1 ------------------------------------------------------------------------------
2 T H E /proc F I L E S Y S T E M
3 ------------------------------------------------------------------------------
4 /proc/sys Terrehon Bowden <terrehon@pacbell.net> October 7 1999
5 Bodo Bauer <bb@ricochet.net>
7 2.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000
8 move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009
9 ------------------------------------------------------------------------------
10 Version 1.3 Kernel version 2.2.12
11 Kernel version 2.4.0-test11-pre4
12 ------------------------------------------------------------------------------
13 fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009
19 0.1 Introduction/Credits
22 1 Collecting System Information
23 1.1 Process-Specific Subdirectories
25 1.3 IDE devices in /proc/ide
26 1.4 Networking info in /proc/net
28 1.6 Parallel port info in /proc/parport
29 1.7 TTY info in /proc/tty
30 1.8 Miscellaneous kernel statistics in /proc/stat
31 1.9 Ext4 file system parameters
33 2 Modifying System Parameters
35 3 Per-Process Parameters
36 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer
38 3.2 /proc/<pid>/oom_score - Display current oom-killer score
39 3.3 /proc/<pid>/io - Display the IO accounting fields
40 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
41 3.5 /proc/<pid>/mountinfo - Information about mounts
42 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
43 3.7 /proc/<pid>/task/<tid>/children - Information about task children
44 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
45 3.9 /proc/<pid>/map_files - Information about memory mapped files
46 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
47 3.11 /proc/<pid>/patch_state - Livepatch patch operation state
48 3.12 /proc/<pid>/arch_status - Task architecture specific information
53 ------------------------------------------------------------------------------
55 ------------------------------------------------------------------------------
57 0.1 Introduction/Credits
58 ------------------------
60 This documentation is part of a soon (or so we hope) to be released book on
61 the SuSE Linux distribution. As there is no complete documentation for the
62 /proc file system and we've used many freely available sources to write these
63 chapters, it seems only fair to give the work back to the Linux community.
64 This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm
65 afraid it's still far from complete, but we hope it will be useful. As far as
66 we know, it is the first 'all-in-one' document about the /proc file system. It
67 is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,
68 SPARC, AXP, etc., features, you probably won't find what you are looking for.
69 It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But
70 additions and patches are welcome and will be added to this document if you
73 We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of
74 other people for help compiling this documentation. We'd also like to extend a
75 special thank you to Andi Kleen for documentation, which we relied on heavily
76 to create this document, as well as the additional information he provided.
77 Thanks to everybody else who contributed source or docs to the Linux kernel
78 and helped create a great piece of software... :)
80 If you have any comments, corrections or additions, please don't hesitate to
81 contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this
84 The latest version of this document is available online at
85 http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html
87 If the above direction does not works for you, you could try the kernel
88 mailing list at linux-kernel@vger.kernel.org and/or try to reach me at
89 comandante@zaralinux.com.
94 We don't guarantee the correctness of this document, and if you come to us
95 complaining about how you screwed up your system because of incorrect
96 documentation, we won't feel responsible...
98 ------------------------------------------------------------------------------
99 CHAPTER 1: COLLECTING SYSTEM INFORMATION
100 ------------------------------------------------------------------------------
102 ------------------------------------------------------------------------------
104 ------------------------------------------------------------------------------
105 * Investigating the properties of the pseudo file system /proc and its
106 ability to provide information on the running Linux system
107 * Examining /proc's structure
108 * Uncovering various information about the kernel and the processes running
110 ------------------------------------------------------------------------------
113 The proc file system acts as an interface to internal data structures in the
114 kernel. It can be used to obtain information about the system and to change
115 certain kernel parameters at runtime (sysctl).
117 First, we'll take a look at the read-only parts of /proc. In Chapter 2, we
118 show you how you can use /proc/sys to change settings.
120 1.1 Process-Specific Subdirectories
121 -----------------------------------
123 The directory /proc contains (among other things) one subdirectory for each
124 process running on the system, which is named after the process ID (PID).
126 The link self points to the process reading the file system. Each process
127 subdirectory has the entries listed in Table 1-1.
129 Note that an open a file descriptor to /proc/<pid> or to any of its
130 contained files or subdirectories does not prevent <pid> being reused
131 for some other process in the event that <pid> exits. Operations on
132 open /proc/<pid> file descriptors corresponding to dead processes
133 never act on any new process that the kernel may, through chance, have
134 also assigned the process ID <pid>. Instead, operations on these FDs
135 usually fail with ESRCH.
137 Table 1-1: Process specific entries in /proc
138 ..............................................................................
140 clear_refs Clears page referenced bits shown in smaps output
141 cmdline Command line arguments
142 cpu Current and last cpu in which it was executed (2.4)(smp)
143 cwd Link to the current working directory
144 environ Values of environment variables
145 exe Link to the executable of this process
146 fd Directory, which contains all file descriptors
147 maps Memory maps to executables and library files (2.4)
148 mem Memory held by this process
149 root Link to the root directory of this process
151 statm Process memory status information
152 status Process status in human readable form
153 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function
154 symbol the task is blocked in - or "0" if not blocked.
156 stack Report full stack trace, enable via CONFIG_STACKTRACE
157 smaps an extension based on maps, showing the memory consumption of
158 each mapping and flags associated with it
159 numa_maps an extension based on maps, showing the memory locality and
160 binding policy as well as mem usage (in pages) of each mapping.
161 ..............................................................................
163 For example, to get the status information of a process, all you have to do is
164 read the file /proc/PID/status:
166 >cat /proc/self/status
196 SigPnd: 0000000000000000
197 ShdPnd: 0000000000000000
198 SigBlk: 0000000000000000
199 SigIgn: 0000000000000000
200 SigCgt: 0000000000000000
201 CapInh: 00000000fffffeff
202 CapPrm: 0000000000000000
203 CapEff: 0000000000000000
204 CapBnd: ffffffffffffffff
205 CapAmb: 0000000000000000
208 Speculation_Store_Bypass: thread vulnerable
209 voluntary_ctxt_switches: 0
210 nonvoluntary_ctxt_switches: 1
212 This shows you nearly the same information you would get if you viewed it with
213 the ps command. In fact, ps uses the proc file system to obtain its
214 information. But you get a more detailed view of the process by reading the
215 file /proc/PID/status. It fields are described in table 1-2.
