory used by FUSE for temporary writeback buffers
792 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
793 this is the total amount of memory currently available to
794 be allocated on the system. This limit is only adhered to
795 if strict overcommit accounting is enabled (mode 2 in
796 'vm.overcommit_memory').
797 The CommitLimit is calculated with the following formula:
798 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
799 For example, on a system with 1G of physical RAM and 7G
800 of swap with a `vm.overcommit_ratio` of 30 it would
801 yield a CommitLimit of 7.3G.
802 For more details, see the memory overcommit documentation
803 in vm/overcommit-accounting.
804 Committed_AS: The amount of memory presently allocated on the system.
805 The committed memory is a sum of all of the memory which
806 has been allocated by processes, even if it has not been
807 "used" by them as of yet. A process which malloc()'s 1G
808 of memory, but only touches 300M of it will only show up
809 as using 300M of memory even if it has the address space
810 allocated for the entire 1G. This 1G is memory which has
811 been "committed" to by the VM and can be used at any time
812 by the allocating application. With strict overcommit
813 enabled on the system (mode 2 in 'vm.overcommit_memory'),
814 allocations which would exceed the CommitLimit (detailed
815 above) will not be permitted. This is useful if one needs
816 to guarantee that processes will not fail due to lack of
817 memory once that memory has been successfully allocated.
818 VmallocTotal: total size of vmalloc memory area
819 VmallocUsed: amount of vmalloc area which is used
820 VmallocChunk: largest contiguous block of vmalloc area which is free
822 ..............................................................................
826 Provides information about vmalloced/vmaped areas. One line per area,
827 containing the virtual address range of the area, size in bytes,
828 caller information of the creator, and optional information depending
829 on the kind of area :
831 pages=nr number of pages
832 phys=addr if a physical address was specified
833 ioremap I/O mapping (ioremap() and friends)
834 vmalloc vmalloc() area
837 vpages buffer for pages pointers was vmalloced (huge area)
838 N<node>=nr (Only on NUMA kernels)
839 Number of pages allocated on memory node <node>
841 > cat /proc/vmallocinfo
842 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
843 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
844 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
845 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
846 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
847 phys=7fee8000 ioremap
848 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
849 phys=7fee7000 ioremap
850 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
851 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
852 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
853 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
855 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
856 /0x130 [x_tables] pages=4 vmalloc N0=4
857 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
858 pages=14 vmalloc N2=14
859 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
861 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
863 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
864 pages=10 vmalloc N0=10
866 ..............................................................................
870 Provides counts of softirq handlers serviced since boot time, for each cpu.
875 TIMER: 27166 27120 27097 27034
880 SCHED: 27035 26983 26971 26746
882 RCU: 1678 1769 2178 2250
885 1.3 IDE devices in /proc/ide
886 ----------------------------
888 The subdirectory /proc/ide contains information about all IDE devices of which
889 the kernel is aware. There is one subdirectory for each IDE controller, the
890 file drivers and a link for each IDE device, pointing to the device directory
891 in the controller specific subtree.
893 The file drivers contains general information about the drivers used for the
896 > cat /proc/ide/drivers
897 ide-cdrom version 4.53
898 ide-disk version 1.08
900 More detailed information can be found in the controller specific
901 subdirectories. These are named ide0, ide1 and so on. Each of these
902 directories contains the files shown in table 1-6.
905 Table 1-6: IDE controller info in /proc/ide/ide?
906 ..............................................................................
908 channel IDE channel (0 or 1)
909 config Configuration (only for PCI/IDE bridge)
911 model Type/Chipset of IDE controller
912 ..............................................................................
914 Each device connected to a controller has a separate subdirectory in the
915 controllers directory. The files listed in table 1-7 are contained in these
919 Table 1-7: IDE device information
920 ..............................................................................
923 capacity Capacity of the medium (in 512Byte blocks)
924 driver driver and version
925 geometry physical and logical geometry
926 identify device identify block
928 model device identifier
929 settings device setup
930 smart_thresholds IDE disk management thresholds
931 smart_values IDE disk management values
932 ..............................................................................
934 The most interesting file is settings. This file contains a nice overview of
935 the drive parameters:
937 # cat /proc/ide/ide0/hda/settings
938 name value min max mode
939 ---- ----- --- --- ----
940 bios_cyl 526 0 65535 rw
941 bios_head 255 0 255 rw
943 breada_readahead 4 0 127 rw
945 file_readahead 72 0 2097151 rw
947 keepsettings 0 0 1 rw
948 max_kb_per_request 122 1 127 rw
952 pio_mode write-only 0 255 w
958 1.4 Networking info in /proc/net
959 --------------------------------
961 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
962 additional values you get for IP version 6 if you configure the kernel to
963 support this. Table 1-9 lists the files and their meaning.
