Loading...
1------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3------------------------------------------------------------------------------
4/proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
6
72.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
9------------------------------------------------------------------------------
10Version 1.3 Kernel version 2.2.12
11 Kernel version 2.4.0-test11-pre4
12------------------------------------------------------------------------------
13fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
14
15Table of Contents
16-----------------
17
18 0 Preface
19 0.1 Introduction/Credits
20 0.2 Legal Stuff
21
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
32
33 2 Modifying System Parameters
34
35 3 Per-Process Parameters
36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
37 score
38 3.2 /proc/<pid>/oom_score - Display current oom-killer score
39 3.3 /proc/<pid>/io - Display the IO accounting fields
40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
41 3.5 /proc/<pid>/mountinfo - Information about mounts
42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
43 3.7 /proc/<pid>/task/<tid>/children - Information about task children
44 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
45
46 4 Configuring procfs
47 4.1 Mount options
48
49------------------------------------------------------------------------------
50Preface
51------------------------------------------------------------------------------
52
530.1 Introduction/Credits
54------------------------
55
56This documentation is part of a soon (or so we hope) to be released book on
57the SuSE Linux distribution. As there is no complete documentation for the
58/proc file system and we've used many freely available sources to write these
59chapters, it seems only fair to give the work back to the Linux community.
60This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
61afraid it's still far from complete, but we hope it will be useful. As far as
62we know, it is the first 'all-in-one' document about the /proc file system. It
63is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
64SPARC, AXP, etc., features, you probably won't find what you are looking for.
65It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
66additions and patches are welcome and will be added to this document if you
67mail them to Bodo.
68
69We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
70other people for help compiling this documentation. We'd also like to extend a
71special thank you to Andi Kleen for documentation, which we relied on heavily
72to create this document, as well as the additional information he provided.
73Thanks to everybody else who contributed source or docs to the Linux kernel
74and helped create a great piece of software... :)
75
76If you have any comments, corrections or additions, please don't hesitate to
77contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
78document.
79
80The latest version of this document is available online at
81http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
82
83If the above direction does not works for you, you could try the kernel
84mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
85comandante@zaralinux.com.
86
870.2 Legal Stuff
88---------------
89
90We don't guarantee the correctness of this document, and if you come to us
91complaining about how you screwed up your system because of incorrect
92documentation, we won't feel responsible...
93
94------------------------------------------------------------------------------
95CHAPTER 1: COLLECTING SYSTEM INFORMATION
96------------------------------------------------------------------------------
97
98------------------------------------------------------------------------------
99In This Chapter
100------------------------------------------------------------------------------
101* Investigating the properties of the pseudo file system /proc and its
102 ability to provide information on the running Linux system
103* Examining /proc's structure
104* Uncovering various information about the kernel and the processes running
105 on the system
106------------------------------------------------------------------------------
107
108
109The proc file system acts as an interface to internal data structures in the
110kernel. It can be used to obtain information about the system and to change
111certain kernel parameters at runtime (sysctl).
112
113First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
114show you how you can use /proc/sys to change settings.
115
1161.1 Process-Specific Subdirectories
117-----------------------------------
118
119The directory /proc contains (among other things) one subdirectory for each
120process running on the system, which is named after the process ID (PID).
121
122The link self points to the process reading the file system. Each process
123subdirectory has the entries listed in Table 1-1.
124
125
126Table 1-1: Process specific entries in /proc
127..............................................................................
128 File Content
129 clear_refs Clears page referenced bits shown in smaps output
130 cmdline Command line arguments
131 cpu Current and last cpu in which it was executed (2.4)(smp)
132 cwd Link to the current working directory
133 environ Values of environment variables
134 exe Link to the executable of this process
135 fd Directory, which contains all file descriptors
136 maps Memory maps to executables and library files (2.4)
137 mem Memory held by this process
138 root Link to the root directory of this process
139 stat Process status
140 statm Process memory status information
141 status Process status in human readable form
142 wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan
143 pagemap Page table
144 stack Report full stack trace, enable via CONFIG_STACKTRACE
145 smaps a extension based on maps, showing the memory consumption of
146 each mapping and flags associated with it
147..............................................................................
148
149For example, to get the status information of a process, all you have to do is
150read the file /proc/PID/status:
151
152 >cat /proc/self/status
153 Name: cat
154 State: R (running)
155 Tgid: 5452
156 Pid: 5452
157 PPid: 743
158 TracerPid: 0 (2.4)
159 Uid: 501 501 501 501
160 Gid: 100 100 100 100
161 FDSize: 256
162 Groups: 100 14 16
163 VmPeak: 5004 kB
164 VmSize: 5004 kB
165 VmLck: 0 kB
166 VmHWM: 476 kB
167 VmRSS: 476 kB
168 VmData: 156 kB
169 VmStk: 88 kB
170 VmExe: 68 kB
171 VmLib: 1412 kB
172 VmPTE: 20 kb
173 VmSwap: 0 kB
174 Threads: 1
175 SigQ: 0/28578
176 SigPnd: 0000000000000000
177 ShdPnd: 0000000000000000
178 SigBlk: 0000000000000000
179 SigIgn: 0000000000000000
180 SigCgt: 0000000000000000
181 CapInh: 00000000fffffeff
182 CapPrm: 0000000000000000
183 CapEff: 0000000000000000
184 CapBnd: ffffffffffffffff
185 Seccomp: 0
186 voluntary_ctxt_switches: 0
187 nonvoluntary_ctxt_switches: 1
188
189This shows you nearly the same information you would get if you viewed it with
190the ps command. In fact, ps uses the proc file system to obtain its
191information. But you get a more detailed view of the process by reading the
192file /proc/PID/status. It fields are described in table 1-2.
193
194The statm file contains more detailed information about the process
195memory usage. Its seven fields are explained in Table 1-3. The stat file
196contains details information about the process itself. Its fields are
197explained in Table 1-4.
198
199(for SMP CONFIG users)
200For making accounting scalable, RSS related information are handled in
201asynchronous manner and the vaule may not be very precise. To see a precise
202snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
203It's slow but very precise.
204
205Table 1-2: Contents of the status files (as of 2.6.30-rc7)
206..............................................................................
207 Field Content
208 Name filename of the executable
209 State state (R is running, S is sleeping, D is sleeping
210 in an uninterruptible wait, Z is zombie,
211 T is traced or stopped)
212 Tgid thread group ID
213 Pid process id
214 PPid process id of the parent process
215 TracerPid PID of process tracing this process (0 if not)
216 Uid Real, effective, saved set, and file system UIDs
217 Gid Real, effective, saved set, and file system GIDs
218 FDSize number of file descriptor slots currently allocated
219 Groups supplementary group list
220 VmPeak peak virtual memory size
221 VmSize total program size
222 VmLck locked memory size
223 VmHWM peak resident set size ("high water mark")
224 VmRSS size of memory portions
225 VmData size of data, stack, and text segments
226 VmStk size of data, stack, and text segments
227 VmExe size of text segment
228 VmLib size of shared library code
229 VmPTE size of page table entries
230 VmSwap size of swap usage (the number of referred swapents)
231 Threads number of threads
232 SigQ number of signals queued/max. number for queue
233 SigPnd bitmap of pending signals for the thread
234 ShdPnd bitmap of shared pending signals for the process
235 SigBlk bitmap of blocked signals
236 SigIgn bitmap of ignored signals
237 SigCgt bitmap of catched signals
238 CapInh bitmap of inheritable capabilities
239 CapPrm bitmap of permitted capabilities
240 CapEff bitmap of effective capabilities
241 CapBnd bitmap of capabilities bounding set
242 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
243 Cpus_allowed mask of CPUs on which this process may run
244 Cpus_allowed_list Same as previous, but in "list format"
245 Mems_allowed mask of memory nodes allowed to this process
246 Mems_allowed_list Same as previous, but in "list format"
247 voluntary_ctxt_switches number of voluntary context switches
248 nonvoluntary_ctxt_switches number of non voluntary context switches
249..............................................................................
250
251Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
252..............................................................................
253 Field Content
254 size total program size (pages) (same as VmSize in status)
255 resident size of memory portions (pages) (same as VmRSS in status)
256 shared number of pages that are shared (i.e. backed by a file)
257 trs number of pages that are 'code' (not including libs; broken,
258 includes data segment)
259 lrs number of pages of library (always 0 on 2.6)
260 drs number of pages of data/stack (including libs; broken,
261 includes library text)
262 dt number of dirty pages (always 0 on 2.6)
263..............................................................................
264
265
266Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
267..............................................................................
268 Field Content
269 pid process id
270 tcomm filename of the executable
271 state state (R is running, S is sleeping, D is sleeping in an
272 uninterruptible wait, Z is zombie, T is traced or stopped)
273 ppid process id of the parent process
274 pgrp pgrp of the process
275 sid session id
276 tty_nr tty the process uses
277 tty_pgrp pgrp of the tty
278 flags task flags
279 min_flt number of minor faults
280 cmin_flt number of minor faults with child's
281 maj_flt number of major faults
282 cmaj_flt number of major faults with child's
283 utime user mode jiffies
284 stime kernel mode jiffies
285 cutime user mode jiffies with child's
286 cstime kernel mode jiffies with child's
287 priority priority level
288 nice nice level
289 num_threads number of threads
290 it_real_value (obsolete, always 0)
291 start_time time the process started after system boot
292 vsize virtual memory size
293 rss resident set memory size
294 rsslim current limit in bytes on the rss
295 start_code address above which program text can run
296 end_code address below which program text can run
297 start_stack address of the start of the main process stack
298 esp current value of ESP
299 eip current value of EIP
300 pending bitmap of pending signals
301 blocked bitmap of blocked signals
302 sigign bitmap of ignored signals
303 sigcatch bitmap of catched signals
304 wchan address where process went to sleep
305 0 (place holder)
306 0 (place holder)
307 exit_signal signal to send to parent thread on exit
308 task_cpu which CPU the task is scheduled on
309 rt_priority realtime priority
310 policy scheduling policy (man sched_setscheduler)
311 blkio_ticks time spent waiting for block IO
312 gtime guest time of the task in jiffies
313 cgtime guest time of the task children in jiffies
314 start_data address above which program data+bss is placed
315 end_data address below which program data+bss is placed
316 start_brk address above which program heap can be expanded with brk()
317 arg_start address above which program command line is placed
318 arg_end address below which program command line is placed
319 env_start address above which program environment is placed
320 env_end address below which program environment is placed
321 exit_code the thread's exit_code in the form reported by the waitpid system call
322..............................................................................
323
324The /proc/PID/maps file containing the currently mapped memory regions and
325their access permissions.
326
327The format is:
328
329address perms offset dev inode pathname
330
33108048000-08049000 r-xp 00000000 03:00 8312 /opt/test
33208049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
3330804a000-0806b000 rw-p 00000000 00:00 0 [heap]
334a7cb1000-a7cb2000 ---p 00000000 00:00 0
335a7cb2000-a7eb2000 rw-p 00000000 00:00 0
336a7eb2000-a7eb3000 ---p 00000000 00:00 0
337a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
338a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
339a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
340a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
341a800b000-a800e000 rw-p 00000000 00:00 0
342a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
343a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
344a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
345a8024000-a8027000 rw-p 00000000 00:00 0
346a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
347a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
348a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
349aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
350ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
351
352where "address" is the address space in the process that it occupies, "perms"
353is a set of permissions:
354
355 r = read
356 w = write
357 x = execute
358 s = shared
359 p = private (copy on write)
360
361"offset" is the offset into the mapping, "dev" is the device (major:minor), and
362"inode" is the inode on that device. 0 indicates that no inode is associated
363with the memory region, as the case would be with BSS (uninitialized data).
364The "pathname" shows the name associated file for this mapping. If the mapping
365is not associated with a file:
366
367 [heap] = the heap of the program
368 [stack] = the stack of the main process
369 [stack:1001] = the stack of the thread with tid 1001
370 [vdso] = the "virtual dynamic shared object",
371 the kernel system call handler
372
373 or if empty, the mapping is anonymous.
374
375The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
376of the individual tasks of a process. In this file you will see a mapping marked
377as [stack] if that task sees it as a stack. This is a key difference from the
378content of /proc/PID/maps, where you will see all mappings that are being used
379as stack by all of those tasks. Hence, for the example above, the task-level
380map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
381
38208048000-08049000 r-xp 00000000 03:00 8312 /opt/test
38308049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
3840804a000-0806b000 rw-p 00000000 00:00 0 [heap]
385a7cb1000-a7cb2000 ---p 00000000 00:00 0
386a7cb2000-a7eb2000 rw-p 00000000 00:00 0
387a7eb2000-a7eb3000 ---p 00000000 00:00 0
388a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
389a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
390a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
391a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
392a800b000-a800e000 rw-p 00000000 00:00 0
393a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
394a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
395a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
396a8024000-a8027000 rw-p 00000000 00:00 0
397a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
398a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
399a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
400aff35000-aff4a000 rw-p 00000000 00:00 0
401ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
402
403The /proc/PID/smaps is an extension based on maps, showing the memory
404consumption for each of the process's mappings. For each of mappings there
405is a series of lines such as the following:
406
40708048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
408Size: 1084 kB
409Rss: 892 kB
410Pss: 374 kB
411Shared_Clean: 892 kB
412Shared_Dirty: 0 kB
413Private_Clean: 0 kB
414Private_Dirty: 0 kB
415Referenced: 892 kB
416Anonymous: 0 kB
417Swap: 0 kB
418KernelPageSize: 4 kB
419MMUPageSize: 4 kB
420Locked: 374 kB
421VmFlags: rd ex mr mw me de
422
423the first of these lines shows the same information as is displayed for the
424mapping in /proc/PID/maps. The remaining lines show the size of the mapping
425(size), the amount of the mapping that is currently resident in RAM (RSS), the
426process' proportional share of this mapping (PSS), the number of clean and
427dirty private pages in the mapping. Note that even a page which is part of a
428MAP_SHARED mapping, but has only a single pte mapped, i.e. is currently used
429by only one process, is accounted as private and not as shared. "Referenced"
430indicates the amount of memory currently marked as referenced or accessed.
431"Anonymous" shows the amount of memory that does not belong to any file. Even
432a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
433and a page is modified, the file page is replaced by a private anonymous copy.