217 The statm file contains more detailed information about the process
218 memory usage. Its seven fields are explained in Table 1-3. The stat file
219 contains details information about the process itself. Its fields are
220 explained in Table 1-4.
222 (for SMP CONFIG users)
223 For making accounting scalable, RSS related information are handled in an
224 asynchronous manner and the value may not be very precise. To see a precise
225 snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.
226 It's slow but very precise.
228 Table 1-2: Contents of the status files (as of 4.19)
229 ..............................................................................
231 Name filename of the executable
232 Umask file mode creation mask
233 State state (R is running, S is sleeping, D is sleeping
234 in an uninterruptible wait, Z is zombie,
235 T is traced or stopped)
237 Ngid NUMA group ID (0 if none)
239 PPid process id of the parent process
240 TracerPid PID of process tracing this process (0 if not)
241 Uid Real, effective, saved set, and file system UIDs
242 Gid Real, effective, saved set, and file system GIDs
243 FDSize number of file descriptor slots currently allocated
244 Groups supplementary group list
245 NStgid descendant namespace thread group ID hierarchy
246 NSpid descendant namespace process ID hierarchy
247 NSpgid descendant namespace process group ID hierarchy
248 NSsid descendant namespace session ID hierarchy
249 VmPeak peak virtual memory size
250 VmSize total program size
251 VmLck locked memory size
252 VmPin pinned memory size
253 VmHWM peak resident set size ("high water mark")
254 VmRSS size of memory portions. It contains the three
255 following parts (VmRSS = RssAnon + RssFile + RssShmem)
256 RssAnon size of resident anonymous memory
257 RssFile size of resident file mappings
258 RssShmem size of resident shmem memory (includes SysV shm,
259 mapping of tmpfs and shared anonymous mappings)
260 VmData size of private data segments
261 VmStk size of stack segments
262 VmExe size of text segment
263 VmLib size of shared library code
264 VmPTE size of page table entries
265 VmSwap amount of swap used by anonymous private data
266 (shmem swap usage is not included)
267 HugetlbPages size of hugetlb memory portions
268 CoreDumping process's memory is currently being dumped
269 (killing the process may lead to a corrupted core)
270 THP_enabled process is allowed to use THP (returns 0 when
271 PR_SET_THP_DISABLE is set on the process
272 Threads number of threads
273 SigQ number of signals queued/max. number for queue
274 SigPnd bitmap of pending signals for the thread
275 ShdPnd bitmap of shared pending signals for the process
276 SigBlk bitmap of blocked signals
277 SigIgn bitmap of ignored signals
278 SigCgt bitmap of caught signals
279 CapInh bitmap of inheritable capabilities
280 CapPrm bitmap of permitted capabilities
281 CapEff bitmap of effective capabilities
282 CapBnd bitmap of capabilities bounding set
283 CapAmb bitmap of ambient capabilities
284 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...)
285 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)
286 Speculation_Store_Bypass speculative store bypass mitigation status
287 Cpus_allowed mask of CPUs on which this process may run
288 Cpus_allowed_list Same as previous, but in "list format"
289 Mems_allowed mask of memory nodes allowed to this process
290 Mems_allowed_list Same as previous, but in "list format"
291 voluntary_ctxt_switches number of voluntary context switches
292 nonvoluntary_ctxt_switches number of non voluntary context switches
293 ..............................................................................
295 Table 1-3: Contents of the statm files (as of 2.6.8-rc3)
296 ..............................................................................
298 size total program size (pages) (same as VmSize in status)
299 resident size of memory portions (pages) (same as VmRSS in status)
300 shared number of pages that are shared (i.e. backed by a file, same
301 as RssFile+RssShmem in status)
302 trs number of pages that are 'code' (not including libs; broken,
303 includes data segment)
304 lrs number of pages of library (always 0 on 2.6)
305 drs number of pages of data/stack (including libs; broken,
306 includes library text)
307 dt number of dirty pages (always 0 on 2.6)
308 ..............................................................................
311 Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
312 ..............................................................................
315 tcomm filename of the executable
316 state state (R is running, S is sleeping, D is sleeping in an
317 uninterruptible wait, Z is zombie, T is traced or stopped)
318 ppid process id of the parent process
319 pgrp pgrp of the process
321 tty_nr tty the process uses
322 tty_pgrp pgrp of the tty
324 min_flt number of minor faults
325 cmin_flt number of minor faults with child's
326 maj_flt number of major faults
327 cmaj_flt number of major faults with child's
328 utime user mode jiffies
329 stime kernel mode jiffies
330 cutime user mode jiffies with child's
331 cstime kernel mode jiffies with child's
332 priority priority level
334 num_threads number of threads
335 it_real_value (obsolete, always 0)
336 start_time time the process started after system boot
337 vsize virtual memory size
338 rss resident set memory size
339 rsslim current limit in bytes on the rss
340 start_code address above which program text can run
341 end_code address below which program text can run
342 start_stack address of the start of the main process stack
343 esp current value of ESP
344 eip current value of EIP
345 pending bitmap of pending signals
346 blocked bitmap of blocked signals
347 sigign bitmap of ignored signals
348 sigcatch bitmap of caught signals
349 0 (place holder, used to be the wchan address, use /proc/PID/wchan instead)
352 exit_signal signal to send to parent thread on exit
353 task_cpu which CPU the task is scheduled on
354 rt_priority realtime priority
355 policy scheduling policy (man sched_setscheduler)
356 blkio_ticks time spent waiting for block IO
357 gtime guest time of the task in jiffies
358 cgtime guest time of the task children in jiffies
359 start_data address above which program data+bss is placed
360 end_data address below which program data+bss is placed
361 start_brk address above which program heap can be expanded with brk()
362 arg_start address above which program command line is placed
363 arg_end address below which program command line is placed
364 env_start address above which program environment is placed
365 env_end address below which program environment is placed
366 exit_code the thread's exit_code in the form reported by the waitpid system call
367 ..............................................................................
369 The /proc/PID/maps file containing the currently mapped memory regions and
370 their access permissions.