966 Table 1-8: IPv6 info in /proc/net
967 ..............................................................................
969 udp6 UDP sockets (IPv6)
970 tcp6 TCP sockets (IPv6)
971 raw6 Raw device statistics (IPv6)
972 igmp6 IP multicast addresses, which this host joined (IPv6)
973 if_inet6 List of IPv6 interface addresses
974 ipv6_route Kernel routing table for IPv6
975 rt6_stats Global IPv6 routing tables statistics
976 sockstat6 Socket statistics (IPv6)
977 snmp6 Snmp data (IPv6)
978 ..............................................................................
981 Table 1-9: Network info in /proc/net
982 ..............................................................................
985 dev network devices with statistics
986 dev_mcast the Layer2 multicast groups a device is listening too
987 (interface index, label, number of references, number of bound
989 dev_stat network device status
990 ip_fwchains Firewall chain linkage
991 ip_fwnames Firewall chain names
992 ip_masq Directory containing the masquerading tables
993 ip_masquerade Major masquerading table
994 netstat Network statistics
995 raw raw device statistics
996 route Kernel routing table
997 rpc Directory containing rpc info
998 rt_cache Routing cache
1000 sockstat Socket statistics
1003 unix UNIX domain sockets
1004 wireless Wireless interface data (Wavelan etc)
1005 igmp IP multicast addresses, which this host joined
1006 psched Global packet scheduler parameters.
1007 netlink List of PF_NETLINK sockets
1008 ip_mr_vifs List of multicast virtual interfaces
1009 ip_mr_cache List of multicast routing cache
1010 ..............................................................................
1012 You can use this information to see which network devices are available in
1013 your system and how much traffic was routed over those devices:
1016 Inter-|Receive |[...
1017 face |bytes packets errs drop fifo frame compressed multicast|[...
1018 lo: 908188 5596 0 0 0 0 0 0 [...
1019 ppp0:15475140 20721 410 0 0 410 0 0 [...
1020 eth0: 614530 7085 0 0 0 0 0 1 [...
1023 ...] bytes packets errs drop fifo colls carrier compressed
1024 ...] 908188 5596 0 0 0 0 0 0
1025 ...] 1375103 17405 0 0 0 0 0 0
1026 ...] 1703981 5535 0 0 0 3 0 0
1028 In addition, each Channel Bond interface has its own directory. For
1029 example, the bond0 device will have a directory called /proc/net/bond0/.
1030 It will contain information that is specific to that bond, such as the
1031 current slaves of the bond, the link status of the slaves, and how
1032 many times the slaves link has failed.
1037 If you have a SCSI host adapter in your system, you'll find a subdirectory
1038 named after the driver for this adapter in /proc/scsi. You'll also see a list
1039 of all recognized SCSI devices in /proc/scsi:
1041 >cat /proc/scsi/scsi
1043 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1044 Vendor: IBM Model: DGHS09U Rev: 03E0
1045 Type: Direct-Access ANSI SCSI revision: 03
1046 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1047 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1048 Type: CD-ROM ANSI SCSI revision: 02
1051 The directory named after the driver has one file for each adapter found in
1052 the system. These files contain information about the controller, including
1053 the used IRQ and the IO address range. The amount of information shown is
1054 dependent on the adapter you use. The example shows the output for an Adaptec
1055 AHA-2940 SCSI adapter:
1057 > cat /proc/scsi/aic7xxx/0
1059 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1061 TCQ Enabled By Default : Disabled
1062 AIC7XXX_PROC_STATS : Disabled
1063 AIC7XXX_RESET_DELAY : 5
1064 Adapter Configuration:
1065 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1066 Ultra Wide Controller