434"Swap" shows how much would-be-anonymous memory is also used, but out on
435swap.
436
437"VmFlags" field deserves a separate description. This member represents the kernel
438flags associated with the particular virtual memory area in two letter encoded
439manner. The codes are the following:
440 rd - readable
441 wr - writeable
442 ex - executable
443 sh - shared
444 mr - may read
445 mw - may write
446 me - may execute
447 ms - may share
448 gd - stack segment growns down
449 pf - pure PFN range
450 dw - disabled write to the mapped file
451 lo - pages are locked in memory
452 io - memory mapped I/O area
453 sr - sequential read advise provided
454 rr - random read advise provided
455 dc - do not copy area on fork
456 de - do not expand area on remapping
457 ac - area is accountable
458 nr - swap space is not reserved for the area
459 ht - area uses huge tlb pages
460 nl - non-linear mapping
461 ar - architecture specific flag
462 dd - do not include area into core dump
463 sd - soft-dirty flag
464 mm - mixed map area
465 hg - huge page advise flag
466 nh - no-huge page advise flag
467 mg - mergable advise flag
468
469Note that there is no guarantee that every flag and associated mnemonic will
470be present in all further kernel releases. Things get changed, the flags may
471be vanished or the reverse -- new added.
472
473This file is only present if the CONFIG_MMU kernel configuration option is
474enabled.
475
476The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
477bits on both physical and virtual pages associated with a process, and the
478soft-dirty bit on pte (see Documentation/vm/soft-dirty.txt for details).
479To clear the bits for all the pages associated with the process
480 > echo 1 > /proc/PID/clear_refs
481
482To clear the bits for the anonymous pages associated with the process
483 > echo 2 > /proc/PID/clear_refs
484
485To clear the bits for the file mapped pages associated with the process
486 > echo 3 > /proc/PID/clear_refs
487
488To clear the soft-dirty bit
489 > echo 4 > /proc/PID/clear_refs
490
491Any other value written to /proc/PID/clear_refs will have no effect.
492
493The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
494using /proc/kpageflags and number of times a page is mapped using
495/proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
496
4971.2 Kernel data
498---------------
499
500Similar to the process entries, the kernel data files give information about
501the running kernel. The files used to obtain this information are contained in
502/proc and are listed in Table 1-5. Not all of these will be present in your
503system. It depends on the kernel configuration and the loaded modules, which
504files are there, and which are missing.
505
506Table 1-5: Kernel info in /proc
507..............................................................................
508 File Content
509 apm Advanced power management info
510 buddyinfo Kernel memory allocator information (see text) (2.5)
511 bus Directory containing bus specific information
512 cmdline Kernel command line
513 cpuinfo Info about the CPU
514 devices Available devices (block and character)
515 dma Used DMS channels
516 filesystems Supported filesystems
517 driver Various drivers grouped here, currently rtc (2.4)
518 execdomains Execdomains, related to security (2.4)
519 fb Frame Buffer devices (2.4)
520 fs File system parameters, currently nfs/exports (2.4)
521 ide Directory containing info about the IDE subsystem
522 interrupts Interrupt usage
523 iomem Memory map (2.4)
524 ioports I/O port usage
525 irq Masks for irq to cpu affinity (2.4)(smp?)
526 isapnp ISA PnP (Plug&Play) Info (2.4)
527 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
528 kmsg Kernel messages
529 ksyms Kernel symbol table
530 loadavg Load average of last 1, 5 & 15 minutes
531 locks Kernel locks
532 meminfo Memory info
533 misc Miscellaneous
534 modules List of loaded modules
535 mounts Mounted filesystems
536 net Networking info (see text)
537 pagetypeinfo Additional page allocator information (see text) (2.5)
538 partitions Table of partitions known to the system
539 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
540 decoupled by lspci (2.4)
541 rtc Real time clock
542 scsi SCSI info (see text)
543 slabinfo Slab pool info
544 softirqs softirq usage
545 stat Overall statistics
546 swaps Swap space utilization
547 sys See chapter 2
548 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
549 tty Info of tty drivers
550 uptime Wall clock since boot, combined idle time of all cpus
551 version Kernel version
552 video bttv info of video resources (2.4)
553 vmallocinfo Show vmalloced areas
554..............................................................................
555
556You can, for example, check which interrupts are currently in use and what
557they are used for by looking in the file /proc/interrupts:
558
559 > cat /proc/interrupts
560 CPU0
561 0: 8728810 XT-PIC timer
562 1: 895 XT-PIC keyboard
563 2: 0 XT-PIC cascade
564 3: 531695 XT-PIC aha152x
565 4: 2014133 XT-PIC serial
566 5: 44401 XT-PIC pcnet_cs
567 8: 2 XT-PIC rtc
568 11: 8 XT-PIC i82365
569 12: 182918 XT-PIC PS/2 Mouse
570 13: 1 XT-PIC fpu
571 14: 1232265 XT-PIC ide0
572 15: 7 XT-PIC ide1
573 NMI: 0
574
575In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
576output of a SMP machine):
577
578 > cat /proc/interrupts
579
580 CPU0 CPU1
581 0: 1243498 1214548 IO-APIC-edge timer
582 1: 8949 8958 IO-APIC-edge keyboard
583 2: 0 0 XT-PIC cascade
584 5: 11286 10161 IO-APIC-edge soundblaster
585 8: 1 0 IO-APIC-edge rtc
586 9: 27422 27407 IO-APIC-edge 3c503
587 12: 113645 113873 IO-APIC-edge PS/2 Mouse
588 13: 0 0 XT-PIC fpu
589 14: 22491 24012 IO-APIC-edge ide0
590 15: 2183 2415 IO-APIC-edge ide1
591 17: 30564 30414 IO-APIC-level eth0
592 18: 177 164 IO-APIC-level bttv
593 NMI: 2457961 2457959
594 LOC: 2457882 2457881
595 ERR: 2155
596
597NMI is incremented in this case because every timer interrupt generates a NMI
598(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
599
600LOC is the local interrupt counter of the internal APIC of every CPU.
601
602ERR is incremented in the case of errors in the IO-APIC bus (the bus that
603connects the CPUs in a SMP system. This means that an error has been detected,
604the IO-APIC automatically retry the transmission, so it should not be a big
605problem, but you should read the SMP-FAQ.
606
607In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
608/proc/interrupts to display every IRQ vector in use by the system, not
609just those considered 'most important'. The new vectors are:
610
611 THR -- interrupt raised when a machine check threshold counter
612 (typically counting ECC corrected errors of memory or cache) exceeds
613 a configurable threshold. Only available on some systems.
614
615 TRM -- a thermal event interrupt occurs when a temperature threshold
616 has been exceeded for the CPU. This interrupt may also be generated
617 when the temperature drops back to normal.
618
619 SPU -- a spurious interrupt is some interrupt that was raised then lowered
620 by some IO device before it could be fully processed by the APIC. Hence
621 the APIC sees the interrupt but does not know what device it came from.
622 For this case the APIC will generate the interrupt with a IRQ vector
623 of 0xff. This might also be generated by chipset bugs.
624
625 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
626 sent from one CPU to another per the needs of the OS. Typically,
627 their statistics are used by kernel developers and interested users to
628 determine the occurrence of interrupts of the given type.
629
630The above IRQ vectors are displayed only when relevant. For example,
631the threshold vector does not exist on x86_64 platforms. Others are
632suppressed when the system is a uniprocessor. As of this writing, only
633i386 and x86_64 platforms support the new IRQ vector displays.
634
635Of some interest is the introduction of the /proc/irq directory to 2.4.
636It could be used to set IRQ to CPU affinity, this means that you can "hook" an
637IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
638irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
639prof_cpu_mask.
640
641For example
642 > ls /proc/irq/
643 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
644 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
645 > ls /proc/irq/0/
646 smp_affinity
647
648smp_affinity is a bitmask, in which you can specify which CPUs can handle the
649IRQ, you can set it by doing:
650
651 > echo 1 > /proc/irq/10/smp_affinity
652
653This means that only the first CPU will handle the IRQ, but you can also echo
6545 which means that only the first and fourth CPU can handle the IRQ.
655
656The contents of each smp_affinity file is the same by default:
657
658 > cat /proc/irq/0/smp_affinity
659 ffffffff
660
661There is an alternate interface, smp_affinity_list which allows specifying
662a cpu range instead of a bitmask:
663
664 > cat /proc/irq/0/smp_affinity_list
665 1024-1031
666
667The default_smp_affinity mask applies to all non-active IRQs, which are the
668IRQs which have not yet been allocated/activated, and hence which lack a
669/proc/irq/[0-9]* directory.
670
671The node file on an SMP system shows the node to which the device using the IRQ
672reports itself as being attached. This hardware locality information does not
673include information about any possible driver locality preference.
674
675prof_cpu_mask specifies which CPUs are to be profiled by the system wide
676profiler. Default value is ffffffff (all cpus if there are only 32 of them).
677
678The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
679between all the CPUs which are allowed to handle it. As usual the kernel has
680more info than you and does a better job than you, so the defaults are the
681best choice for almost everyone. [Note this applies only to those IO-APIC's
682that support "Round Robin" interrupt distribution.]
683
684There are three more important subdirectories in /proc: net, scsi, and sys.
685The general rule is that the contents, or even the existence of these
686directories, depend on your kernel configuration. If SCSI is not enabled, the
687directory scsi may not exist. The same is true with the net, which is there
688only when networking support is present in the running kernel.
689
690The slabinfo file gives information about memory usage at the slab level.
691Linux uses slab pools for memory management above page level in version 2.2.
692Commonly used objects have their own slab pool (such as network buffers,
693directory cache, and so on).
694
695..............................................................................
696
697> cat /proc/buddyinfo
698
699Node 0, zone DMA 0 4 5 4 4 3 ...
700Node 0, zone Normal 1 0 0 1 101 8 ...
701Node 0, zone HighMem 2 0 0 1 1 0 ...
702
703External fragmentation is a problem under some workloads, and buddyinfo is a
704useful tool for helping diagnose these problems. Buddyinfo will give you a
705clue as to how big an area you can safely allocate, or why a previous
706allocation failed.
707
708Each column represents the number of pages of a certain order which are
709available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
710ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
711available in ZONE_NORMAL, etc...
712
713More information relevant to external fragmentation can be found in
714pagetypeinfo.
715
716> cat /proc/pagetypeinfo
717Page block order: 9
718Pages per block: 512
719
720Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
721Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
722Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
723Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
724Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
725Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
726Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
727Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
728Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
729Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
730Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
731
732Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
733Node 0, zone DMA 2 0 5 1 0
734Node 0, zone DMA32 41 6 967 2 0
735
736Fragmentation avoidance in the kernel works by grouping pages of different
737migrate types into the same contiguous regions of memory called page blocks.
738A page block is typically the size of the default hugepage size e.g. 2MB on
739X86-64. By keeping pages grouped based on their ability to move, the kernel
740can reclaim pages within a page block to satisfy a high-order allocation.
741
742The pagetypinfo begins with information on the size of a page block. It
743then gives the same type of information as buddyinfo except broken down
744by migrate-type and finishes with details on how many page blocks of each
745type exist.
746
747If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
748from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
749make an estimate of the likely number of huge pages that can be allocated
750at a given point in time. All the "Movable" blocks should be allocatable
751unless memory has been mlock()'d. Some of the Reclaimable blocks should
752also be allocatable although a lot of filesystem metadata may have to be
753reclaimed to achieve this.
754
755..............................................................................
756
757meminfo:
758
759Provides information about distribution and utilization of memory. This
760varies by architecture and compile options. The following is from a
76116GB PIII, which has highmem enabled. You may not have all of these fields.
762
763> cat /proc/meminfo
764
765The "Locked" indicates whether the mapping is locked in memory or not.
766
767
768MemTotal: 16344972 kB
769MemFree: 13634064 kB
770MemAvailable: 14836172 kB
771Buffers: 3656 kB
772Cached: 1195708 kB
773SwapCached: 0 kB
774Active: 891636 kB
775Inactive: 1077224 kB
776HighTotal: 15597528 kB
777HighFree: 13629632 kB
778LowTotal: 747444 kB
779LowFree: 4432 kB
780SwapTotal: 0 kB
781SwapFree: 0 kB
782Dirty: 968 kB
783Writeback: 0 kB
784AnonPages: 861800 kB
785Mapped: 280372 kB
786Slab: 284364 kB
787SReclaimable: 159856 kB
788SUnreclaim: 124508 kB
789PageTables: 24448 kB
790NFS_Unstable: 0 kB
791Bounce: 0 kB
792WritebackTmp: 0 kB
793CommitLimit: 7669796 kB
794Committed_AS: 100056 kB
795VmallocTotal: 112216 kB
796VmallocUsed: 428 kB
797VmallocChunk: 111088 kB
798AnonHugePages: 49152 kB
799
800 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
801 bits and the kernel binary code)
802 MemFree: The sum of LowFree+HighFree
803MemAvailable: An estimate of how much memory is available for starting new
804 applications, without swapping. Calculated from MemFree,
805 SReclaimable, the size of the file LRU lists, and the low
806 watermarks in each zone.
807 The estimate takes into account that the system needs some
808 page cache to function well, and that not all reclaimable
809 slab will be reclaimable, due to items being in use. The
810 impact of those factors will vary from system to system.
811 Buffers: Relatively temporary storage for raw disk blocks
812 shouldn't get tremendously large (20MB or so)
813 Cached: in-memory cache for files read from the disk (the
814 pagecache). Doesn't include SwapCached
815 SwapCached: Memory that once was swapped out, is swapped back in but
816 still also is in the swapfile (if memory is needed it
817 doesn't need to be swapped out AGAIN because it is already
818 in the swapfile. This saves I/O)
819 Active: Memory that has been used more recently and usually not
820 reclaimed unless absolutely necessary.
821 Inactive: Memory which has been less recently used. It is more
822 eligible to be reclaimed for other purposes
823 HighTotal:
824 HighFree: Highmem is all memory above ~860MB of physical memory
825 Highmem areas are for use by userspace programs, or
826 for the pagecache. The kernel must use tricks to access
827 this memory, making it slower to access than lowmem.
828 LowTotal:
829 LowFree: Lowmem is memory which can be used for everything that
830 highmem can be used for, but it is also available for the
831 kernel's use for its own data structures. Among many
832 other things, it is where everything from the Slab is
833 allocated. Bad things happen when you're out of lowmem.