374 address perms offset dev inode pathname
376 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
377 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test
378 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]
379 a7cb1000-a7cb2000 ---p 00000000 00:00 0
380 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
381 a7eb2000-a7eb3000 ---p 00000000 00:00 0
382 a7eb3000-a7ed5000 rw-p 00000000 00:00 0
383 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
384 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
385 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
386 a800b000-a800e000 rw-p 00000000 00:00 0
387 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0
388 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0
389 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0
390 a8024000-a8027000 rw-p 00000000 00:00 0
391 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2
392 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2
393 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2
394 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]
395 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]
397 where "address" is the address space in the process that it occupies, "perms"
398 is a set of permissions:
404 p = private (copy on write)
406 "offset" is the offset into the mapping, "dev" is the device (major:minor), and
407 "inode" is the inode on that device. 0 indicates that no inode is associated
408 with the memory region, as the case would be with BSS (uninitialized data).
409 The "pathname" shows the name associated file for this mapping. If the mapping
410 is not associated with a file:
412 [heap] = the heap of the program
413 [stack] = the stack of the main process
414 [vdso] = the "virtual dynamic shared object",
415 the kernel system call handler
417 or if empty, the mapping is anonymous.
419 The /proc/PID/smaps is an extension based on maps, showing the memory
420 consumption for each of the process's mappings. For each of mappings there
421 is a series of lines such as the following:
423 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash
437 Private_Hugetlb: 0 kB
444 VmFlags: rd ex mr mw me dw
446 the first of these lines shows the same information as is displayed for the
447 mapping in /proc/PID/maps. The remaining lines show the size of the mapping
448 (size), the amount of the mapping that is currently resident in RAM (RSS), the
449 process' proportional share of this mapping (PSS), the number of clean and
450 dirty private pages in the mapping.
452 The "proportional set size" (PSS) of a process is the count of pages it has
453 in memory, where each page is divided by the number of processes sharing it.
454 So if a process has 1000 pages all to itself, and 1000 shared with one other
455 process, its PSS will be 1500.
456 Note that even a page which is part of a MAP_SHARED mapping, but has only
457 a single pte mapped, i.e. is currently used by only one process, is accounted
458 as private and not as shared.
459 "Referenced" indicates the amount of memory currently marked as referenced or
461 "Anonymous" shows the amount of memory that does not belong to any file. Even
462 a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE
463 and a page is modified, the file page is replaced by a private anonymous copy.
464 "LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE).
465 The memory isn't freed immediately with madvise(). It's freed in memory
466 pressure if the memory is clean. Please note that the printed value might
467 be lower than the real value due to optimizations used in the current
468 implementation. If this is not desirable please file a bug report.
469 "AnonHugePages" shows the ammount of memory backed by transparent hugepage.
470 "ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by
472 "Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by
473 hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical
474 reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.
475 "Swap" shows how much would-be-anonymous memory is also used, but out on swap.
476 For shmem mappings, "Swap" includes also the size of the mapped (and not
477 replaced by copy-on-write) part of the underlying shmem object out on swap.
478 "SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this
479 does not take into account swapped out page of underlying shmem objects.
480 "Locked" indicates whether the mapping is locked in memory or not.
481 "THPeligible" indicates whether the mapping is eligible for THP pages - 1 if
484 "VmFlags" field deserves a separate description. This member represents the kernel
485 flags associated with the particular virtual memory area in two letter encoded
486 manner. The codes are the following:
495 gd - stack segment growns down
497 dw - disabled write to the mapped file
498 lo - pages are locked in memory
499 io - memory mapped I/O area
500 sr - sequential read advise provided
501 rr - random read advise provided
502 dc - do not copy area on fork
503 de - do not expand area on remapping
504 ac - area is accountable
505 nr - swap space is not reserved for the area
506 ht - area uses huge tlb pages
507 ar - architecture specific flag
508 dd - do not include area into core dump
511 hg - huge page advise flag
512 nh - no-huge page advise flag
513 mg - mergable advise flag
515 Note that there is no guarantee that every flag and associated mnemonic will
516 be present in all further kernel releases. Things get changed, the flags may
517 be vanished or the reverse -- new added. Interpretation of their meaning
518 might change in future as well. So each consumer of these flags has to
519 follow each specific kernel version for the exact semantic.
521 This file is only present if the CONFIG_MMU kernel configuration option is
524 Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent
525 output can be achieved only in the single read call).
526 This typically manifests when doing partial reads of these files while the
527 memory map is being modified. Despite the races, we do provide the following
530 1) The mapped addresses never go backwards, which implies no two
531 regions will ever overlap.
532 2) If there is something at a given vaddr during the entirety of the
533 life of the smaps/maps walk, there will be some output for it.
536 The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG
537 bits on both physical and virtual pages associated with a process, and the
538 soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst
540 To clear the bits for all the pages associated with the process
541 > echo 1 > /proc/PID/clear_refs
543 To clear the bits for the anonymous pages associated with the process
544 > echo 2 > /proc/PID/clear_refs
546 To clear the bits for the file mapped pages associated with the process
547 > echo 3 > /proc/PID/clear_refs
549 To clear the soft-dirty bit
550 > echo 4 > /proc/PID/clear_refs
552 To reset the peak resident set size ("high water mark") to the process's
554 > echo 5 > /proc/PID/clear_refs
556 Any other value written to /proc/PID/clear_refs will have no effect.
558 The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags
559 using /proc/kpageflags and number of times a page is mapped using
560 /proc/kpagecount. For detailed explanation, see
561 Documentation/admin-guide/mm/pagemap.rst.
563 The /proc/pid/numa_maps is an extension based on maps, showing the memory
564 locality and binding policy, as well as the memory usage (in pages) of
565 each mapping. The output follows a general format where mapping details get
566 summarized separated by blank spaces, one mapping per each file line:
568 address policy mapping details
570 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4
571 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4
572 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4
573 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
574 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
575 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4
576 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4
577 320698b000 default file=/lib64/libc-2.12.so
578 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4
579 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4
580 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4
581 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4
582 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4
583 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048
584 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4
585 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4
588 "address" is the starting address for the mapping;
589 "policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst);
590 "mapping details" summarizes mapping data such as mapping type, page usage counters,
591 node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page
592 size, in KB, that is backing the mapping up.
597 Similar to the process entries, the kernel data files give information about
598 the running kernel. The files used to obtain this information are contained in
599 /proc and are listed in Table 1-5. Not all of these will be present in your
600 system. It depends on the kernel configuration and the loaded modules, which
601 files are there, and which are missing.