1067 PCI MMAPed I/O Base: 0xeb001000
1068 Adapter SEEPROM Config: SEEPROM found and used.
1069 Adaptec SCSI BIOS: Enabled
1071 SCBs: Active 0, Max Active 2,
1072 Allocated 15, HW 16, Page 255
1074 BIOS Control Word: 0x18b6
1075 Adapter Control Word: 0x005b
1076 Extended Translation: Enabled
1077 Disconnect Enable Flags: 0xffff
1078 Ultra Enable Flags: 0x0001
1079 Tag Queue Enable Flags: 0x0000
1080 Ordered Queue Tag Flags: 0x0000
1081 Default Tag Queue Depth: 8
1082 Tagged Queue By Device array for aic7xxx host instance 0:
1083 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1084 Actual queue depth per device for aic7xxx host instance 0:
1085 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1088 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1089 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1090 Total transfers 160151 (74577 reads and 85574 writes)
1092 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1093 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1094 Total transfers 0 (0 reads and 0 writes)
1097 1.6 Parallel port info in /proc/parport
1098 ---------------------------------------
1100 The directory /proc/parport contains information about the parallel ports of
1101 your system. It has one subdirectory for each port, named after the port
1104 These directories contain the four files shown in Table 1-10.
1107 Table 1-10: Files in /proc/parport
1108 ..............................................................................
1110 autoprobe Any IEEE-1284 device ID information that has been acquired.
1111 devices list of the device drivers using that port. A + will appear by the
1112 name of the device currently using the port (it might not appear
1114 hardware Parallel port's base address, IRQ line and DMA channel.
1115 irq IRQ that parport is using for that port. This is in a separate
1116 file to allow you to alter it by writing a new value in (IRQ
1118 ..............................................................................
1120 1.7 TTY info in /proc/tty
1121 -------------------------
1123 Information about the available and actually used tty's can be found in the
1124 directory /proc/tty.You'll find entries for drivers and line disciplines in
1125 this directory, as shown in Table 1-11.
1128 Table 1-11: Files in /proc/tty
1129 ..............................................................................
1131 drivers list of drivers and their usage
1132 ldiscs registered line disciplines
1133 driver/serial usage statistic and status of single tty lines
1134 ..............................................................................
1136 To see which tty's are currently in use, you can simply look into the file
1139 > cat /proc/tty/drivers
1140 pty_slave /dev/pts 136 0-255 pty:slave
1141 pty_master /dev/ptm 128 0-255 pty:master
1142 pty_slave /dev/ttyp 3 0-255 pty:slave
1143 pty_master /dev/pty 2 0-255 pty:master
1144 serial /dev/cua 5 64-67 serial:callout
1145 serial /dev/ttyS 4 64-67 serial
1146 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1147 /dev/ptmx /dev/ptmx 5 2 system
1148 /dev/console /dev/console 5 1 system:console
1149 /dev/tty /dev/tty 5 0 system:/dev/tty
1150 unknown /dev/tty 4 1-63 console
1153 1.8 Miscellaneous kernel statistics in /proc/stat
1154 -------------------------------------------------
1156 Various pieces of information about kernel activity are available in the
1157 /proc/stat file. All of the numbers reported in this file are aggregates
1158 since the system first booted. For a quick look, simply cat the file:
1161 cpu 2255 34 2290 22625563 6290 127 456 0 0
1162 cpu0 1132 34 1441 11311718 3675 127 438 0 0
1163 cpu1 1123 0 849 11313845 2614 0 18 0 0
1164 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1170 softirq 183433 0 21755 12 39 1137 231 21459 2263
1172 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1173 lines. These numbers identify the amount of time the CPU has spent performing
1174 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1175 second). The meanings of the columns are as follows, from left to right:
1177 - user: normal processes executing in user mode
1178 - nice: niced processes executing in user mode
1179 - system: processes executing in kernel mode
1180 - idle: twiddling thumbs
1181 - iowait: waiting for I/O to complete
1182 - irq: servicing interrupts
1183 - softirq: servicing softirqs
1184 - steal: involuntary wait
1185 - guest: running a normal guest
1186 - guest_nice: running a niced guest
1188 The "intr" line gives counts of interrupts serviced since boot time, for each
1189 of the possible system interrupts. The first column is the total of all
1190 interrupts serviced; each subsequent column is the total for that particular
1193 The "ctxt" line gives the total number of context switches across all CPUs.
1195 The "btime" line gives the time at which the system booted, in seconds since
1198 The "processes" line gives the number of processes and threads created, which
1199 includes (but is not limited to) those created by calls to the fork() and
1200 clone() system calls.
1202 The "procs_running" line gives the total number of threads that are
1203 running or ready to run (i.e., the total number of runnable threads).
1205 The "procs_blocked" line gives the number of processes currently blocked,
1206 waiting for I/O to complete.