834 SwapTotal: total amount of swap space available
835 SwapFree: Memory which has been evicted from RAM, and is temporarily
836 on the disk
837 Dirty: Memory which is waiting to get written back to the disk
838 Writeback: Memory which is actively being written back to the disk
839 AnonPages: Non-file backed pages mapped into userspace page tables
840AnonHugePages: Non-file backed huge pages mapped into userspace page tables
841 Mapped: files which have been mmaped, such as libraries
842 Slab: in-kernel data structures cache
843SReclaimable: Part of Slab, that might be reclaimed, such as caches
844 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
845 PageTables: amount of memory dedicated to the lowest level of page
846 tables.
847NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
848 storage
849 Bounce: Memory used for block device "bounce buffers"
850WritebackTmp: Memory used by FUSE for temporary writeback buffers
851 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
852 this is the total amount of memory currently available to
853 be allocated on the system. This limit is only adhered to
854 if strict overcommit accounting is enabled (mode 2 in
855 'vm.overcommit_memory').
856 The CommitLimit is calculated with the following formula:
857 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
858 For example, on a system with 1G of physical RAM and 7G
859 of swap with a `vm.overcommit_ratio` of 30 it would
860 yield a CommitLimit of 7.3G.
861 For more details, see the memory overcommit documentation
862 in vm/overcommit-accounting.
863Committed_AS: The amount of memory presently allocated on the system.
864 The committed memory is a sum of all of the memory which
865 has been allocated by processes, even if it has not been
866 "used" by them as of yet. A process which malloc()'s 1G
867 of memory, but only touches 300M of it will show up as
868 using 1G. This 1G is memory which has been "committed" to
869 by the VM and can be used at any time by the allocating
870 application. With strict overcommit enabled on the system
871 (mode 2 in 'vm.overcommit_memory'),allocations which would
872 exceed the CommitLimit (detailed above) will not be permitted.
873 This is useful if one needs to guarantee that processes will
874 not fail due to lack of memory once that memory has been
875 successfully allocated.
876VmallocTotal: total size of vmalloc memory area
877 VmallocUsed: amount of vmalloc area which is used
878VmallocChunk: largest contiguous block of vmalloc area which is free
879
880..............................................................................
881
882vmallocinfo:
883
884Provides information about vmalloced/vmaped areas. One line per area,
885containing the virtual address range of the area, size in bytes,
886caller information of the creator, and optional information depending
887on the kind of area :
888
889 pages=nr number of pages
890 phys=addr if a physical address was specified
891 ioremap I/O mapping (ioremap() and friends)
892 vmalloc vmalloc() area
893 vmap vmap()ed pages
894 user VM_USERMAP area
895 vpages buffer for pages pointers was vmalloced (huge area)
896 N<node>=nr (Only on NUMA kernels)
897 Number of pages allocated on memory node <node>
898
899> cat /proc/vmallocinfo
9000xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
901 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
9020xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
903 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
9040xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
905 phys=7fee8000 ioremap
9060xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
907 phys=7fee7000 ioremap
9080xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
9090xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
910 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
9110xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
912 pages=2 vmalloc N1=2
9130xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
914 /0x130 [x_tables] pages=4 vmalloc N0=4
9150xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
916 pages=14 vmalloc N2=14
9170xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
918 pages=4 vmalloc N1=4
9190xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
920 pages=2 vmalloc N1=2
9210xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
922 pages=10 vmalloc N0=10
923
924..............................................................................
925
926softirqs:
927
928Provides counts of softirq handlers serviced since boot time, for each cpu.
929
930> cat /proc/softirqs
931 CPU0 CPU1 CPU2 CPU3
932 HI: 0 0 0 0
933 TIMER: 27166 27120 27097 27034
934 NET_TX: 0 0 0 17
935 NET_RX: 42 0 0 39
936 BLOCK: 0 0 107 1121
937 TASKLET: 0 0 0 290
938 SCHED: 27035 26983 26971 26746
939 HRTIMER: 0 0 0 0
940 RCU: 1678 1769 2178 2250
941
942
9431.3 IDE devices in /proc/ide
944----------------------------
945
946The subdirectory /proc/ide contains information about all IDE devices of which
947the kernel is aware. There is one subdirectory for each IDE controller, the
948file drivers and a link for each IDE device, pointing to the device directory
949in the controller specific subtree.
950
951The file drivers contains general information about the drivers used for the
952IDE devices:
953
954 > cat /proc/ide/drivers
955 ide-cdrom version 4.53
956 ide-disk version 1.08
957
958More detailed information can be found in the controller specific
959subdirectories. These are named ide0, ide1 and so on. Each of these
960directories contains the files shown in table 1-6.
961
962
963Table 1-6: IDE controller info in /proc/ide/ide?
964..............................................................................
965 File Content
966 channel IDE channel (0 or 1)
967 config Configuration (only for PCI/IDE bridge)
968 mate Mate name
969 model Type/Chipset of IDE controller
970..............................................................................
971
972Each device connected to a controller has a separate subdirectory in the
973controllers directory. The files listed in table 1-7 are contained in these
974directories.
975
976
977Table 1-7: IDE device information
978..............................................................................
979 File Content
980 cache The cache
981 capacity Capacity of the medium (in 512Byte blocks)
982 driver driver and version
983 geometry physical and logical geometry
984 identify device identify block
985 media media type
986 model device identifier
987 settings device setup
988 smart_thresholds IDE disk management thresholds
989 smart_values IDE disk management values
990..............................................................................
991
992The most interesting file is settings. This file contains a nice overview of
993the drive parameters:
994
995 # cat /proc/ide/ide0/hda/settings
996 name value min max mode
997 ---- ----- --- --- ----
998 bios_cyl 526 0 65535 rw
999 bios_head 255 0 255 rw
1000 bios_sect 63 0 63 rw
1001 breada_readahead 4 0 127 rw
1002 bswap 0 0 1 r
1003 file_readahead 72 0 2097151 rw
1004 io_32bit 0 0 3 rw
1005 keepsettings 0 0 1 rw
1006 max_kb_per_request 122 1 127 rw
1007 multcount 0 0 8 rw
1008 nice1 1 0 1 rw
1009 nowerr 0 0 1 rw
1010 pio_mode write-only 0 255 w
1011 slow 0 0 1 rw
1012 unmaskirq 0 0 1 rw
1013 using_dma 0 0 1 rw
1014
1015
10161.4 Networking info in /proc/net
1017--------------------------------
1018
1019The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1020additional values you get for IP version 6 if you configure the kernel to
1021support this. Table 1-9 lists the files and their meaning.
1022
1023
1024Table 1-8: IPv6 info in /proc/net
1025..............................................................................
1026 File Content
1027 udp6 UDP sockets (IPv6)
1028 tcp6 TCP sockets (IPv6)
1029 raw6 Raw device statistics (IPv6)
1030 igmp6 IP multicast addresses, which this host joined (IPv6)
1031 if_inet6 List of IPv6 interface addresses
1032 ipv6_route Kernel routing table for IPv6
1033 rt6_stats Global IPv6 routing tables statistics
1034 sockstat6 Socket statistics (IPv6)
1035 snmp6 Snmp data (IPv6)
1036..............................................................................
1037
1038
1039Table 1-9: Network info in /proc/net
1040..............................................................................
1041 File Content
1042 arp Kernel ARP table
1043 dev network devices with statistics
1044 dev_mcast the Layer2 multicast groups a device is listening too
1045 (interface index, label, number of references, number of bound
1046 addresses).
1047 dev_stat network device status
1048 ip_fwchains Firewall chain linkage
1049 ip_fwnames Firewall chain names
1050 ip_masq Directory containing the masquerading tables
1051 ip_masquerade Major masquerading table
1052 netstat Network statistics
1053 raw raw device statistics
1054 route Kernel routing table
1055 rpc Directory containing rpc info
1056 rt_cache Routing cache
1057 snmp SNMP data
1058 sockstat Socket statistics
1059 tcp TCP sockets
1060 udp UDP sockets
1061 unix UNIX domain sockets
1062 wireless Wireless interface data (Wavelan etc)
1063 igmp IP multicast addresses, which this host joined
1064 psched Global packet scheduler parameters.
1065 netlink List of PF_NETLINK sockets
1066 ip_mr_vifs List of multicast virtual interfaces
1067 ip_mr_cache List of multicast routing cache
1068..............................................................................
1069
1070You can use this information to see which network devices are available in
1071your system and how much traffic was routed over those devices:
1072
1073 > cat /proc/net/dev
1074 Inter-|Receive |[...
1075 face |bytes packets errs drop fifo frame compressed multicast|[...
1076 lo: 908188 5596 0 0 0 0 0 0 [...
1077 ppp0:15475140 20721 410 0 0 410 0 0 [...
1078 eth0: 614530 7085 0 0 0 0 0 1 [...
1079
1080 ...] Transmit
1081 ...] bytes packets errs drop fifo colls carrier compressed
1082 ...] 908188 5596 0 0 0 0 0 0
1083 ...] 1375103 17405 0 0 0 0 0 0
1084 ...] 1703981 5535 0 0 0 3 0 0
1085
1086In addition, each Channel Bond interface has its own directory. For
1087example, the bond0 device will have a directory called /proc/net/bond0/.
1088It will contain information that is specific to that bond, such as the
1089current slaves of the bond, the link status of the slaves, and how
1090many times the slaves link has failed.
1091
10921.5 SCSI info
1093-------------
1094
1095If you have a SCSI host adapter in your system, you'll find a subdirectory
1096named after the driver for this adapter in /proc/scsi. You'll also see a list
1097of all recognized SCSI devices in /proc/scsi:
1098
1099 >cat /proc/scsi/scsi
1100 Attached devices:
1101 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1102 Vendor: IBM Model: DGHS09U Rev: 03E0
1103 Type: Direct-Access ANSI SCSI revision: 03
1104 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1105 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1106 Type: CD-ROM ANSI SCSI revision: 02
1107
1108
1109The directory named after the driver has one file for each adapter found in
1110the system. These files contain information about the controller, including
1111the used IRQ and the IO address range. The amount of information shown is
1112dependent on the adapter you use. The example shows the output for an Adaptec
1113AHA-2940 SCSI adapter:
1114
1115 > cat /proc/scsi/aic7xxx/0
1116
1117 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1118 Compile Options:
1119 TCQ Enabled By Default : Disabled
1120 AIC7XXX_PROC_STATS : Disabled
1121 AIC7XXX_RESET_DELAY : 5
1122 Adapter Configuration:
1123 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1124 Ultra Wide Controller
1125 PCI MMAPed I/O Base: 0xeb001000
1126 Adapter SEEPROM Config: SEEPROM found and used.
1127 Adaptec SCSI BIOS: Enabled
1128 IRQ: 10
1129 SCBs: Active 0, Max Active 2,
1130 Allocated 15, HW 16, Page 255
1131 Interrupts: 160328
1132 BIOS Control Word: 0x18b6
1133 Adapter Control Word: 0x005b
1134 Extended Translation: Enabled
1135 Disconnect Enable Flags: 0xffff
1136 Ultra Enable Flags: 0x0001
1137 Tag Queue Enable Flags: 0x0000
1138 Ordered Queue Tag Flags: 0x0000
1139 Default Tag Queue Depth: 8
1140 Tagged Queue By Device array for aic7xxx host instance 0:
1141 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1142 Actual queue depth per device for aic7xxx host instance 0:
1143 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1144 Statistics:
1145 (scsi0:0:0:0)
1146 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1147 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1148 Total transfers 160151 (74577 reads and 85574 writes)
1149 (scsi0:0:6:0)
1150 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1151 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1152 Total transfers 0 (0 reads and 0 writes)
1153
1154
11551.6 Parallel port info in /proc/parport
1156---------------------------------------
1157
1158The directory /proc/parport contains information about the parallel ports of
1159your system. It has one subdirectory for each port, named after the port
1160number (0,1,2,...).
1161
1162These directories contain the four files shown in Table 1-10.
1163
1164
1165Table 1-10: Files in /proc/parport
1166..............................................................................
1167 File Content
1168 autoprobe Any IEEE-1284 device ID information that has been acquired.
1169 devices list of the device drivers using that port. A + will appear by the
1170 name of the device currently using the port (it might not appear
1171 against any).
1172 hardware Parallel port's base address, IRQ line and DMA channel.
1173 irq IRQ that parport is using for that port. This is in a separate
1174 file to allow you to alter it by writing a new value in (IRQ
1175 number or none).
1176..............................................................................
1177
11781.7 TTY info in /proc/tty
1179-------------------------
1180
1181Information about the available and actually used tty's can be found in the
1182directory /proc/tty.You'll find entries for drivers and line disciplines in
1183this directory, as shown in Table 1-11.
1184
1185
1186Table 1-11: Files in /proc/tty
1187..............................................................................
1188 File Content
1189 drivers list of drivers and their usage
1190 ldiscs registered line disciplines
1191 driver/serial usage statistic and status of single tty lines
1192..............................................................................
1193
1194To see which tty's are currently in use, you can simply look into the file
1195/proc/tty/drivers:
1196
1197 > cat /proc/tty/drivers
1198 pty_slave /dev/pts 136 0-255 pty:slave
1199 pty_master /dev/ptm 128 0-255 pty:master
1200 pty_slave /dev/ttyp 3 0-255 pty:slave
1201 pty_master /dev/pty 2 0-255 pty:master
1202 serial /dev/cua 5 64-67 serial:callout
1203 serial /dev/ttyS 4 64-67 serial
1204 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1205 /dev/ptmx /dev/ptmx 5 2 system
1206 /dev/console /dev/console 5 1 system:console
1207 /dev/tty /dev/tty 5 0 system:/dev/tty
1208 unknown /dev/tty 4 1-63 console
1209
1210
12111.8 Miscellaneous kernel statistics in /proc/stat
1212-------------------------------------------------
1213
1214Various pieces of information about kernel activity are available in the
1215/proc/stat file. All of the numbers reported in this file are aggregates
1216since the system first booted. For a quick look, simply cat the file:
1217
1218 > cat /proc/stat
1219 cpu 2255 34 2290 22625563 6290 127 456 0 0
1220 cpu0 1132 34 1441 11311718 3675 127 438 0 0
1221 cpu1 1123 0 849 11313845 2614 0 18 0 0
1222 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1223 ctxt 1990473
1224 btime 1062191376
1225 processes 2915
1226 procs_running 1
1227 procs_blocked 0
1228 softirq 183433 0 21755 12 39 1137 231 21459 2263
1229
1230The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1231lines. These numbers identify the amount of time the CPU has spent performing
1232different kinds of work. Time units are in USER_HZ (typically hundredths of a
1233second). The meanings of the columns are as follows, from left to right:
1234
1235- user: normal processes executing in user mode
1236- nice: niced processes executing in user mode
1237- system: processes executing in kernel mode
1238- idle: twiddling thumbs
1239- iowait: waiting for I/O to complete
1240- irq: servicing interrupts
1241- softirq: servicing softirqs
1242- steal: involuntary wait
1243- guest: running a normal guest
1244- guest_nice: running a niced guest
1245
1246The "intr" line gives counts of interrupts serviced since boot time, for each
1247of the possible system interrupts. The first column is the total of all
1248interrupts serviced including unnumbered architecture specific interrupts;
1249each subsequent column is the total for that particular numbered interrupt.