603 Table 1-5: Kernel info in /proc
604 ..............................................................................
606 apm Advanced power management info
607 buddyinfo Kernel memory allocator information (see text) (2.5)
608 bus Directory containing bus specific information
609 cmdline Kernel command line
610 cpuinfo Info about the CPU
611 devices Available devices (block and character)
612 dma Used DMS channels
613 filesystems Supported filesystems
614 driver Various drivers grouped here, currently rtc (2.4)
615 execdomains Execdomains, related to security (2.4)
616 fb Frame Buffer devices (2.4)
617 fs File system parameters, currently nfs/exports (2.4)
618 ide Directory containing info about the IDE subsystem
619 interrupts Interrupt usage
620 iomem Memory map (2.4)
621 ioports I/O port usage
622 irq Masks for irq to cpu affinity (2.4)(smp?)
623 isapnp ISA PnP (Plug&Play) Info (2.4)
624 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))
626 ksyms Kernel symbol table
627 loadavg Load average of last 1, 5 & 15 minutes
631 modules List of loaded modules
632 mounts Mounted filesystems
633 net Networking info (see text)
634 pagetypeinfo Additional page allocator information (see text) (2.5)
635 partitions Table of partitions known to the system
636 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,
637 decoupled by lspci (2.4)
639 scsi SCSI info (see text)
640 slabinfo Slab pool info
641 softirqs softirq usage
642 stat Overall statistics
643 swaps Swap space utilization
645 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)
646 tty Info of tty drivers
647 uptime Wall clock since boot, combined idle time of all cpus
648 version Kernel version
649 video bttv info of video resources (2.4)
650 vmallocinfo Show vmalloced areas
651 ..............................................................................
653 You can, for example, check which interrupts are currently in use and what
654 they are used for by looking in the file /proc/interrupts:
656 > cat /proc/interrupts
658 0: 8728810 XT-PIC timer
659 1: 895 XT-PIC keyboard
661 3: 531695 XT-PIC aha152x
662 4: 2014133 XT-PIC serial
663 5: 44401 XT-PIC pcnet_cs
666 12: 182918 XT-PIC PS/2 Mouse
668 14: 1232265 XT-PIC ide0
672 In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the
673 output of a SMP machine):
675 > cat /proc/interrupts
678 0: 1243498 1214548 IO-APIC-edge timer
679 1: 8949 8958 IO-APIC-edge keyboard
680 2: 0 0 XT-PIC cascade
681 5: 11286 10161 IO-APIC-edge soundblaster
682 8: 1 0 IO-APIC-edge rtc
683 9: 27422 27407 IO-APIC-edge 3c503
684 12: 113645 113873 IO-APIC-edge PS/2 Mouse
686 14: 22491 24012 IO-APIC-edge ide0
687 15: 2183 2415 IO-APIC-edge ide1
688 17: 30564 30414 IO-APIC-level eth0
689 18: 177 164 IO-APIC-level bttv
694 NMI is incremented in this case because every timer interrupt generates a NMI
695 (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.
697 LOC is the local interrupt counter of the internal APIC of every CPU.
699 ERR is incremented in the case of errors in the IO-APIC bus (the bus that
700 connects the CPUs in a SMP system. This means that an error has been detected,
701 the IO-APIC automatically retry the transmission, so it should not be a big
702 problem, but you should read the SMP-FAQ.
704 In 2.6.2* /proc/interrupts was expanded again. This time the goal was for
705 /proc/interrupts to display every IRQ vector in use by the system, not
706 just those considered 'most important'. The new vectors are:
708 THR -- interrupt raised when a machine check threshold counter
709 (typically counting ECC corrected errors of memory or cache) exceeds
710 a configurable threshold. Only available on some systems.
712 TRM -- a thermal event interrupt occurs when a temperature threshold
713 has been exceeded for the CPU. This interrupt may also be generated
714 when the temperature drops back to normal.
716 SPU -- a spurious interrupt is some interrupt that was raised then lowered
717 by some IO device before it could be fully processed by the APIC. Hence
718 the APIC sees the interrupt but does not know what device it came from.
719 For this case the APIC will generate the interrupt with a IRQ vector
720 of 0xff. This might also be generated by chipset bugs.
722 RES, CAL, TLB -- rescheduling, call and TLB flush interrupts are
723 sent from one CPU to another per the needs of the OS. Typically,
724 their statistics are used by kernel developers and interested users to
725 determine the occurrence of interrupts of the given type.
727 The above IRQ vectors are displayed only when relevant. For example,
728 the threshold vector does not exist on x86_64 platforms. Others are
729 suppressed when the system is a uniprocessor. As of this writing, only
730 i386 and x86_64 platforms support the new IRQ vector displays.
732 Of some interest is the introduction of the /proc/irq directory to 2.4.
733 It could be used to set IRQ to CPU affinity, this means that you can "hook" an
734 IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the
735 irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and
740 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask
741 1 11 13 15 17 19 3 5 7 9 default_smp_affinity
745 smp_affinity is a bitmask, in which you can specify which CPUs can handle the
746 IRQ, you can set it by doing:
748 > echo 1 > /proc/irq/10/smp_affinity
750 This means that only the first CPU will handle the IRQ, but you can also echo
751 5 which means that only the first and third CPU can handle the IRQ.
753 The contents of each smp_affinity file is the same by default:
755 > cat /proc/irq/0/smp_affinity
758 There is an alternate interface, smp_affinity_list which allows specifying
759 a cpu range instead of a bitmask:
761 > cat /proc/irq/0/smp_affinity_list
764 The default_smp_affinity mask applies to all non-active IRQs, which are the
765 IRQs which have not yet been allocated/activated, and hence which lack a
766 /proc/irq/[0-9]* directory.
768 The node file on an SMP system shows the node to which the device using the IRQ
769 reports itself as being attached. This hardware locality information does not
770 include information about any possible driver locality preference.
772 prof_cpu_mask specifies which CPUs are to be profiled by the system wide
773 profiler. Default value is ffffffff (all cpus if there are only 32 of them).
775 The way IRQs are routed is handled by the IO-APIC, and it's Round Robin
776 between all the CPUs which are allowed to handle it. As usual the kernel has
777 more info than you and does a better job than you, so the defaults are the
778 best choice for almost everyone. [Note this applies only to those IO-APIC's
779 that support "Round Robin" interrupt distribution.]