1208 The "softirq" line gives counts of softirqs serviced since boot time, for each
1209 of the possible system softirqs. The first column is the total of all
1210 softirqs serviced; each subsequent column is the total for that particular
1214 1.9 Ext4 file system parameters
1215 ------------------------------
1217 Information about mounted ext4 file systems can be found in
1218 /proc/fs/ext4. Each mounted filesystem will have a directory in
1219 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1220 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1221 in Table 1-12, below.
1223 Table 1-12: Files in /proc/fs/ext4/<devname>
1224 ..............................................................................
1226 mb_groups details of multiblock allocator buddy cache of free blocks
1227 ..............................................................................
1231 Shows registered system console lines.
1233 To see which character device lines are currently used for the system console
1234 /dev/console, you may simply look into the file /proc/consoles:
1236 > cat /proc/consoles
1242 device name of the device
1243 operations R = can do read operations
1244 W = can do write operations
1246 flags E = it is enabled
1247 C = it is preferred console
1248 B = it is primary boot console
1249 p = it is used for printk buffer
1250 b = it is not a TTY but a Braille device
1251 a = it is safe to use when cpu is offline
1252 major:minor major and minor number of the device separated by a colon
1254 ------------------------------------------------------------------------------
1256 ------------------------------------------------------------------------------
1257 The /proc file system serves information about the running system. It not only
1258 allows access to process data but also allows you to request the kernel status
1259 by reading files in the hierarchy.
1261 The directory structure of /proc reflects the types of information and makes
1262 it easy, if not obvious, where to look for specific data.
1263 ------------------------------------------------------------------------------
1265 ------------------------------------------------------------------------------
1266 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1267 ------------------------------------------------------------------------------
1269 ------------------------------------------------------------------------------
1271 ------------------------------------------------------------------------------
1272 * Modifying kernel parameters by writing into files found in /proc/sys
1273 * Exploring the files which modify certain parameters
1274 * Review of the /proc/sys file tree
1275 ------------------------------------------------------------------------------
1278 A very interesting part of /proc is the directory /proc/sys. This is not only
1279 a source of information, it also allows you to change parameters within the
1280 kernel. Be very careful when attempting this. You can optimize your system,
1281 but you can also cause it to crash. Never alter kernel parameters on a
1282 production system. Set up a development machine and test to make sure that
1283 everything works the way you want it to. You may have no alternative but to
1284 reboot the machine once an error has been made.
1286 To change a value, simply echo the new value into the file. An example is
1287 given below in the section on the file system data. You need to be root to do
1288 this. You can create your own boot script to perform this every time your
1291 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1292 general things in the operation of the Linux kernel. Since some of the files
1293 can inadvertently disrupt your system, it is advisable to read both
1294 documentation and source before actually making adjustments. In any case, be
1295 very careful when writing to any of these files. The entries in /proc may
1296 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1297 review the kernel documentation in the directory /usr/src/linux/Documentation.
1298 This chapter is heavily based on the documentation included in the pre 2.2
1299 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1301 Please see: Documentation/sysctl/ directory for descriptions of these
1304 ------------------------------------------------------------------------------
1306 ------------------------------------------------------------------------------
1307 Certain aspects of kernel behavior can be modified at runtime, without the
1308 need to recompile the kernel, or even to reboot the system. The files in the
1309 /proc/sys tree can not only be read, but also modified. You can use the echo
1310 command to write value into these files, thereby changing the default settings
1312 ------------------------------------------------------------------------------
1314 ------------------------------------------------------------------------------
1315 CHAPTER 3: PER-PROCESS PARAMETERS
1316 ------------------------------------------------------------------------------
1318 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1319 --------------------------------------------------------------------------------
1321 These file can be used to adjust the badness heuristic used to select which
1322 process gets killed in out of memory conditions.
1324 The badness heuristic assigns a value to each candidate task ranging from 0
1325 (never kill) to 1000 (always kill) to determine which process is targeted. The
1326 units are roughly a proportion along that range of allowed memory the process
1327 may allocate from based on an estimation of its current memory and swap use.
1328 For example, if a task is using all allowed memory, its badness score will be
1329 1000. If it is using half of its allowed memory, its score will be 500.
1331 There is an additional factor included in the badness score: root
1332 processes are given 3% extra memory over other tasks.