1250Unnumbered interrupts are not shown, only summed into the total.
1251
1252The "ctxt" line gives the total number of context switches across all CPUs.
1253
1254The "btime" line gives the time at which the system booted, in seconds since
1255the Unix epoch.
1256
1257The "processes" line gives the number of processes and threads created, which
1258includes (but is not limited to) those created by calls to the fork() and
1259clone() system calls.
1260
1261The "procs_running" line gives the total number of threads that are
1262running or ready to run (i.e., the total number of runnable threads).
1263
1264The "procs_blocked" line gives the number of processes currently blocked,
1265waiting for I/O to complete.
1266
1267The "softirq" line gives counts of softirqs serviced since boot time, for each
1268of the possible system softirqs. The first column is the total of all
1269softirqs serviced; each subsequent column is the total for that particular
1270softirq.
1271
1272
12731.9 Ext4 file system parameters
1274------------------------------
1275
1276Information about mounted ext4 file systems can be found in
1277/proc/fs/ext4. Each mounted filesystem will have a directory in
1278/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1279/proc/fs/ext4/dm-0). The files in each per-device directory are shown
1280in Table 1-12, below.
1281
1282Table 1-12: Files in /proc/fs/ext4/<devname>
1283..............................................................................
1284 File Content
1285 mb_groups details of multiblock allocator buddy cache of free blocks
1286..............................................................................
1287
12882.0 /proc/consoles
1289------------------
1290Shows registered system console lines.
1291
1292To see which character device lines are currently used for the system console
1293/dev/console, you may simply look into the file /proc/consoles:
1294
1295 > cat /proc/consoles
1296 tty0 -WU (ECp) 4:7
1297 ttyS0 -W- (Ep) 4:64
1298
1299The columns are:
1300
1301 device name of the device
1302 operations R = can do read operations
1303 W = can do write operations
1304 U = can do unblank
1305 flags E = it is enabled
1306 C = it is preferred console
1307 B = it is primary boot console
1308 p = it is used for printk buffer
1309 b = it is not a TTY but a Braille device
1310 a = it is safe to use when cpu is offline
1311 major:minor major and minor number of the device separated by a colon
1312
1313------------------------------------------------------------------------------
1314Summary
1315------------------------------------------------------------------------------
1316The /proc file system serves information about the running system. It not only
1317allows access to process data but also allows you to request the kernel status
1318by reading files in the hierarchy.
1319
1320The directory structure of /proc reflects the types of information and makes
1321it easy, if not obvious, where to look for specific data.
1322------------------------------------------------------------------------------
1323
1324------------------------------------------------------------------------------
1325CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1326------------------------------------------------------------------------------
1327
1328------------------------------------------------------------------------------
1329In This Chapter
1330------------------------------------------------------------------------------
1331* Modifying kernel parameters by writing into files found in /proc/sys
1332* Exploring the files which modify certain parameters
1333* Review of the /proc/sys file tree
1334------------------------------------------------------------------------------
1335
1336
1337A very interesting part of /proc is the directory /proc/sys. This is not only
1338a source of information, it also allows you to change parameters within the
1339kernel. Be very careful when attempting this. You can optimize your system,
1340but you can also cause it to crash. Never alter kernel parameters on a
1341production system. Set up a development machine and test to make sure that
1342everything works the way you want it to. You may have no alternative but to
1343reboot the machine once an error has been made.
1344
1345To change a value, simply echo the new value into the file. An example is
1346given below in the section on the file system data. You need to be root to do
1347this. You can create your own boot script to perform this every time your
1348system boots.
1349
1350The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1351general things in the operation of the Linux kernel. Since some of the files
1352can inadvertently disrupt your system, it is advisable to read both
1353documentation and source before actually making adjustments. In any case, be
1354very careful when writing to any of these files. The entries in /proc may
1355change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1356review the kernel documentation in the directory /usr/src/linux/Documentation.
1357This chapter is heavily based on the documentation included in the pre 2.2
1358kernels, and became part of it in version 2.2.1 of the Linux kernel.
1359
1360Please see: Documentation/sysctl/ directory for descriptions of these
1361entries.
1362
1363------------------------------------------------------------------------------
1364Summary
1365------------------------------------------------------------------------------
1366Certain aspects of kernel behavior can be modified at runtime, without the
1367need to recompile the kernel, or even to reboot the system. The files in the
1368/proc/sys tree can not only be read, but also modified. You can use the echo
1369command to write value into these files, thereby changing the default settings
1370of the kernel.
1371------------------------------------------------------------------------------
1372
1373------------------------------------------------------------------------------
1374CHAPTER 3: PER-PROCESS PARAMETERS
1375------------------------------------------------------------------------------
1376
13773.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1378--------------------------------------------------------------------------------
1379
1380These file can be used to adjust the badness heuristic used to select which
1381process gets killed in out of memory conditions.
1382
1383The badness heuristic assigns a value to each candidate task ranging from 0
1384(never kill) to 1000 (always kill) to determine which process is targeted. The
1385units are roughly a proportion along that range of allowed memory the process
1386may allocate from based on an estimation of its current memory and swap use.
1387For example, if a task is using all allowed memory, its badness score will be
13881000. If it is using half of its allowed memory, its score will be 500.
1389
1390There is an additional factor included in the badness score: the current memory
1391and swap usage is discounted by 3% for root processes.
1392
1393The amount of "allowed" memory depends on the context in which the oom killer
1394was called. If it is due to the memory assigned to the allocating task's cpuset
1395being exhausted, the allowed memory represents the set of mems assigned to that
1396cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1397memory represents the set of mempolicy nodes. If it is due to a memory
1398limit (or swap limit) being reached, the allowed memory is that configured
1399limit. Finally, if it is due to the entire system being out of memory, the
1400allowed memory represents all allocatable resources.
1401
1402The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1403is used to determine which task to kill. Acceptable values range from -1000
1404(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1405polarize the preference for oom killing either by always preferring a certain
1406task or completely disabling it. The lowest possible value, -1000, is
1407equivalent to disabling oom killing entirely for that task since it will always
1408report a badness score of 0.
1409
1410Consequently, it is very simple for userspace to define the amount of memory to
1411consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1412example, is roughly equivalent to allowing the remainder of tasks sharing the
1413same system, cpuset, mempolicy, or memory controller resources to use at least
141450% more memory. A value of -500, on the other hand, would be roughly
1415equivalent to discounting 50% of the task's allowed memory from being considered
1416as scoring against the task.
1417
1418For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1419be used to tune the badness score. Its acceptable values range from -16
1420(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1421(OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1422scaled linearly with /proc/<pid>/oom_score_adj.
1423
1424The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1425value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1426requires CAP_SYS_RESOURCE.
1427
1428Caveat: when a parent task is selected, the oom killer will sacrifice any first
1429generation children with separate address spaces instead, if possible. This
1430avoids servers and important system daemons from being killed and loses the
1431minimal amount of work.
1432
1433
14343.2 /proc/<pid>/oom_score - Display current oom-killer score
1435-------------------------------------------------------------
1436
1437This file can be used to check the current score used by the oom-killer is for
1438any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1439process should be killed in an out-of-memory situation.
1440
1441
14423.3 /proc/<pid>/io - Display the IO accounting fields
1443-------------------------------------------------------
1444
1445This file contains IO statistics for each running process
1446
1447Example
1448-------
1449
1450test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1451[1] 3828
1452
1453test:/tmp # cat /proc/3828/io
1454rchar: 323934931
1455wchar: 323929600
1456syscr: 632687
1457syscw: 632675
1458read_bytes: 0
1459write_bytes: 323932160
1460cancelled_write_bytes: 0
1461
1462
1463Description
1464-----------
1465
1466rchar
1467-----
1468
1469I/O counter: chars read
1470The number of bytes which this task has caused to be read from storage. This
1471is simply the sum of bytes which this process passed to read() and pread().
1472It includes things like tty IO and it is unaffected by whether or not actual
1473physical disk IO was required (the read might have been satisfied from
1474pagecache)
1475
1476
1477wchar
1478-----
1479
1480I/O counter: chars written
1481The number of bytes which this task has caused, or shall cause to be written
1482to disk. Similar caveats apply here as with rchar.
1483
1484
1485syscr
1486-----
1487
1488I/O counter: read syscalls
1489Attempt to count the number of read I/O operations, i.e. syscalls like read()
1490and pread().
1491
1492
1493syscw
1494-----
1495
1496I/O counter: write syscalls
1497Attempt to count the number of write I/O operations, i.e. syscalls like
1498write() and pwrite().
1499
1500
1501read_bytes
1502----------
1503
1504I/O counter: bytes read
1505Attempt to count the number of bytes which this process really did cause to
1506be fetched from the storage layer. Done at the submit_bio() level, so it is
1507accurate for block-backed filesystems. <please add status regarding NFS and
1508CIFS at a later time>
1509
1510
1511write_bytes
1512-----------
1513
1514I/O counter: bytes written
1515Attempt to count the number of bytes which this process caused to be sent to
1516the storage layer. This is done at page-dirtying time.
1517
1518
1519cancelled_write_bytes
1520---------------------
1521
1522The big inaccuracy here is truncate. If a process writes 1MB to a file and
1523then deletes the file, it will in fact perform no writeout. But it will have
1524been accounted as having caused 1MB of write.
1525In other words: The number of bytes which this process caused to not happen,
1526by truncating pagecache. A task can cause "negative" IO too. If this task
1527truncates some dirty pagecache, some IO which another task has been accounted
1528for (in its write_bytes) will not be happening. We _could_ just subtract that
1529from the truncating task's write_bytes, but there is information loss in doing
1530that.
1531
1532
1533Note
1534----
1535
1536At its current implementation state, this is a bit racy on 32-bit machines: if
1537process A reads process B's /proc/pid/io while process B is updating one of
1538those 64-bit counters, process A could see an intermediate result.
1539
1540
1541More information about this can be found within the taskstats documentation in
1542Documentation/accounting.
1543
15443.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1545---------------------------------------------------------------
1546When a process is dumped, all anonymous memory is written to a core file as
1547long as the size of the core file isn't limited. But sometimes we don't want
1548to dump some memory segments, for example, huge shared memory. Conversely,
1549sometimes we want to save file-backed memory segments into a core file, not
1550only the individual files.
1551
1552/proc/<pid>/coredump_filter allows you to customize which memory segments
1553will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1554of memory types. If a bit of the bitmask is set, memory segments of the
1555corresponding memory type are dumped, otherwise they are not dumped.
1556
1557The following 7 memory types are supported:
1558 - (bit 0) anonymous private memory
1559 - (bit 1) anonymous shared memory
1560 - (bit 2) file-backed private memory
1561 - (bit 3) file-backed shared memory
1562 - (bit 4) ELF header pages in file-backed private memory areas (it is
1563 effective only if the bit 2 is cleared)
1564 - (bit 5) hugetlb private memory
1565 - (bit 6) hugetlb shared memory
1566
1567 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1568 are always dumped regardless of the bitmask status.
1569
1570 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1571 effected by bit 5-6.
1572
1573Default value of coredump_filter is 0x23; this means all anonymous memory
1574segments and hugetlb private memory are dumped.
1575
1576If you don't want to dump all shared memory segments attached to pid 1234,
1577write 0x21 to the process's proc file.
1578
1579 $ echo 0x21 > /proc/1234/coredump_filter
1580
1581When a new process is created, the process inherits the bitmask status from its
1582parent. It is useful to set up coredump_filter before the program runs.
1583For example:
1584
1585 $ echo 0x7 > /proc/self/coredump_filter
1586 $ ./some_program
1587
15883.5 /proc/<pid>/mountinfo - Information about mounts
1589--------------------------------------------------------
1590
1591This file contains lines of the form:
1592
159336 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1594(1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1595
1596(1) mount ID: unique identifier of the mount (may be reused after umount)
1597(2) parent ID: ID of parent (or of self for the top of the mount tree)
1598(3) major:minor: value of st_dev for files on filesystem
1599(4) root: root of the mount within the filesystem
1600(5) mount point: mount point relative to the process's root
1601(6) mount options: per mount options
1602(7) optional fields: zero or more fields of the form "tag[:value]"
1603(8) separator: marks the end of the optional fields
1604(9) filesystem type: name of filesystem of the form "type[.subtype]"
1605(10) mount source: filesystem specific information or "none"
1606(11) super options: per super block options
1607
1608Parsers should ignore all unrecognised optional fields. Currently the
1609possible optional fields are:
1610
1611shared:X mount is shared in peer group X
1612master:X mount is slave to peer group X
1613propagate_from:X mount is slave and receives propagation from peer group X (*)
1614unbindable mount is unbindable
1615
1616(*) X is the closest dominant peer group under the process's root. If
1617X is the immediate master of the mount, or if there's no dominant peer
1618group under the same root, then only the "master:X" field is present
1619and not the "propagate_from:X" field.
1620
1621For more information on mount propagation see:
1622
1623 Documentation/filesystems/sharedsubtree.txt
1624
1625
16263.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1627--------------------------------------------------------
1628These files provide a method to access a tasks comm value. It also allows for
1629a task to set its own or one of its thread siblings comm value. The comm value
1630is limited in size compared to the cmdline value, so writing anything longer
1631then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1632comm value.