781 There are three more important subdirectories in /proc: net, scsi, and sys.
782 The general rule is that the contents, or even the existence of these
783 directories, depend on your kernel configuration. If SCSI is not enabled, the
784 directory scsi may not exist. The same is true with the net, which is there
785 only when networking support is present in the running kernel.
787 The slabinfo file gives information about memory usage at the slab level.
788 Linux uses slab pools for memory management above page level in version 2.2.
789 Commonly used objects have their own slab pool (such as network buffers,
790 directory cache, and so on).
792 ..............................................................................
794 > cat /proc/buddyinfo
796 Node 0, zone DMA 0 4 5 4 4 3 ...
797 Node 0, zone Normal 1 0 0 1 101 8 ...
798 Node 0, zone HighMem 2 0 0 1 1 0 ...
800 External fragmentation is a problem under some workloads, and buddyinfo is a
801 useful tool for helping diagnose these problems. Buddyinfo will give you a
802 clue as to how big an area you can safely allocate, or why a previous
805 Each column represents the number of pages of a certain order which are
806 available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in
807 ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE
808 available in ZONE_NORMAL, etc...
810 More information relevant to external fragmentation can be found in
813 > cat /proc/pagetypeinfo
817 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10
818 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0
819 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0
820 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2
821 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0
822 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0
823 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9
824 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0
825 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452
826 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0
827 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0
829 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate
830 Node 0, zone DMA 2 0 5 1 0
831 Node 0, zone DMA32 41 6 967 2 0
833 Fragmentation avoidance in the kernel works by grouping pages of different
834 migrate types into the same contiguous regions of memory called page blocks.
835 A page block is typically the size of the default hugepage size e.g. 2MB on
836 X86-64. By keeping pages grouped based on their ability to move, the kernel
837 can reclaim pages within a page block to satisfy a high-order allocation.
839 The pagetypinfo begins with information on the size of a page block. It
840 then gives the same type of information as buddyinfo except broken down
841 by migrate-type and finishes with details on how many page blocks of each
844 If min_free_kbytes has been tuned correctly (recommendations made by hugeadm
845 from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can
846 make an estimate of the likely number of huge pages that can be allocated
847 at a given point in time. All the "Movable" blocks should be allocatable
848 unless memory has been mlock()'d. Some of the Reclaimable blocks should
849 also be allocatable although a lot of filesystem metadata may have to be
850 reclaimed to achieve this.
852 ..............................................................................
856 Provides information about distribution and utilization of memory. This
857 varies by architecture and compile options. The following is from a
858 16GB PIII, which has highmem enabled. You may not have all of these fields.
862 MemTotal: 16344972 kB
864 MemAvailable: 14836172 kB
870 HighTotal: 15597528 kB
871 HighFree: 13629632 kB
881 KReclaimable: 168048 kB
883 SReclaimable: 159856 kB
884 SUnreclaim: 124508 kB
889 CommitLimit: 7669796 kB
890 Committed_AS: 100056 kB
891 VmallocTotal: 112216 kB
893 VmallocChunk: 111088 kB
895 HardwareCorrupted: 0 kB
896 AnonHugePages: 49152 kB
901 MemTotal: Total usable ram (i.e. physical ram minus a few reserved
902 bits and the kernel binary code)
903 MemFree: The sum of LowFree+HighFree
904 MemAvailable: An estimate of how much memory is available for starting new
905 applications, without swapping. Calculated from MemFree,
906 SReclaimable, the size of the file LRU lists, and the low
907 watermarks in each zone.
908 The estimate takes into account that the system needs some
909 page cache to function well, and that not all reclaimable
910 slab will be reclaimable, due to items being in use. The
911 impact of those factors will vary from system to system.
912 Buffers: Relatively temporary storage for raw disk blocks
913 shouldn't get tremendously large (20MB or so)
914 Cached: in-memory cache for files read from the disk (the
915 pagecache). Doesn't include SwapCached
916 SwapCached: Memory that once was swapped out, is swapped back in but
917 still also is in the swapfile (if memory is needed it
918 doesn't need to be swapped out AGAIN because it is already
919 in the swapfile. This saves I/O)
920 Active: Memory that has been used more recently and usually not
921 reclaimed unless absolutely necessary.
922 Inactive: Memory which has been less recently used. It is more
923 eligible to be reclaimed for other purposes
925 HighFree: Highmem is all memory above ~860MB of physical memory
926 Highmem areas are for use by userspace programs, or
927 for the pagecache. The kernel must use tricks to access
928 this memory, making it slower to access than lowmem.
930 LowFree: Lowmem is memory which can be used for everything that
931 highmem can be used for, but it is also available for the
932 kernel's use for its own data structures. Among many
933 other things, it is where everything from the Slab is
934 allocated. Bad things happen when you're out of lowmem.
935 SwapTotal: total amount of swap space available
936 SwapFree: Memory which has been evicted from RAM, and is temporarily
938 Dirty: Memory which is waiting to get written back to the disk
939 Writeback: Memory which is actively being written back to the disk
940 AnonPages: Non-file backed pages mapped into userspace page tables
941 HardwareCorrupted: The amount of RAM/memory in KB, the kernel identifies as
943 AnonHugePages: Non-file backed huge pages mapped into userspace page tables
944 Mapped: files which have been mmaped, such as libraries
945 Shmem: Total memory used by shared memory (shmem) and tmpfs
946 ShmemHugePages: Memory used by shared memory (shmem) and tmpfs allocated
948 ShmemPmdMapped: Shared memory mapped into userspace with huge pages
949 KReclaimable: Kernel allocations that the kernel will attempt to reclaim
950 under memory pressure. Includes SReclaimable (below), and other
951 direct allocations with a shrinker.
952 Slab: in-kernel data structures cache
953 SReclaimable: Part of Slab, that might be reclaimed, such as caches
954 SUnreclaim: Part of Slab, that cannot be reclaimed on memory pressure
955 PageTables: amount of memory dedicated to the lowest level of page
957 NFS_Unstable: NFS pages sent to the server, but not yet committed to stable
959 Bounce: Memory used for block device "bounce buffers"
960 WritebackTmp: Memory used by FUSE for temporary writeback buffers
961 CommitLimit: Based on the overcommit ratio ('vm.overcommit_ratio'),
962 this is the total amount of memory currently available to
963 be allocated on the system. This limit is only adhered to
964 if strict overcommit accounting is enabled (mode 2 in
965 'vm.overcommit_memory').