1334 The amount of "allowed" memory depends on the context in which the oom killer
1335 was called. If it is due to the memory assigned to the allocating task's cpuset
1336 being exhausted, the allowed memory represents the set of mems assigned to that
1337 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1338 memory represents the set of mempolicy nodes. If it is due to a memory
1339 limit (or swap limit) being reached, the allowed memory is that configured
1340 limit. Finally, if it is due to the entire system being out of memory, the
1341 allowed memory represents all allocatable resources.
1343 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1344 is used to determine which task to kill. Acceptable values range from -1000
1345 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1346 polarize the preference for oom killing either by always preferring a certain
1347 task or completely disabling it. The lowest possible value, -1000, is
1348 equivalent to disabling oom killing entirely for that task since it will always
1349 report a badness score of 0.
1351 Consequently, it is very simple for userspace to define the amount of memory to
1352 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1353 example, is roughly equivalent to allowing the remainder of tasks sharing the
1354 same system, cpuset, mempolicy, or memory controller resources to use at least
1355 50% more memory. A value of -500, on the other hand, would be roughly
1356 equivalent to discounting 50% of the task's allowed memory from being considered
1357 as scoring against the task.
1359 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1360 be used to tune the badness score. Its acceptable values range from -16
1361 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1362 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1363 scaled linearly with /proc/<pid>/oom_score_adj.
1365 Writing to /proc/<pid>/oom_score_adj or /proc/<pid>/oom_adj will change the
1366 other with its scaled value.
1368 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1369 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1370 requires CAP_SYS_RESOURCE.
1372 NOTICE: /proc/<pid>/oom_adj is deprecated and will be removed, please see
1373 Documentation/feature-removal-schedule.txt.
1375 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1376 generation children with separate address spaces instead, if possible. This
1377 avoids servers and important system daemons from being killed and loses the
1378 minimal amount of work.
1381 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1382 -------------------------------------------------------------
1384 This file can be used to check the current score used by the oom-killer is for
1385 any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
1386 process should be killed in an out-of-memory situation.
1389 3.3 /proc/<pid>/io - Display the IO accounting fields
1390 -------------------------------------------------------
1392 This file contains IO statistics for each running process
1397 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1400 test:/tmp # cat /proc/3828/io
1406 write_bytes: 323932160
1407 cancelled_write_bytes: 0
1416 I/O counter: chars read
1417 The number of bytes which this task has caused to be read from storage. This
1418 is simply the sum of bytes which this process passed to read() and pread().
1419 It includes things like tty IO and it is unaffected by whether or not actual
1420 physical disk IO was required (the read might have been satisfied from
1427 I/O counter: chars written
1428 The number of bytes which this task has caused, or shall cause to be written
1429 to disk. Similar caveats apply here as with rchar.
1435 I/O counter: read syscalls
1436 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1443 I/O counter: write syscalls
1444 Attempt to count the number of write I/O operations, i.e. syscalls like
1445 write() and pwrite().
1451 I/O counter: bytes read
1452 Attempt to count the number of bytes which this process really did cause to
1453 be fetched from the storage layer. Done at the submit_bio() level, so it is
1454 accurate for block-backed filesystems. <please add status regarding NFS and
1455 CIFS at a later time>
1461 I/O counter: bytes written
1462 Attempt to count the number of bytes which this process caused to be sent to
1463 the storage layer. This is done at page-dirtying time.
1466 cancelled_write_bytes
1467 ---------------------
1469 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1470 then deletes the file, it will in fact perform no writeout. But it will have
1471 been accounted as having caused 1MB of write.
1472 In other words: The number of bytes which this process caused to not happen,
1473 by truncating pagecache. A task can cause "negative" IO too. If this task
1474 truncates some dirty pagecache, some IO which another task has been accounted
1475 for (in its write_bytes) will not be happening. We _could_ just subtract that
1476 from the truncating task's write_bytes, but there is information loss in doing
1483 At its current implementation state, this is a bit racy on 32-bit machines: if
1484 process A reads process B's /proc/pid/io while process B is updating one of
1485 those 64-bit counters, process A could see an intermediate result.
1488 More information about this can be found within the taskstats documentation in
1489 Documentation/accounting.
1491 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1492 ---------------------------------------------------------------
1493 When a process is dumped, all anonymous memory is written to a core file as
1494 long as the size of the core file isn't limited. But sometimes we don't want
1495 to dump some memory segments, for example, huge shared memory. Conversely,
1496 sometimes we want to save file-backed memory segments into a core file, not
1497 only the individual files.