1633
1634
16353.7 /proc/<pid>/task/<tid>/children - Information about task children
1636-------------------------------------------------------------------------
1637This file provides a fast way to retrieve first level children pids
1638of a task pointed by <pid>/<tid> pair. The format is a space separated
1639stream of pids.
1640
1641Note the "first level" here -- if a child has own children they will
1642not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1643to obtain the descendants.
1644
1645Since this interface is intended to be fast and cheap it doesn't
1646guarantee to provide precise results and some children might be
1647skipped, especially if they've exited right after we printed their
1648pids, so one need to either stop or freeze processes being inspected
1649if precise results are needed.
1650
1651
16523.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1653---------------------------------------------------------------
1654This file provides information associated with an opened file. The regular
1655files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1656represents the current offset of the opened file in decimal form [see lseek(2)
1657for details], 'flags' denotes the octal O_xxx mask the file has been
1658created with [see open(2) for details] and 'mnt_id' represents mount ID of
1659the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1660for details].
1661
1662A typical output is
1663
1664 pos: 0
1665 flags: 0100002
1666 mnt_id: 19
1667
1668The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1669pair provide additional information particular to the objects they represent.
1670
1671 Eventfd files
1672 ~~~~~~~~~~~~~
1673 pos: 0
1674 flags: 04002
1675 mnt_id: 9
1676 eventfd-count: 5a
1677
1678 where 'eventfd-count' is hex value of a counter.
1679
1680 Signalfd files
1681 ~~~~~~~~~~~~~~
1682 pos: 0
1683 flags: 04002
1684 mnt_id: 9
1685 sigmask: 0000000000000200
1686
1687 where 'sigmask' is hex value of the signal mask associated
1688 with a file.
1689
1690 Epoll files
1691 ~~~~~~~~~~~
1692 pos: 0
1693 flags: 02
1694 mnt_id: 9
1695 tfd: 5 events: 1d data: ffffffffffffffff
1696
1697 where 'tfd' is a target file descriptor number in decimal form,
1698 'events' is events mask being watched and the 'data' is data
1699 associated with a target [see epoll(7) for more details].
1700
1701 Fsnotify files
1702 ~~~~~~~~~~~~~~
1703 For inotify files the format is the following
1704
1705 pos: 0
1706 flags: 02000000
1707 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1708
1709 where 'wd' is a watch descriptor in decimal form, ie a target file
1710 descriptor number, 'ino' and 'sdev' are inode and device where the
1711 target file resides and the 'mask' is the mask of events, all in hex
1712 form [see inotify(7) for more details].
1713
1714 If the kernel was built with exportfs support, the path to the target
1715 file is encoded as a file handle. The file handle is provided by three
1716 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1717 format.
1718
1719 If the kernel is built without exportfs support the file handle won't be
1720 printed out.
1721
1722 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1723
1724 For fanotify files the format is
1725
1726 pos: 0
1727 flags: 02
1728 mnt_id: 9
1729 fanotify flags:10 event-flags:0
1730 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1731 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1732
1733 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1734 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1735 flags associated with mark which are tracked separately from events
1736 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1737 mask and 'ignored_mask' is the mask of events which are to be ignored.
1738 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1739 does provide information about flags and mask used in fanotify_mark
1740 call [see fsnotify manpage for details].
1741
1742 While the first three lines are mandatory and always printed, the rest is
1743 optional and may be omitted if no marks created yet.
1744
1745
1746------------------------------------------------------------------------------
1747Configuring procfs
1748------------------------------------------------------------------------------
1749
17504.1 Mount options
1751---------------------
1752
1753The following mount options are supported:
1754
1755 hidepid= Set /proc/<pid>/ access mode.
1756 gid= Set the group authorized to learn processes information.
1757
1758hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1759(default).
1760
1761hidepid=1 means users may not access any /proc/<pid>/ directories but their
1762own. Sensitive files like cmdline, sched*, status are now protected against
1763other users. This makes it impossible to learn whether any user runs
1764specific program (given the program doesn't reveal itself by its behaviour).
1765As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1766poorly written programs passing sensitive information via program arguments are
1767now protected against local eavesdroppers.
1768
1769hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1770users. It doesn't mean that it hides a fact whether a process with a specific
1771pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1772but it hides process' uid and gid, which may be learned by stat()'ing
1773/proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1774information about running processes, whether some daemon runs with elevated
1775privileges, whether other user runs some sensitive program, whether other users
1776run any program at all, etc.
1777
1778gid= defines a group authorized to learn processes information otherwise
1779prohibited by hidepid=. If you use some daemon like identd which needs to learn
1780information about processes information, just add identd to this group.
1------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3------------------------------------------------------------------------------
4/proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
6
72.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
9------------------------------------------------------------------------------
10Version 1.3 Kernel version 2.2.12
11 Kernel version 2.4.0-test11-pre4
12------------------------------------------------------------------------------
13fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
14
15Table of Contents
16-----------------
17
18 0 Preface
19 0.1 Introduction/Credits
20 0.2 Legal Stuff
21
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
32
33 2 Modifying System Parameters
34
35 3 Per-Process Parameters
36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
37 score
38 3.2 /proc/<pid>/oom_score - Display current oom-killer score
39 3.3 /proc/<pid>/io - Display the IO accounting fields
40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
41 3.5 /proc/<pid>/mountinfo - Information about mounts
42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
43 3.7 /proc/<pid>/task/<tid>/children - Information about task children
44
45 4 Configuring procfs
46 4.1 Mount options
47
48------------------------------------------------------------------------------
49Preface
50------------------------------------------------------------------------------
51
520.1 Introduction/Credits
53------------------------
54
55This documentation is part of a soon (or so we hope) to be released book on
56the 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
58chapters, it seems only fair to give the work back to the Linux community.
59This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
60afraid it's still far from complete, but we hope it will be useful. As far as
61we know, it is the first 'all-in-one' document about the /proc file system. It
62is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
63SPARC, AXP, etc., features, you probably won't find what you are looking for.
64It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
65additions and patches are welcome and will be added to this document if you
66mail them to Bodo.
67
68We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
69other people for help compiling this documentation. We'd also like to extend a
70special thank you to Andi Kleen for documentation, which we relied on heavily
71to create this document, as well as the additional information he provided.
72Thanks to everybody else who contributed source or docs to the Linux kernel
73and helped create a great piece of software... :)
74
75If you have any comments, corrections or additions, please don't hesitate to
76contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
77document.
78
79The latest version of this document is available online at
80http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
81
82If the above direction does not works for you, you could try the kernel
83mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
84comandante@zaralinux.com.
85
860.2 Legal Stuff
87---------------
88
89We don't guarantee the correctness of this document, and if you come to us
90complaining about how you screwed up your system because of incorrect
91documentation, we won't feel responsible...
92
93------------------------------------------------------------------------------
94CHAPTER 1: COLLECTING SYSTEM INFORMATION
95------------------------------------------------------------------------------
96
97------------------------------------------------------------------------------
98In This Chapter
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
104 on the system
105------------------------------------------------------------------------------
106
107
108The proc file system acts as an interface to internal data structures in the
109kernel. It can be used to obtain information about the system and to change
110certain kernel parameters at runtime (sysctl).
111
112First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
113show you how you can use /proc/sys to change settings.
114
1151.1 Process-Specific Subdirectories
116-----------------------------------
117
118The directory /proc contains (among other things) one subdirectory for each
119process running on the system, which is named after the process ID (PID).
120
121The link self points to the process reading the file system. Each process
122subdirectory has the entries listed in Table 1-1.
123
124
125Table 1-1: Process specific entries in /proc
126..............................................................................
127 File Content
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
138 stat Process status
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
142 pagemap Page table
143 stack Report full stack trace, enable via CONFIG_STACKTRACE
144 smaps a extension based on maps, showing the memory consumption of
145 each mapping
146..............................................................................
147
148For example, to get the status information of a process, all you have to do is
149read the file /proc/PID/status:
150
151 >cat /proc/self/status
152 Name: cat
153 State: R (running)
154 Tgid: 5452
155 Pid: 5452
156 PPid: 743
157 TracerPid: 0 (2.4)
158 Uid: 501 501 501 501
159 Gid: 100 100 100 100
160 FDSize: 256
161 Groups: 100 14 16
162 VmPeak: 5004 kB
163 VmSize: 5004 kB
164 VmLck: 0 kB
165 VmHWM: 476 kB
166 VmRSS: 476 kB
167 VmData: 156 kB
168 VmStk: 88 kB
169 VmExe: 68 kB
170 VmLib: 1412 kB
171 VmPTE: 20 kb
172 VmSwap: 0 kB
173 Threads: 1
174 SigQ: 0/28578
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
186
187This shows you nearly the same information you would get if you viewed it with
188the ps command. In fact, ps uses the proc file system to obtain its
189information. But you get a more detailed view of the process by reading the
190file /proc/PID/status. It fields are described in table 1-2.
191
192The statm file contains more detailed information about the process
193memory usage. Its seven fields are explained in Table 1-3. The stat file
194contains details information about the process itself. Its fields are
195explained in Table 1-4.
196
197(for SMP CONFIG users)
198For making accounting scalable, RSS related information are handled in
199asynchronous manner and the vaule may not be very precise. To see a precise
200snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
201It's slow but very precise.
202
203Table 1-2: Contents of the status files (as of 2.6.30-rc7)
204..............................................................................
205 Field Content
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)
210 Tgid thread group ID
211 Pid process id
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..............................................................................
247
248Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
249..............................................................................
250 Field Content
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..............................................................................
261
262
263Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
264..............................................................................
265 Field Content
266 pid process id
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
272 sid session id
273 tty_nr tty the process uses
274 tty_pgrp pgrp of the tty
275 flags task flags
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
285 nice nice 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
302 0 (place holder)
303 0 (place holder)
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 arg_start address above which program command line is placed
315 arg_end address below which program command line is placed
316 env_start address above which program environment is placed
317 env_end address below which program environment is placed
318 exit_code the thread's exit_code in the form reported by the waitpid system call
319..............................................................................
320
321The /proc/PID/maps file containing the currently mapped memory regions and
322their access permissions.
323
324The format is:
325
326address perms offset dev inode pathname
327
32808048000-08049000 r-xp 00000000 03:00 8312 /opt/test
32908049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
3300804a000-0806b000 rw-p 00000000 00:00 0 [heap]
331a7cb1000-a7cb2000 ---p 00000000 00:00 0
332a7cb2000-a7eb2000 rw-p 00000000 00:00 0
333a7eb2000-a7eb3000 ---p 00000000 00:00 0
334a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001]
335a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
336a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
337a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
338a800b000-a800e000 rw-p 00000000 00:00 0
339a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
340a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
341a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
342a8024000-a8027000 rw-p 00000000 00:00 0
343a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
344a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
345a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
346aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
347ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
348
349where "address" is the address space in the process that it occupies, "perms"
350is a set of permissions:
351
352 r = read
353 w = write
354 x = execute
355 s = shared
356 p = private (copy on write)
357
358"offset" is the offset into the mapping, "dev" is the device (major:minor), and
359"inode" is the inode on that device. 0 indicates that no inode is associated
360with the memory region, as the case would be with BSS (uninitialized data).
361The "pathname" shows the name associated file for this mapping. If the mapping
362is not associated with a file:
363
364 [heap] = the heap of the program
365 [stack] = the stack of the main process
366 [stack:1001] = the stack of the thread with tid 1001
367 [vdso] = the "virtual dynamic shared object",
368 the kernel system call handler
369
370 or if empty, the mapping is anonymous.
371
372The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
373of the individual tasks of a process. In this file you will see a mapping marked
374as [stack] if that task sees it as a stack. This is a key difference from the
375content of /proc/PID/maps, where you will see all mappings that are being used
376as stack by all of those tasks. Hence, for the example above, the task-level
377map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
378
37908048000-08049000 r-xp 00000000 03:00 8312 /opt/test
38008049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
3810804a000-0806b000 rw-p 00000000 00:00 0 [heap]
382a7cb1000-a7cb2000 ---p 00000000 00:00 0
383a7cb2000-a7eb2000 rw-p 00000000 00:00 0
384a7eb2000-a7eb3000 ---p 00000000 00:00 0
385a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack]
386a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
387a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
388a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
389a800b000-a800e000 rw-p 00000000 00:00 0
390a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
391a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
392a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
393a8024000-a8027000 rw-p 00000000 00:00 0
394a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
395a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
396a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
397aff35000-aff4a000 rw-p 00000000 00:00 0
398ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
399
400The /proc/PID/smaps is an extension based on maps, showing the memory
401consumption for each of the process's mappings. For each of mappings there
402is a series of lines such as the following:
403
40408048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
405Size: 1084 kB
406Rss: 892 kB
407Pss: 374 kB
408Shared_Clean: 892 kB
409Shared_Dirty: 0 kB
410Private_Clean: 0 kB
411Private_Dirty: 0 kB
412Referenced: 892 kB
413Anonymous: 0 kB
414Swap: 0 kB
415KernelPageSize: 4 kB
416MMUPageSize: 4 kB
417Locked: 374 kB
418
419The first of these lines shows the same information as is displayed for the
420mapping in /proc/PID/maps. The remaining lines show the size of the mapping
421(size), the amount of the mapping that is currently resident in RAM (RSS), the
422process' proportional share of this mapping (PSS), the number of clean and
423dirty private pages in the mapping. Note that even a page which is part of a
424MAP_SHARED mapping, but has only a single pte mapped, i.e. is currently used
425by only one process, is accounted as private and not as shared. "Referenced"
426indicates the amount of memory currently marked as referenced or accessed.
427"Anonymous" shows the amount of memory that does not belong to any file. Even
428a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
429and a page is modified, the file page is replaced by a private anonymous copy.
430"Swap" shows how much would-be-anonymous memory is also used, but out on
431swap.
432
433This file is only present if the CONFIG_MMU kernel configuration option is
434enabled.
435
436The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
437bits on both physical and virtual pages associated with a process.
438To clear the bits for all the pages associated with the process
439 > echo 1 > /proc/PID/clear_refs
440
441To clear the bits for the anonymous pages associated with the process
442 > echo 2 > /proc/PID/clear_refs
443
444To clear the bits for the file mapped pages associated with the process
445 > echo 3 > /proc/PID/clear_refs
446Any other value written to /proc/PID/clear_refs will have no effect.