966 The CommitLimit is calculated with the following formula:
967 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *
968 overcommit_ratio / 100 + [total swap pages]
969 For example, on a system with 1G of physical RAM and 7G
970 of swap with a `vm.overcommit_ratio` of 30 it would
971 yield a CommitLimit of 7.3G.
972 For more details, see the memory overcommit documentation
973 in vm/overcommit-accounting.
974 Committed_AS: The amount of memory presently allocated on the system.
975 The committed memory is a sum of all of the memory which
976 has been allocated by processes, even if it has not been
977 "used" by them as of yet. A process which malloc()'s 1G
978 of memory, but only touches 300M of it will show up as
979 using 1G. This 1G is memory which has been "committed" to
980 by the VM and can be used at any time by the allocating
981 application. With strict overcommit enabled on the system
982 (mode 2 in 'vm.overcommit_memory'),allocations which would
983 exceed the CommitLimit (detailed above) will not be permitted.
984 This is useful if one needs to guarantee that processes will
985 not fail due to lack of memory once that memory has been
986 successfully allocated.
987 VmallocTotal: total size of vmalloc memory area
988 VmallocUsed: amount of vmalloc area which is used
989 VmallocChunk: largest contiguous block of vmalloc area which is free
990 Percpu: Memory allocated to the percpu allocator used to back percpu
991 allocations. This stat excludes the cost of metadata.
993 ..............................................................................
997 Provides information about vmalloced/vmaped areas. One line per area,
998 containing the virtual address range of the area, size in bytes,
999 caller information of the creator, and optional information depending
1000 on the kind of area :
1002 pages=nr number of pages
1003 phys=addr if a physical address was specified
1004 ioremap I/O mapping (ioremap() and friends)
1005 vmalloc vmalloc() area
1007 user VM_USERMAP area
1008 vpages buffer for pages pointers was vmalloced (huge area)
1009 N<node>=nr (Only on NUMA kernels)
1010 Number of pages allocated on memory node <node>
1012 > cat /proc/vmallocinfo
1013 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...
1014 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128
1015 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...
1016 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64
1017 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...
1018 phys=7fee8000 ioremap
1019 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...
1020 phys=7fee7000 ioremap
1021 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210
1022 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...
1023 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3
1024 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...
1025 pages=2 vmalloc N1=2
1026 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...
1027 /0x130 [x_tables] pages=4 vmalloc N0=4
1028 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...
1029 pages=14 vmalloc N2=14
1030 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...
1031 pages=4 vmalloc N1=4
1032 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...
1033 pages=2 vmalloc N1=2
1034 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...
1035 pages=10 vmalloc N0=10
1037 ..............................................................................
1041 Provides counts of softirq handlers serviced since boot time, for each cpu.
1043 > cat /proc/softirqs
1046 TIMER: 27166 27120 27097 27034
1051 SCHED: 27035 26983 26971 26746
1053 RCU: 1678 1769 2178 2250
1056 1.3 IDE devices in /proc/ide
1057 ----------------------------
1059 The subdirectory /proc/ide contains information about all IDE devices of which
1060 the kernel is aware. There is one subdirectory for each IDE controller, the
1061 file drivers and a link for each IDE device, pointing to the device directory
1062 in the controller specific subtree.
1064 The file drivers contains general information about the drivers used for the
1067 > cat /proc/ide/drivers
1068 ide-cdrom version 4.53
1069 ide-disk version 1.08
1071 More detailed information can be found in the controller specific
1072 subdirectories. These are named ide0, ide1 and so on. Each of these
1073 directories contains the files shown in table 1-6.
1076 Table 1-6: IDE controller info in /proc/ide/ide?
1077 ..............................................................................
1079 channel IDE channel (0 or 1)
1080 config Configuration (only for PCI/IDE bridge)
1082 model Type/Chipset of IDE controller
1083 ..............................................................................
1085 Each device connected to a controller has a separate subdirectory in the
1086 controllers directory. The files listed in table 1-7 are contained in these
1090 Table 1-7: IDE device information
1091 ..............................................................................
1094 capacity Capacity of the medium (in 512Byte blocks)
1095 driver driver and version
1096 geometry physical and logical geometry
1097 identify device identify block
1099 model device identifier
1100 settings device setup
1101 smart_thresholds IDE disk management thresholds
1102 smart_values IDE disk management values
1103 ..............................................................................
1105 The most interesting file is settings. This file contains a nice overview of
1106 the drive parameters:
1108 # cat /proc/ide/ide0/hda/settings
1109 name value min max mode
1110 ---- ----- --- --- ----
1111 bios_cyl 526 0 65535 rw
1112 bios_head 255 0 255 rw
1113 bios_sect 63 0 63 rw
1114 breada_readahead 4 0 127 rw
1116 file_readahead 72 0 2097151 rw
1118 keepsettings 0 0 1 rw
1119 max_kb_per_request 122 1 127 rw
1123 pio_mode write-only 0 255 w
1129 1.4 Networking info in /proc/net
1130 --------------------------------
1132 The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the
1133 additional values you get for IP version 6 if you configure the kernel to
1134 support this. Table 1-9 lists the files and their meaning.
1137 Table 1-8: IPv6 info in /proc/net
1138 ..............................................................................
1140 udp6 UDP sockets (IPv6)
1141 tcp6 TCP sockets (IPv6)
1142 raw6 Raw device statistics (IPv6)
1143 igmp6 IP multicast addresses, which this host joined (IPv6)
1144 if_inet6 List of IPv6 interface addresses
1145 ipv6_route Kernel routing table for IPv6
1146 rt6_stats Global IPv6 routing tables statistics
1147 sockstat6 Socket statistics (IPv6)
1148 snmp6 Snmp data (IPv6)
1149 ..............................................................................
1152 Table 1-9: Network info in /proc/net
1153 ..............................................................................