1499 /proc/<pid>/coredump_filter allows you to customize which memory segments
1500 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1501 of memory types. If a bit of the bitmask is set, memory segments of the
1502 corresponding memory type are dumped, otherwise they are not dumped.
1504 The following 7 memory types are supported:
1505 - (bit 0) anonymous private memory
1506 - (bit 1) anonymous shared memory
1507 - (bit 2) file-backed private memory
1508 - (bit 3) file-backed shared memory
1509 - (bit 4) ELF header pages in file-backed private memory areas (it is
1510 effective only if the bit 2 is cleared)
1511 - (bit 5) hugetlb private memory
1512 - (bit 6) hugetlb shared memory
1514 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1515 are always dumped regardless of the bitmask status.
1517 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1518 effected by bit 5-6.
1520 Default value of coredump_filter is 0x23; this means all anonymous memory
1521 segments and hugetlb private memory are dumped.
1523 If you don't want to dump all shared memory segments attached to pid 1234,
1524 write 0x21 to the process's proc file.
1526 $ echo 0x21 > /proc/1234/coredump_filter
1528 When a new process is created, the process inherits the bitmask status from its
1529 parent. It is useful to set up coredump_filter before the program runs.
1532 $ echo 0x7 > /proc/self/coredump_filter
1535 3.5 /proc/<pid>/mountinfo - Information about mounts
1536 --------------------------------------------------------
1538 This file contains lines of the form:
1540 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1541 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1543 (1) mount ID: unique identifier of the mount (may be reused after umount)
1544 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1545 (3) major:minor: value of st_dev for files on filesystem
1546 (4) root: root of the mount within the filesystem
1547 (5) mount point: mount point relative to the process's root
1548 (6) mount options: per mount options
1549 (7) optional fields: zero or more fields of the form "tag[:value]"
1550 (8) separator: marks the end of the optional fields
1551 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1552 (10) mount source: filesystem specific information or "none"
1553 (11) super options: per super block options
1555 Parsers should ignore all unrecognised optional fields. Currently the
1556 possible optional fields are:
1558 shared:X mount is shared in peer group X
1559 master:X mount is slave to peer group X
1560 propagate_from:X mount is slave and receives propagation from peer group X (*)
1561 unbindable mount is unbindable
1563 (*) X is the closest dominant peer group under the process's root. If
1564 X is the immediate master of the mount, or if there's no dominant peer
1565 group under the same root, then only the "master:X" field is present
1566 and not the "propagate_from:X" field.
1568 For more information on mount propagation see:
1570 Documentation/filesystems/sharedsubtree.txt
1573 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1574 --------------------------------------------------------
1575 These files provide a method to access a tasks comm value. It also allows for
1576 a task to set its own or one of its thread siblings comm value. The comm value
1577 is limited in size compared to the cmdline value, so writing anything longer
1578 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1582 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1583 -------------------------------------------------------------------------
1584 This file provides a fast way to retrieve first level children pids
1585 of a task pointed by <pid>/<tid> pair. The format is a space separated
1588 Note the "first level" here -- if a child has own children they will
1589 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1590 to obtain the descendants.
1592 Since this interface is intended to be fast and cheap it doesn't
1593 guarantee to provide precise results and some children might be
1594 skipped, especially if they've exited right after we printed their
1595 pids, so one need to either stop or freeze processes being inspected
1596 if precise results are needed.
1599 ------------------------------------------------------------------------------
1601 ------------------------------------------------------------------------------
1604 ---------------------
1606 The following mount options are supported:
1608 hidepid= Set /proc/<pid>/ access mode.
1609 gid= Set the group authorized to learn processes information.
1611 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1614 hidepid=1 means users may not access any /proc/<pid>/ directories but their
1615 own. Sensitive files like cmdline, sched*, status are now protected against
1616 other users. This makes it impossible to learn whether any user runs
1617 specific program (given the program doesn't reveal itself by its behaviour).
1618 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1619 poorly written programs passing sensitive information via program arguments are
1620 now protected against local eavesdroppers.
1622 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1623 users. It doesn't mean that it hides a fact whether a process with a specific
1624 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1625 but it hides process' uid and gid, which may be learned by stat()'ing
1626 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1627 information about running processes, whether some daemon runs with elevated
1628 privileges, whether other user runs some sensitive program, whether other users
1629 run any program at all, etc.
1631 gid= defines a group authorized to learn processes information otherwise
1632 prohibited by hidepid=. If you use some daemon like identd which needs to learn
1633 information about processes information, just add identd to this group.