447
448The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
449using /proc/kpageflags and number of times a page is mapped using
450/proc/kpagecount. For detailed explanation, see Documentation/vm/pagemap.txt.
451
4521.2 Kernel data
453---------------
454
455Similar to the process entries, the kernel data files give information about
456the running kernel. The files used to obtain this information are contained in
457/proc and are listed in Table 1-5. Not all of these will be present in your
458system. It depends on the kernel configuration and the loaded modules, which
459files are there, and which are missing.
460
461Table 1-5: Kernel info in /proc
462..............................................................................
463 File Content
464 apm Advanced power management info
465 buddyinfo Kernel memory allocator information (see text) (2.5)
466 bus Directory containing bus specific information
467 cmdline Kernel command line
468 cpuinfo Info about the CPU
469 devices Available devices (block and character)
470 dma Used DMS channels
471 filesystems Supported filesystems
472 driver Various drivers grouped here, currently rtc (2.4)
473 execdomains Execdomains, related to security (2.4)
474 fb Frame Buffer devices (2.4)
475 fs File system parameters, currently nfs/exports (2.4)
476 ide Directory containing info about the IDE subsystem
477 interrupts Interrupt usage
478 iomem Memory map (2.4)
479 ioports I/O port usage
480 irq Masks for irq to cpu affinity (2.4)(smp?)
481 isapnp ISA PnP (Plug&Play) Info (2.4)
482 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
483 kmsg Kernel messages
484 ksyms Kernel symbol table
485 loadavg Load average of last 1, 5 & 15 minutes
486 locks Kernel locks
487 meminfo Memory info
488 misc Miscellaneous
489 modules List of loaded modules
490 mounts Mounted filesystems
491 net Networking info (see text)
492 pagetypeinfo Additional page allocator information (see text) (2.5)
493 partitions Table of partitions known to the system
494 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
495 decoupled by lspci (2.4)
496 rtc Real time clock
497 scsi SCSI info (see text)
498 slabinfo Slab pool info
499 softirqs softirq usage
500 stat Overall statistics
501 swaps Swap space utilization
502 sys See chapter 2
503 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
504 tty Info of tty drivers
505 uptime System uptime
506 version Kernel version
507 video bttv info of video resources (2.4)
508 vmallocinfo Show vmalloced areas
509..............................................................................
510
511You can, for example, check which interrupts are currently in use and what
512they are used for by looking in the file /proc/interrupts:
513
514 > cat /proc/interrupts
515 CPU0
516 0: 8728810 XT-PIC timer
517 1: 895 XT-PIC keyboard
518 2: 0 XT-PIC cascade
519 3: 531695 XT-PIC aha152x
520 4: 2014133 XT-PIC serial
521 5: 44401 XT-PIC pcnet_cs
522 8: 2 XT-PIC rtc
523 11: 8 XT-PIC i82365
524 12: 182918 XT-PIC PS/2 Mouse
525 13: 1 XT-PIC fpu
526 14: 1232265 XT-PIC ide0
527 15: 7 XT-PIC ide1
528 NMI: 0
529
530In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
531output of a SMP machine):
532
533 > cat /proc/interrupts
534
535 CPU0 CPU1
536 0: 1243498 1214548 IO-APIC-edge timer
537 1: 8949 8958 IO-APIC-edge keyboard
538 2: 0 0 XT-PIC cascade
539 5: 11286 10161 IO-APIC-edge soundblaster
540 8: 1 0 IO-APIC-edge rtc
541 9: 27422 27407 IO-APIC-edge 3c503
542 12: 113645 113873 IO-APIC-edge PS/2 Mouse
543 13: 0 0 XT-PIC fpu
544 14: 22491 24012 IO-APIC-edge ide0
545 15: 2183 2415 IO-APIC-edge ide1
546 17: 30564 30414 IO-APIC-level eth0
547 18: 177 164 IO-APIC-level bttv
548 NMI: 2457961 2457959
549 LOC: 2457882 2457881
550 ERR: 2155
551
552NMI is incremented in this case because every timer interrupt generates a NMI
553(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
554
555LOC is the local interrupt counter of the internal APIC of every CPU.
556
557ERR is incremented in the case of errors in the IO-APIC bus (the bus that
558connects the CPUs in a SMP system. This means that an error has been detected,
559the IO-APIC automatically retry the transmission, so it should not be a big
560problem, but you should read the SMP-FAQ.
561
562In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
563/proc/interrupts to display every IRQ vector in use by the system, not
564just those considered 'most important'. The new vectors are:
565
566 THR -- interrupt raised when a machine check threshold counter
567 (typically counting ECC corrected errors of memory or cache) exceeds
568 a configurable threshold. Only available on some systems.
569
570 TRM -- a thermal event interrupt occurs when a temperature threshold
571 has been exceeded for the CPU. This interrupt may also be generated
572 when the temperature drops back to normal.
573
574 SPU -- a spurious interrupt is some interrupt that was raised then lowered
575 by some IO device before it could be fully processed by the APIC. Hence
576 the APIC sees the interrupt but does not know what device it came from.
577 For this case the APIC will generate the interrupt with a IRQ vector
578 of 0xff. This might also be generated by chipset bugs.
579
580 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
581 sent from one CPU to another per the needs of the OS. Typically,
582 their statistics are used by kernel developers and interested users to
583 determine the occurrence of interrupts of the given type.
584
585The above IRQ vectors are displayed only when relevant. For example,
586the threshold vector does not exist on x86_64 platforms. Others are
587suppressed when the system is a uniprocessor. As of this writing, only
588i386 and x86_64 platforms support the new IRQ vector displays.
589
590Of some interest is the introduction of the /proc/irq directory to 2.4.
591It could be used to set IRQ to CPU affinity, this means that you can "hook" an
592IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
593irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
594prof_cpu_mask.
595
596For example
597 > ls /proc/irq/
598 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
599 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
600 > ls /proc/irq/0/
601 smp_affinity
602
603smp_affinity is a bitmask, in which you can specify which CPUs can handle the
604IRQ, you can set it by doing:
605
606 > echo 1 > /proc/irq/10/smp_affinity
607
608This means that only the first CPU will handle the IRQ, but you can also echo
6095 which means that only the first and fourth CPU can handle the IRQ.
610
611The contents of each smp_affinity file is the same by default:
612
613 > cat /proc/irq/0/smp_affinity
614 ffffffff
615
616There is an alternate interface, smp_affinity_list which allows specifying
617a cpu range instead of a bitmask:
618
619 > cat /proc/irq/0/smp_affinity_list
620 1024-1031
621
622The default_smp_affinity mask applies to all non-active IRQs, which are the
623IRQs which have not yet been allocated/activated, and hence which lack a
624/proc/irq/[0-9]* directory.
625
626The node file on an SMP system shows the node to which the device using the IRQ
627reports itself as being attached. This hardware locality information does not
628include information about any possible driver locality preference.
629
630prof_cpu_mask specifies which CPUs are to be profiled by the system wide
631profiler. Default value is ffffffff (all cpus if there are only 32 of them).
632
633The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
634between all the CPUs which are allowed to handle it. As usual the kernel has
635more info than you and does a better job than you, so the defaults are the
636best choice for almost everyone. [Note this applies only to those IO-APIC's
637that support "Round Robin" interrupt distribution.]
638
639There are three more important subdirectories in /proc: net, scsi, and sys.
640The general rule is that the contents, or even the existence of these
641directories, depend on your kernel configuration. If SCSI is not enabled, the
642directory scsi may not exist. The same is true with the net, which is there
643only when networking support is present in the running kernel.
644
645The slabinfo file gives information about memory usage at the slab level.
646Linux uses slab pools for memory management above page level in version 2.2.
647Commonly used objects have their own slab pool (such as network buffers,
648directory cache, and so on).
649
650..............................................................................
651
652> cat /proc/buddyinfo
653
654Node 0, zone DMA 0 4 5 4 4 3 ...
655Node 0, zone Normal 1 0 0 1 101 8 ...
656Node 0, zone HighMem 2 0 0 1 1 0 ...
657
658External fragmentation is a problem under some workloads, and buddyinfo is a
659useful tool for helping diagnose these problems. Buddyinfo will give you a
660clue as to how big an area you can safely allocate, or why a previous
661allocation failed.
662
663Each column represents the number of pages of a certain order which are
664available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
665ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
666available in ZONE_NORMAL, etc...
667
668More information relevant to external fragmentation can be found in
669pagetypeinfo.
670
671> cat /proc/pagetypeinfo
672Page block order: 9
673Pages per block: 512
674
675Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
676Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
677Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
678Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
679Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
680Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
681Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
682Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
683Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
684Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
685Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
686
687Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
688Node 0, zone DMA 2 0 5 1 0
689Node 0, zone DMA32 41 6 967 2 0
690
691Fragmentation avoidance in the kernel works by grouping pages of different
692migrate types into the same contiguous regions of memory called page blocks.
693A page block is typically the size of the default hugepage size e.g. 2MB on
694X86-64. By keeping pages grouped based on their ability to move, the kernel
695can reclaim pages within a page block to satisfy a high-order allocation.
696
697The pagetypinfo begins with information on the size of a page block. It
698then gives the same type of information as buddyinfo except broken down
699by migrate-type and finishes with details on how many page blocks of each
700type exist.
701
702If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
703from libhugetlbfs http://sourceforge.net/projects/libhugetlbfs/), one can
704make an estimate of the likely number of huge pages that can be allocated
705at a given point in time. All the "Movable" blocks should be allocatable
706unless memory has been mlock()'d. Some of the Reclaimable blocks should
707also be allocatable although a lot of filesystem metadata may have to be
708reclaimed to achieve this.
709
710..............................................................................
711
712meminfo:
713
714Provides information about distribution and utilization of memory. This
715varies by architecture and compile options. The following is from a
71616GB PIII, which has highmem enabled. You may not have all of these fields.
717
718> cat /proc/meminfo
719
720The "Locked" indicates whether the mapping is locked in memory or not.
721
722
723MemTotal: 16344972 kB
724MemFree: 13634064 kB
725Buffers: 3656 kB
726Cached: 1195708 kB
727SwapCached: 0 kB
728Active: 891636 kB
729Inactive: 1077224 kB
730HighTotal: 15597528 kB
731HighFree: 13629632 kB
732LowTotal: 747444 kB
733LowFree: 4432 kB
734SwapTotal: 0 kB
735SwapFree: 0 kB
736Dirty: 968 kB
737Writeback: 0 kB
738AnonPages: 861800 kB
739Mapped: 280372 kB
740Slab: 284364 kB
741SReclaimable: 159856 kB
742SUnreclaim: 124508 kB
743PageTables: 24448 kB
744NFS_Unstable: 0 kB
745Bounce: 0 kB
746WritebackTmp: 0 kB
747CommitLimit: 7669796 kB
748Committed_AS: 100056 kB
749VmallocTotal: 112216 kB
750VmallocUsed: 428 kB
751VmallocChunk: 111088 kB
752AnonHugePages: 49152 kB
753
754 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
755 bits and the kernel binary code)
756 MemFree: The sum of LowFree+HighFree
757 Buffers: Relatively temporary storage for raw disk blocks
758 shouldn't get tremendously large (20MB or so)
759 Cached: in-memory cache for files read from the disk (the
760 pagecache). Doesn't include SwapCached
761 SwapCached: Memory that once was swapped out, is swapped back in but
762 still also is in the swapfile (if memory is needed it
763 doesn't need to be swapped out AGAIN because it is already
764 in the swapfile. This saves I/O)
765 Active: Memory that has been used more recently and usually not
766 reclaimed unless absolutely necessary.
767 Inactive: Memory which has been less recently used. It is more
768 eligible to be reclaimed for other purposes
769 HighTotal:
770 HighFree: Highmem is all memory above ~860MB of physical memory
771 Highmem areas are for use by userspace programs, or
772 for the pagecache. The kernel must use tricks to access
773 this memory, making it slower to access than lowmem.
774 LowTotal:
775 LowFree: Lowmem is memory which can be used for everything that
776 highmem can be used for, but it is also available for the
777 kernel's use for its own data structures. Among many
778 other things, it is where everything from the Slab is
779 allocated. Bad things happen when you're out of lowmem.
780 SwapTotal: total amount of swap space available
781 SwapFree: Memory which has been evicted from RAM, and is temporarily
782 on the disk
783 Dirty: Memory which is waiting to get written back to the disk
784 Writeback: Memory which is actively being written back to the disk
785 AnonPages: Non-file backed pages mapped into userspace page tables
786AnonHugePages: Non-file backed huge pages mapped into userspace page tables
787 Mapped: files which have been mmaped, such as libraries
788 Slab: in-kernel data structures cache
789SReclaimable: Part of Slab, that might be reclaimed, such as caches
790 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
791 PageTables: amount of memory dedicated to the lowest level of page
792 tables.
793NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
794 storage
795 Bounce: Memory used for block device "bounce buffers"
796WritebackTmp: Memory used by FUSE for temporary writeback buffers
797 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
798 this is the total amount of memory currently available to
799 be allocated on the system. This limit is only adhered to
800 if strict overcommit accounting is enabled (mode 2 in
801 'vm.overcommit_memory').
802 The CommitLimit is calculated with the following formula:
803 CommitLimit = ('vm.overcommit_ratio' * Physical RAM) + Swap
804 For example, on a system with 1G of physical RAM and 7G
805 of swap with a `vm.overcommit_ratio` of 30 it would
806 yield a CommitLimit of 7.3G.
807 For more details, see the memory overcommit documentation
808 in vm/overcommit-accounting.
809Committed_AS: The amount of memory presently allocated on the system.
810 The committed memory is a sum of all of the memory which
811 has been allocated by processes, even if it has not been
812 "used" by them as of yet. A process which malloc()'s 1G
813 of memory, but only touches 300M of it will only show up
814 as using 300M of memory even if it has the address space
815 allocated for the entire 1G. This 1G is memory which has
816 been "committed" to by the VM and can be used at any time
817 by the allocating application. With strict overcommit
818 enabled on the system (mode 2 in 'vm.overcommit_memory'),
819 allocations which would exceed the CommitLimit (detailed
820 above) will not be permitted. This is useful if one needs
821 to guarantee that processes will not fail due to lack of
822 memory once that memory has been successfully allocated.