1155 arp Kernel ARP table
1156 dev network devices with statistics
1157 dev_mcast the Layer2 multicast groups a device is listening too
1158 (interface index, label, number of references, number of bound
1160 dev_stat network device status
1161 ip_fwchains Firewall chain linkage
1162 ip_fwnames Firewall chain names
1163 ip_masq Directory containing the masquerading tables
1164 ip_masquerade Major masquerading table
1165 netstat Network statistics
1166 raw raw device statistics
1167 route Kernel routing table
1168 rpc Directory containing rpc info
1169 rt_cache Routing cache
1171 sockstat Socket statistics
1174 unix UNIX domain sockets
1175 wireless Wireless interface data (Wavelan etc)
1176 igmp IP multicast addresses, which this host joined
1177 psched Global packet scheduler parameters.
1178 netlink List of PF_NETLINK sockets
1179 ip_mr_vifs List of multicast virtual interfaces
1180 ip_mr_cache List of multicast routing cache
1181 ..............................................................................
1183 You can use this information to see which network devices are available in
1184 your system and how much traffic was routed over those devices:
1187 Inter-|Receive |[...
1188 face |bytes packets errs drop fifo frame compressed multicast|[...
1189 lo: 908188 5596 0 0 0 0 0 0 [...
1190 ppp0:15475140 20721 410 0 0 410 0 0 [...
1191 eth0: 614530 7085 0 0 0 0 0 1 [...
1194 ...] bytes packets errs drop fifo colls carrier compressed
1195 ...] 908188 5596 0 0 0 0 0 0
1196 ...] 1375103 17405 0 0 0 0 0 0
1197 ...] 1703981 5535 0 0 0 3 0 0
1199 In addition, each Channel Bond interface has its own directory. For
1200 example, the bond0 device will have a directory called /proc/net/bond0/.
1201 It will contain information that is specific to that bond, such as the
1202 current slaves of the bond, the link status of the slaves, and how
1203 many times the slaves link has failed.
1208 If you have a SCSI host adapter in your system, you'll find a subdirectory
1209 named after the driver for this adapter in /proc/scsi. You'll also see a list
1210 of all recognized SCSI devices in /proc/scsi:
1212 >cat /proc/scsi/scsi
1214 Host: scsi0 Channel: 00 Id: 00 Lun: 00
1215 Vendor: IBM Model: DGHS09U Rev: 03E0
1216 Type: Direct-Access ANSI SCSI revision: 03
1217 Host: scsi0 Channel: 00 Id: 06 Lun: 00
1218 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04
1219 Type: CD-ROM ANSI SCSI revision: 02
1222 The directory named after the driver has one file for each adapter found in
1223 the system. These files contain information about the controller, including
1224 the used IRQ and the IO address range. The amount of information shown is
1225 dependent on the adapter you use. The example shows the output for an Adaptec
1226 AHA-2940 SCSI adapter:
1228 > cat /proc/scsi/aic7xxx/0
1230 Adaptec AIC7xxx driver version: 5.1.19/3.2.4
1232 TCQ Enabled By Default : Disabled
1233 AIC7XXX_PROC_STATS : Disabled
1234 AIC7XXX_RESET_DELAY : 5
1235 Adapter Configuration:
1236 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter
1237 Ultra Wide Controller
1238 PCI MMAPed I/O Base: 0xeb001000
1239 Adapter SEEPROM Config: SEEPROM found and used.
1240 Adaptec SCSI BIOS: Enabled
1242 SCBs: Active 0, Max Active 2,
1243 Allocated 15, HW 16, Page 255
1245 BIOS Control Word: 0x18b6
1246 Adapter Control Word: 0x005b
1247 Extended Translation: Enabled
1248 Disconnect Enable Flags: 0xffff
1249 Ultra Enable Flags: 0x0001
1250 Tag Queue Enable Flags: 0x0000
1251 Ordered Queue Tag Flags: 0x0000
1252 Default Tag Queue Depth: 8
1253 Tagged Queue By Device array for aic7xxx host instance 0:
1254 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
1255 Actual queue depth per device for aic7xxx host instance 0:
1256 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
1259 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
1260 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)
1261 Total transfers 160151 (74577 reads and 85574 writes)
1263 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15
1264 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)
1265 Total transfers 0 (0 reads and 0 writes)
1268 1.6 Parallel port info in /proc/parport
1269 ---------------------------------------
1271 The directory /proc/parport contains information about the parallel ports of
1272 your system. It has one subdirectory for each port, named after the port
1275 These directories contain the four files shown in Table 1-10.
1278 Table 1-10: Files in /proc/parport
1279 ..............................................................................
1281 autoprobe Any IEEE-1284 device ID information that has been acquired.
1282 devices list of the device drivers using that port. A + will appear by the
1283 name of the device currently using the port (it might not appear
1285 hardware Parallel port's base address, IRQ line and DMA channel.
1286 irq IRQ that parport is using for that port. This is in a separate
1287 file to allow you to alter it by writing a new value in (IRQ
1289 ..............................................................................
1291 1.7 TTY info in /proc/tty
1292 -------------------------
1294 Information about the available and actually used tty's can be found in the
1295 directory /proc/tty.You'll find entries for drivers and line disciplines in
1296 this directory, as shown in Table 1-11.
1299 Table 1-11: Files in /proc/tty
1300 ..............................................................................
1302 drivers list of drivers and their usage
1303 ldiscs registered line disciplines
1304 driver/serial usage statistic and status of single tty lines
1305 ..............................................................................
1307 To see which tty's are currently in use, you can simply look into the file
1310 > cat /proc/tty/drivers
1311 pty_slave /dev/pts 136 0-255 pty:slave
1312 pty_master /dev/ptm 128 0-255 pty:master
1313 pty_slave /dev/ttyp 3 0-255 pty:slave
1314 pty_master /dev/pty 2 0-255 pty:master
1315 serial /dev/cua 5 64-67 serial:callout
1316 serial /dev/ttyS 4 64-67 serial
1317 /dev/tty0 /dev/tty0 4 0 system:vtmaster
1318 /dev/ptmx /dev/ptmx 5 2 system
1319 /dev/console /dev/console 5 1 system:console
1320 /dev/tty /dev/tty 5 0 system:/dev/tty
1321 unknown /dev/tty 4 1-63 console
1324 1.8 Miscellaneous kernel statistics in /proc/stat
1325 -------------------------------------------------
1327 Various pieces of information about kernel activity are available in the
1328 /proc/stat file. All of the numbers reported in this file are aggregates
1329 since the system first booted. For a quick look, simply cat the file:
1332 cpu 2255 34 2290 22625563 6290 127 456 0 0 0
1333 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0
1334 cpu1 1123 0 849 11313845 2614 0 18 0 0 0
1335 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...]