823VmallocTotal: total size of vmalloc memory area
824 VmallocUsed: amount of vmalloc area which is used
825VmallocChunk: largest contiguous block of vmalloc area which is free
826
827..............................................................................
828
829vmallocinfo:
830
831Provides information about vmalloced/vmaped areas. One line per area,
832containing the virtual address range of the area, size in bytes,
833caller information of the creator, and optional information depending
834on the kind of area :
835
836 pages=nr number of pages
837 phys=addr if a physical address was specified
838 ioremap I/O mapping (ioremap() and friends)
839 vmalloc vmalloc() area
840 vmap vmap()ed pages
841 user VM_USERMAP area
842 vpages buffer for pages pointers was vmalloced (huge area)
843 N<node>=nr (Only on NUMA kernels)
844 Number of pages allocated on memory node <node>
845
846> cat /proc/vmallocinfo
8470xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
848 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
8490xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
850 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
8510xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
852 phys=7fee8000 ioremap
8530xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
854 phys=7fee7000 ioremap
8550xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
8560xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
857 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
8580xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
859 pages=2 vmalloc N1=2
8600xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
861 /0x130 [x_tables] pages=4 vmalloc N0=4
8620xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
863 pages=14 vmalloc N2=14
8640xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
865 pages=4 vmalloc N1=4
8660xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
867 pages=2 vmalloc N1=2
8680xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
869 pages=10 vmalloc N0=10
870
871..............................................................................
872
873softirqs:
874
875Provides counts of softirq handlers serviced since boot time, for each cpu.
876
877> cat /proc/softirqs
878 CPU0 CPU1 CPU2 CPU3
879 HI: 0 0 0 0
880 TIMER: 27166 27120 27097 27034
881 NET_TX: 0 0 0 17
882 NET_RX: 42 0 0 39
883 BLOCK: 0 0 107 1121
884 TASKLET: 0 0 0 290
885 SCHED: 27035 26983 26971 26746
886 HRTIMER: 0 0 0 0
887 RCU: 1678 1769 2178 2250
888
889
8901.3 IDE devices in /proc/ide
891----------------------------
892
893The subdirectory /proc/ide contains information about all IDE devices of which
894the kernel is aware. There is one subdirectory for each IDE controller, the
895file drivers and a link for each IDE device, pointing to the device directory
896in the controller specific subtree.
897
898The file drivers contains general information about the drivers used for the
899IDE devices:
900
901 > cat /proc/ide/drivers
902 ide-cdrom version 4.53
903 ide-disk version 1.08
904
905More detailed information can be found in the controller specific
906subdirectories. These are named ide0, ide1 and so on. Each of these
907directories contains the files shown in table 1-6.
908
909
910Table 1-6: IDE controller info in /proc/ide/ide?
911..............................................................................
912 File Content
913 channel IDE channel (0 or 1)
914 config Configuration (only for PCI/IDE bridge)
915 mate Mate name
916 model Type/Chipset of IDE controller
917..............................................................................
918
919Each device connected to a controller has a separate subdirectory in the
920controllers directory. The files listed in table 1-7 are contained in these
921directories.
922
923
924Table 1-7: IDE device information
925..............................................................................
926 File Content
927 cache The cache
928 capacity Capacity of the medium (in 512Byte blocks)
929 driver driver and version
930 geometry physical and logical geometry
931 identify device identify block
932 media media type
933 model device identifier
934 settings device setup
935 smart_thresholds IDE disk management thresholds
936 smart_values IDE disk management values
937..............................................................................
938
939The most interesting file is settings. This file contains a nice overview of
940the drive parameters:
941
942 # cat /proc/ide/ide0/hda/settings
943 name value min max mode
944 ---- ----- --- --- ----
945 bios_cyl 526 0 65535 rw
946 bios_head 255 0 255 rw
947 bios_sect 63 0 63 rw
948 breada_readahead 4 0 127 rw
949 bswap 0 0 1 r
950 file_readahead 72 0 2097151 rw
951 io_32bit 0 0 3 rw
952 keepsettings 0 0 1 rw
953 max_kb_per_request 122 1 127 rw
954 multcount 0 0 8 rw
955 nice1 1 0 1 rw
956 nowerr 0 0 1 rw
957 pio_mode write-only 0 255 w
958 slow 0 0 1 rw
959 unmaskirq 0 0 1 rw
960 using_dma 0 0 1 rw
961
962
9631.4 Networking info in /proc/net
964--------------------------------
965
966The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
967additional values you get for IP version 6 if you configure the kernel to
968support this. Table 1-9 lists the files and their meaning.
969
970
971Table 1-8: IPv6 info in /proc/net
972..............................................................................
973 File Content
974 udp6 UDP sockets (IPv6)
975 tcp6 TCP sockets (IPv6)
976 raw6 Raw device statistics (IPv6)
977 igmp6 IP multicast addresses, which this host joined (IPv6)
978 if_inet6 List of IPv6 interface addresses
979 ipv6_route Kernel routing table for IPv6
980 rt6_stats Global IPv6 routing tables statistics
981 sockstat6 Socket statistics (IPv6)
982 snmp6 Snmp data (IPv6)
983..............................................................................
984
985
986Table 1-9: Network info in /proc/net
987..............................................................................
988 File Content
989 arp Kernel ARP table
990 dev network devices with statistics
991 dev_mcast the Layer2 multicast groups a device is listening too
992 (interface index, label, number of references, number of bound
993 addresses).
994 dev_stat network device status
995 ip_fwchains Firewall chain linkage
996 ip_fwnames Firewall chain names
997 ip_masq Directory containing the masquerading tables
998 ip_masquerade Major masquerading table
999 netstat Network statistics
1000 raw raw device statistics
1001 route Kernel routing table
1002 rpc Directory containing rpc info
1003 rt_cache Routing cache
1004 snmp SNMP data
1005 sockstat Socket statistics
1006 tcp TCP sockets
1007 udp UDP sockets
1008 unix UNIX domain sockets
1009 wireless Wireless interface data (Wavelan etc)
1010 igmp IP multicast addresses, which this host joined
1011 psched Global packet scheduler parameters.
1012 netlink List of PF_NETLINK sockets
1013 ip_mr_vifs List of multicast virtual interfaces
1014 ip_mr_cache List of multicast routing cache
1015..............................................................................
1016
1017You can use this information to see which network devices are available in
1018your system and how much traffic was routed over those devices:
1019
1020 > cat /proc/net/dev
1021 Inter-|Receive |[...
1022 face |bytes packets errs drop fifo frame compressed multicast|[...
1023 lo: 908188 5596 0 0 0 0 0 0 [...
1024 ppp0:15475140 20721 410 0 0 410 0 0 [...
1025 eth0: 614530 7085 0 0 0 0 0 1 [...
1026
1027 ...] Transmit
1028 ...] bytes packets errs drop fifo colls carrier compressed
1029 ...] 908188 5596 0 0 0 0 0 0
1030 ...] 1375103 17405 0 0 0 0 0 0
1031 ...] 1703981 5535 0 0 0 3 0 0
1032
1033In addition, each Channel Bond interface has its own directory. For
1034example, the bond0 device will have a directory called /proc/net/bond0/.
1035It will contain information that is specific to that bond, such as the
1036current slaves of the bond, the link status of the slaves, and how
1037many times the slaves link has failed.
1038
10391.5 SCSI info
1040-------------
1041
1042If you have a SCSI host adapter in your system, you'll find a subdirectory
1043named after the driver for this adapter in /proc/scsi. You'll also see a list
1044of all recognized SCSI devices in /proc/scsi:
1045
1046 >cat /proc/scsi/scsi
1047 Attached devices:
1048 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1049 Vendor: IBM Model: DGHS09U Rev: 03E0
1050 Type: Direct-Access ANSI SCSI revision: 03
1051 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1052 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1053 Type: CD-ROM ANSI SCSI revision: 02
1054
1055
1056The directory named after the driver has one file for each adapter found in
1057the system. These files contain information about the controller, including
1058the used IRQ and the IO address range. The amount of information shown is
1059dependent on the adapter you use. The example shows the output for an Adaptec
1060AHA-2940 SCSI adapter:
1061
1062 > cat /proc/scsi/aic7xxx/0
1063
1064 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1065 Compile Options:
1066 TCQ Enabled By Default : Disabled
1067 AIC7XXX_PROC_STATS : Disabled
1068 AIC7XXX_RESET_DELAY : 5
1069 Adapter Configuration:
1070 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1071 Ultra Wide Controller
1072 PCI MMAPed I/O Base: 0xeb001000
1073 Adapter SEEPROM Config: SEEPROM found and used.
1074 Adaptec SCSI BIOS: Enabled
1075 IRQ: 10
1076 SCBs: Active 0, Max Active 2,
1077 Allocated 15, HW 16, Page 255
1078 Interrupts: 160328
1079 BIOS Control Word: 0x18b6
1080 Adapter Control Word: 0x005b
1081 Extended Translation: Enabled
1082 Disconnect Enable Flags: 0xffff
1083 Ultra Enable Flags: 0x0001
1084 Tag Queue Enable Flags: 0x0000
1085 Ordered Queue Tag Flags: 0x0000
1086 Default Tag Queue Depth: 8
1087 Tagged Queue By Device array for aic7xxx host instance 0:
1088 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1089 Actual queue depth per device for aic7xxx host instance 0:
1090 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1091 Statistics:
1092 (scsi0:0:0:0)
1093 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1094 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1095 Total transfers 160151 (74577 reads and 85574 writes)
1096 (scsi0:0:6:0)
1097 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1098 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1099 Total transfers 0 (0 reads and 0 writes)
1100
1101
11021.6 Parallel port info in /proc/parport
1103---------------------------------------
1104
1105The directory /proc/parport contains information about the parallel ports of
1106your system. It has one subdirectory for each port, named after the port
1107number (0,1,2,...).
1108
1109These directories contain the four files shown in Table 1-10.
1110
1111
1112Table 1-10: Files in /proc/parport
1113..............................................................................
1114 File Content
1115 autoprobe Any IEEE-1284 device ID information that has been acquired.
1116 devices list of the device drivers using that port. A + will appear by the
1117 name of the device currently using the port (it might not appear
1118 against any).
1119 hardware Parallel port's base address, IRQ line and DMA channel.
1120 irq IRQ that parport is using for that port. This is in a separate
1121 file to allow you to alter it by writing a new value in (IRQ
1122 number or none).
1123..............................................................................
1124
11251.7 TTY info in /proc/tty
1126-------------------------
1127
1128Information about the available and actually used tty's can be found in the
1129directory /proc/tty.You'll find entries for drivers and line disciplines in
1130this directory, as shown in Table 1-11.
1131
1132
1133Table 1-11: Files in /proc/tty
1134..............................................................................
1135 File Content
1136 drivers list of drivers and their usage
1137 ldiscs registered line disciplines
1138 driver/serial usage statistic and status of single tty lines
1139..............................................................................
1140
1141To see which tty's are currently in use, you can simply look into the file
1142/proc/tty/drivers:
1143
1144 > cat /proc/tty/drivers
1145 pty_slave /dev/pts 136 0-255 pty:slave
1146 pty_master /dev/ptm 128 0-255 pty:master
1147 pty_slave /dev/ttyp 3 0-255 pty:slave
1148 pty_master /dev/pty 2 0-255 pty:master
1149 serial /dev/cua 5 64-67 serial:callout
1150 serial /dev/ttyS 4 64-67 serial
1151 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1152 /dev/ptmx /dev/ptmx 5 2 system
1153 /dev/console /dev/console 5 1 system:console
1154 /dev/tty /dev/tty 5 0 system:/dev/tty
1155 unknown /dev/tty 4 1-63 console
1156
1157
11581.8 Miscellaneous kernel statistics in /proc/stat
1159-------------------------------------------------
1160
1161Various pieces of information about kernel activity are available in the
1162/proc/stat file. All of the numbers reported in this file are aggregates
1163since the system first booted. For a quick look, simply cat the file:
1164
1165 > cat /proc/stat
1166 cpu 2255 34 2290 22625563 6290 127 456 0 0
1167 cpu0 1132 34 1441 11311718 3675 127 438 0 0
1168 cpu1 1123 0 849 11313845 2614 0 18 0 0
1169 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1170 ctxt 1990473
1171 btime 1062191376
1172 processes 2915
1173 procs_running 1
1174 procs_blocked 0
1175 softirq 183433 0 21755 12 39 1137 231 21459 2263
1176
1177The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1178lines. These numbers identify the amount of time the CPU has spent performing
1179different kinds of work. Time units are in USER_HZ (typically hundredths of a
1180second). The meanings of the columns are as follows, from left to right:
1181
1182- user: normal processes executing in user mode
1183- nice: niced processes executing in user mode
1184- system: processes executing in kernel mode
1185- idle: twiddling thumbs
1186- iowait: waiting for I/O to complete
1187- irq: servicing interrupts
1188- softirq: servicing softirqs
1189- steal: involuntary wait
1190- guest: running a normal guest
1191- guest_nice: running a niced guest
1192
1193The "intr" line gives counts of interrupts serviced since boot time, for each
1194of the possible system interrupts. The first column is the total of all
1195interrupts serviced; each subsequent column is the total for that particular
1196interrupt.
1197
1198The "ctxt" line gives the total number of context switches across all CPUs.
1199
1200The "btime" line gives the time at which the system booted, in seconds since
1201the Unix epoch.
1202
1203The "processes" line gives the number of processes and threads created, which
1204includes (but is not limited to) those created by calls to the fork() and
1205clone() system calls.
1206
1207The "procs_running" line gives the total number of threads that are
1208running or ready to run (i.e., the total number of runnable threads).
1209
1210The "procs_blocked" line gives the number of processes currently blocked,
1211waiting for I/O to complete.
1212
1213The "softirq" line gives counts of softirqs serviced since boot time, for each
1214of the possible system softirqs. The first column is the total of all
1215softirqs serviced; each subsequent column is the total for that particular
1216softirq.
1217
1218
12191.9 Ext4 file system parameters
1220------------------------------
1221
1222Information about mounted ext4 file systems can be found in
1223/proc/fs/ext4. Each mounted filesystem will have a directory in
1224/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1225/proc/fs/ext4/dm-0). The files in each per-device directory are shown
1226in Table 1-12, below.
1227
1228Table 1-12: Files in /proc/fs/ext4/<devname>
1229..............................................................................
1230 File Content
1231 mb_groups details of multiblock allocator buddy cache of free blocks
1232..............................................................................
1233
12342.0 /proc/consoles
1235------------------
1236Shows registered system console lines.
1237
1238To see which character device lines are currently used for the system console
1239/dev/console, you may simply look into the file /proc/consoles:
1240
1241 > cat /proc/consoles
1242 tty0 -WU (ECp) 4:7
1243 ttyS0 -W- (Ep) 4:64
1244
1245The columns are:
1246
1247 device name of the device
1248 operations R = can do read operations
1249 W = can do write operations
1250 U = can do unblank
1251 flags E = it is enabled
1252 C = it is preferred console
1253 B = it is primary boot console
1254 p = it is used for printk buffer
1255 b = it is not a TTY but a Braille device
1256 a = it is safe to use when cpu is offline
1257 major:minor major and minor number of the device separated by a colon
1258
1259------------------------------------------------------------------------------
1260Summary
1261------------------------------------------------------------------------------
1262The /proc file system serves information about the running system. It not only
1263allows access to process data but also allows you to request the kernel status
1264by reading files in the hierarchy.
1265
1266The directory structure of /proc reflects the types of information and makes
1267it easy, if not obvious, where to look for specific data.
1268------------------------------------------------------------------------------
1269
1270------------------------------------------------------------------------------
1271CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1272------------------------------------------------------------------------------
1273
1274------------------------------------------------------------------------------
1275In This Chapter
1276------------------------------------------------------------------------------
1277* Modifying kernel parameters by writing into files found in /proc/sys
1278* Exploring the files which modify certain parameters
1279* Review of the /proc/sys file tree
1280------------------------------------------------------------------------------
1281
1282
1283A very interesting part of /proc is the directory /proc/sys. This is not only
1284a source of information, it also allows you to change parameters within the
1285kernel. Be very careful when attempting this. You can optimize your system,
1286but you can also cause it to crash. Never alter kernel parameters on a
1287production system. Set up a development machine and test to make sure that
1288everything works the way you want it to. You may have no alternative but to
1289reboot the machine once an error has been made.
1290
1291To change a value, simply echo the new value into the file. An example is
1292given below in the section on the file system data. You need to be root to do
1293this. You can create your own boot script to perform this every time your
1294system boots.
1295
1296The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1297general things in the operation of the Linux kernel. Since some of the files
1298can inadvertently disrupt your system, it is advisable to read both
1299documentation and source before actually making adjustments. In any case, be
1300very careful when writing to any of these files. The entries in /proc may
1301change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1302review the kernel documentation in the directory /usr/src/linux/Documentation.
1303This chapter is heavily based on the documentation included in the pre 2.2
1304kernels, and became part of it in version 2.2.1 of the Linux kernel.
1305
1306Please see: Documentation/sysctl/ directory for descriptions of these
1307entries.
1308
1309------------------------------------------------------------------------------
1310Summary
1311------------------------------------------------------------------------------
1312Certain aspects of kernel behavior can be modified at runtime, without the
1313need to recompile the kernel, or even to reboot the system. The files in the
1314/proc/sys tree can not only be read, but also modified. You can use the echo
1315command to write value into these files, thereby changing the default settings
1316of the kernel.
1317------------------------------------------------------------------------------
1318
1319------------------------------------------------------------------------------
1320CHAPTER 3: PER-PROCESS PARAMETERS
1321------------------------------------------------------------------------------
1322
13233.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1324--------------------------------------------------------------------------------
1325
1326These file can be used to adjust the badness heuristic used to select which
1327process gets killed in out of memory conditions.
1328
1329The badness heuristic assigns a value to each candidate task ranging from 0
1330(never kill) to 1000 (always kill) to determine which process is targeted. The
1331units are roughly a proportion along that range of allowed memory the process
1332may allocate from based on an estimation of its current memory and swap use.
1333For example, if a task is using all allowed memory, its badness score will be
13341000. If it is using half of its allowed memory, its score will be 500.
1335
1336There is an additional factor included in the badness score: root
1337processes are given 3% extra memory over other tasks.
1338
1339The amount of "allowed" memory depends on the context in which the oom killer
1340was called. If it is due to the memory assigned to the allocating task's cpuset
1341being exhausted, the allowed memory represents the set of mems assigned to that
1342cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1343memory represents the set of mempolicy nodes. If it is due to a memory
1344limit (or swap limit) being reached, the allowed memory is that configured
1345limit. Finally, if it is due to the entire system being out of memory, the
1346allowed memory represents all allocatable resources.
1347
1348The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1349is used to determine which task to kill. Acceptable values range from -1000
1350(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1351polarize the preference for oom killing either by always preferring a certain
1352task or completely disabling it. The lowest possible value, -1000, is
1353equivalent to disabling oom killing entirely for that task since it will always
1354report a badness score of 0.
1355
1356Consequently, it is very simple for userspace to define the amount of memory to
1357consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1358example, is roughly equivalent to allowing the remainder of tasks sharing the
1359same system, cpuset, mempolicy, or memory controller resources to use at least
136050% more memory. A value of -500, on the other hand, would be roughly
1361equivalent to discounting 50% of the task's allowed memory from being considered
1362as scoring against the task.
1363
1364For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1365be used to tune the badness score. Its acceptable values range from -16
1366(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1367(OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1368scaled linearly with /proc/<pid>/oom_score_adj.
1369
1370Writing to /proc/<pid>/oom_score_adj or /proc/<pid>/oom_adj will change the
1371other with its scaled value.
1372
1373The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1374value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1375requires CAP_SYS_RESOURCE.
1376
1377NOTICE: /proc/<pid>/oom_adj is deprecated and will be removed, please see
1378Documentation/feature-removal-schedule.txt.
1379
1380Caveat: when a parent task is selected, the oom killer will sacrifice any first
1381generation children with separate address spaces instead, if possible. This
1382avoids servers and important system daemons from being killed and loses the
1383minimal amount of work.
1384
1385
13863.2 /proc/<pid>/oom_score - Display current oom-killer score
1387-------------------------------------------------------------
1388
1389This file can be used to check the current score used by the oom-killer is for
1390any given <pid>. Use it together with /proc/<pid>/oom_adj to tune which
1391process should be killed in an out-of-memory situation.
1392
1393
13943.3 /proc/<pid>/io - Display the IO accounting fields
1395-------------------------------------------------------
1396
1397This file contains IO statistics for each running process
1398
1399Example
1400-------
1401
1402test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1403[1] 3828
1404
1405test:/tmp # cat /proc/3828/io
1406rchar: 323934931
1407wchar: 323929600
1408syscr: 632687
1409syscw: 632675
1410read_bytes: 0
1411write_bytes: 323932160
1412cancelled_write_bytes: 0
1413
1414
1415Description
1416-----------
1417
1418rchar
1419-----
1420
1421I/O counter: chars read
1422The number of bytes which this task has caused to be read from storage. This
1423is simply the sum of bytes which this process passed to read() and pread().
1424It includes things like tty IO and it is unaffected by whether or not actual
1425physical disk IO was required (the read might have been satisfied from
1426pagecache)
1427
1428
1429wchar
1430-----
1431
1432I/O counter: chars written
1433The number of bytes which this task has caused, or shall cause to be written
1434to disk. Similar caveats apply here as with rchar.
1435
1436
1437syscr
1438-----
1439
1440I/O counter: read syscalls
1441Attempt to count the number of read I/O operations, i.e. syscalls like read()
1442and pread().
1443
1444
1445syscw
1446-----
1447
1448I/O counter: write syscalls
1449Attempt to count the number of write I/O operations, i.e. syscalls like
1450write() and pwrite().
1451
1452
1453read_bytes
1454----------
1455
1456I/O counter: bytes read
1457Attempt to count the number of bytes which this process really did cause to
1458be fetched from the storage layer. Done at the submit_bio() level, so it is
1459accurate for block-backed filesystems. <please add status regarding NFS and
1460CIFS at a later time>
1461
1462
1463write_bytes
1464-----------
1465
1466I/O counter: bytes written
1467Attempt to count the number of bytes which this process caused to be sent to
1468the storage layer. This is done at page-dirtying time.
1469
1470
1471cancelled_write_bytes
1472---------------------
1473
1474The big inaccuracy here is truncate. If a process writes 1MB to a file and
1475then deletes the file, it will in fact perform no writeout. But it will have
1476been accounted as having caused 1MB of write.
1477In other words: The number of bytes which this process caused to not happen,
1478by truncating pagecache. A task can cause "negative" IO too. If this task
1479truncates some dirty pagecache, some IO which another task has been accounted
1480for (in its write_bytes) will not be happening. We _could_ just subtract that
1481from the truncating task's write_bytes, but there is information loss in doing
1482that.
1483
1484
1485Note
1486----
1487
1488At its current implementation state, this is a bit racy on 32-bit machines: if
1489process A reads process B's /proc/pid/io while process B is updating one of
1490those 64-bit counters, process A could see an intermediate result.
1491
1492
1493More information about this can be found within the taskstats documentation in
1494Documentation/accounting.
1495
14963.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1497---------------------------------------------------------------
1498When a process is dumped, all anonymous memory is written to a core file as
1499long as the size of the core file isn't limited. But sometimes we don't want
1500to dump some memory segments, for example, huge shared memory. Conversely,
1501sometimes we want to save file-backed memory segments into a core file, not
1502only the individual files.
1503
1504/proc/<pid>/coredump_filter allows you to customize which memory segments
1505will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1506of memory types. If a bit of the bitmask is set, memory segments of the
1507corresponding memory type are dumped, otherwise they are not dumped.
1508
1509The following 7 memory types are supported:
1510 - (bit 0) anonymous private memory
1511 - (bit 1) anonymous shared memory
1512 - (bit 2) file-backed private memory
1513 - (bit 3) file-backed shared memory
1514 - (bit 4) ELF header pages in file-backed private memory areas (it is
1515 effective only if the bit 2 is cleared)
1516 - (bit 5) hugetlb private memory
1517 - (bit 6) hugetlb shared memory
1518
1519 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1520 are always dumped regardless of the bitmask status.
1521
1522 Note bit 0-4 doesn't effect any hugetlb memory. hugetlb memory are only
1523 effected by bit 5-6.
1524
1525Default value of coredump_filter is 0x23; this means all anonymous memory
1526segments and hugetlb private memory are dumped.
1527
1528If you don't want to dump all shared memory segments attached to pid 1234,
1529write 0x21 to the process's proc file.
1530
1531 $ echo 0x21 > /proc/1234/coredump_filter
1532
1533When a new process is created, the process inherits the bitmask status from its
1534parent. It is useful to set up coredump_filter before the program runs.
1535For example:
1536
1537 $ echo 0x7 > /proc/self/coredump_filter
1538 $ ./some_program
1539
15403.5 /proc/<pid>/mountinfo - Information about mounts
1541--------------------------------------------------------
1542
1543This file contains lines of the form:
1544
154536 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1546(1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1547
1548(1) mount ID: unique identifier of the mount (may be reused after umount)
1549(2) parent ID: ID of parent (or of self for the top of the mount tree)
1550(3) major:minor: value of st_dev for files on filesystem
1551(4) root: root of the mount within the filesystem
1552(5) mount point: mount point relative to the process's root
1553(6) mount options: per mount options
1554(7) optional fields: zero or more fields of the form "tag[:value]"
1555(8) separator: marks the end of the optional fields
1556(9) filesystem type: name of filesystem of the form "type[.subtype]"
1557(10) mount source: filesystem specific information or "none"
1558(11) super options: per super block options
1559
1560Parsers should ignore all unrecognised optional fields. Currently the
1561possible optional fields are:
1562
1563shared:X mount is shared in peer group X
1564master:X mount is slave to peer group X
1565propagate_from:X mount is slave and receives propagation from peer group X (*)
1566unbindable mount is unbindable
1567
1568(*) X is the closest dominant peer group under the process's root. If
1569X is the immediate master of the mount, or if there's no dominant peer
1570group under the same root, then only the "master:X" field is present
1571and not the "propagate_from:X" field.
1572
1573For more information on mount propagation see:
1574
1575 Documentation/filesystems/sharedsubtree.txt
1576
1577
15783.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1579--------------------------------------------------------
1580These files provide a method to access a tasks comm value. It also allows for
1581a task to set its own or one of its thread siblings comm value. The comm value
1582is limited in size compared to the cmdline value, so writing anything longer
1583then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1584comm value.
1585
1586
15873.7 /proc/<pid>/task/<tid>/children - Information about task children
1588-------------------------------------------------------------------------
1589This file provides a fast way to retrieve first level children pids
1590of a task pointed by <pid>/<tid> pair. The format is a space separated
1591stream of pids.
1592
1593Note the "first level" here -- if a child has own children they will
1594not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1595to obtain the descendants.
1596
1597Since this interface is intended to be fast and cheap it doesn't
1598guarantee to provide precise results and some children might be
1599skipped, especially if they've exited right after we printed their
1600pids, so one need to either stop or freeze processes being inspected
1601if precise results are needed.
1602
1603
1604------------------------------------------------------------------------------
1605Configuring procfs
1606------------------------------------------------------------------------------
1607
16084.1 Mount options
1609---------------------
1610
1611The following mount options are supported:
1612
1613 hidepid= Set /proc/<pid>/ access mode.
1614 gid= Set the group authorized to learn processes information.
1615
1616hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
1617(default).
1618
1619hidepid=1 means users may not access any /proc/<pid>/ directories but their
1620own. Sensitive files like cmdline, sched*, status are now protected against
1621other users. This makes it impossible to learn whether any user runs
1622specific program (given the program doesn't reveal itself by its behaviour).
1623As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
1624poorly written programs passing sensitive information via program arguments are
1625now protected against local eavesdroppers.
1626
1627hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
1628users. It doesn't mean that it hides a fact whether a process with a specific
1629pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
1630but it hides process' uid and gid, which may be learned by stat()'ing
1631/proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
1632information about running processes, whether some daemon runs with elevated
1633privileges, whether other user runs some sensitive program, whether other users
1634run any program at all, etc.
1635
1636gid= defines a group authorized to learn processes information otherwise
1637prohibited by hidepid=. If you use some daemon like identd which needs to learn
1638information about processes information, just add identd to this group.