1341 softirq 183433 0 21755 12 39 1137 231 21459 2263
1343 The very first "cpu" line aggregates the numbers in all of the other "cpuN"
1344 lines. These numbers identify the amount of time the CPU has spent performing
1345 different kinds of work. Time units are in USER_HZ (typically hundredths of a
1346 second). The meanings of the columns are as follows, from left to right:
1348 - user: normal processes executing in user mode
1349 - nice: niced processes executing in user mode
1350 - system: processes executing in kernel mode
1351 - idle: twiddling thumbs
1352 - iowait: In a word, iowait stands for waiting for I/O to complete. But there
1353 are several problems:
1354 1. Cpu will not wait for I/O to complete, iowait is the time that a task is
1355 waiting for I/O to complete. When cpu goes into idle state for
1356 outstanding task io, another task will be scheduled on this CPU.
1357 2. In a multi-core CPU, the task waiting for I/O to complete is not running
1358 on any CPU, so the iowait of each CPU is difficult to calculate.
1359 3. The value of iowait field in /proc/stat will decrease in certain
1361 So, the iowait is not reliable by reading from /proc/stat.
1362 - irq: servicing interrupts
1363 - softirq: servicing softirqs
1364 - steal: involuntary wait
1365 - guest: running a normal guest
1366 - guest_nice: running a niced guest
1368 The "intr" line gives counts of interrupts serviced since boot time, for each
1369 of the possible system interrupts. The first column is the total of all
1370 interrupts serviced including unnumbered architecture specific interrupts;
1371 each subsequent column is the total for that particular numbered interrupt.
1372 Unnumbered interrupts are not shown, only summed into the total.
1374 The "ctxt" line gives the total number of context switches across all CPUs.
1376 The "btime" line gives the time at which the system booted, in seconds since
1379 The "processes" line gives the number of processes and threads created, which
1380 includes (but is not limited to) those created by calls to the fork() and
1381 clone() system calls.
1383 The "procs_running" line gives the total number of threads that are
1384 running or ready to run (i.e., the total number of runnable threads).
1386 The "procs_blocked" line gives the number of processes currently blocked,
1387 waiting for I/O to complete.
1389 The "softirq" line gives counts of softirqs serviced since boot time, for each
1390 of the possible system softirqs. The first column is the total of all
1391 softirqs serviced; each subsequent column is the total for that particular
1395 1.9 Ext4 file system parameters
1396 -------------------------------
1398 Information about mounted ext4 file systems can be found in
1399 /proc/fs/ext4. Each mounted filesystem will have a directory in
1400 /proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or
1401 /proc/fs/ext4/dm-0). The files in each per-device directory are shown
1402 in Table 1-12, below.
1404 Table 1-12: Files in /proc/fs/ext4/<devname>
1405 ..............................................................................
1407 mb_groups details of multiblock allocator buddy cache of free blocks
1408 ..............................................................................
1412 Shows registered system console lines.
1414 To see which character device lines are currently used for the system console
1415 /dev/console, you may simply look into the file /proc/consoles:
1417 > cat /proc/consoles
1423 device name of the device
1424 operations R = can do read operations
1425 W = can do write operations
1427 flags E = it is enabled
1428 C = it is preferred console
1429 B = it is primary boot console
1430 p = it is used for printk buffer
1431 b = it is not a TTY but a Braille device
1432 a = it is safe to use when cpu is offline
1433 major:minor major and minor number of the device separated by a colon
1435 ------------------------------------------------------------------------------
1437 ------------------------------------------------------------------------------
1438 The /proc file system serves information about the running system. It not only
1439 allows access to process data but also allows you to request the kernel status
1440 by reading files in the hierarchy.
1442 The directory structure of /proc reflects the types of information and makes
1443 it easy, if not obvious, where to look for specific data.
1444 ------------------------------------------------------------------------------
1446 ------------------------------------------------------------------------------
1447 CHAPTER 2: MODIFYING SYSTEM PARAMETERS
1448 ------------------------------------------------------------------------------
1450 ------------------------------------------------------------------------------
1452 ------------------------------------------------------------------------------
1453 * Modifying kernel parameters by writing into files found in /proc/sys
1454 * Exploring the files which modify certain parameters
1455 * Review of the /proc/sys file tree
1456 ------------------------------------------------------------------------------
1459 A very interesting part of /proc is the directory /proc/sys. This is not only
1460 a source of information, it also allows you to change parameters within the
1461 kernel. Be very careful when attempting this. You can optimize your system,
1462 but you can also cause it to crash. Never alter kernel parameters on a
1463 production system. Set up a development machine and test to make sure that
1464 everything works the way you want it to. You may have no alternative but to
1465 reboot the machine once an error has been made.
1467 To change a value, simply echo the new value into the file. An example is
1468 given below in the section on the file system data. You need to be root to do
1469 this. You can create your own boot script to perform this every time your
1472 The files in /proc/sys can be used to fine tune and monitor miscellaneous and
1473 general things in the operation of the Linux kernel. Since some of the files
1474 can inadvertently disrupt your system, it is advisable to read both
1475 documentation and source before actually making adjustments. In any case, be
1476 very careful when writing to any of these files. The entries in /proc may
1477 change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt
1478 review the kernel documentation in the directory /usr/src/linux/Documentation.
1479 This chapter is heavily based on the documentation included in the pre 2.2
1480 kernels, and became part of it in version 2.2.1 of the Linux kernel.
1482 Please see: Documentation/sysctl/ directory for descriptions of these
1485 ------------------------------------------------------------------------------
1487 ------------------------------------------------------------------------------
1488 Certain aspects of kernel behavior can be modified at runtime, without the
1489 need to recompile the kernel, or even to reboot the system. The files in the
1490 /proc/sys tree can not only be read, but also modified. You can use the echo
1491 command to write value into these files, thereby changing the default settings
1493 ------------------------------------------------------------------------------
1495 ------------------------------------------------------------------------------
1496 CHAPTER 3: PER-PROCESS PARAMETERS
1497 ------------------------------------------------------------------------------
1499 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score
1500 --------------------------------------------------------------------------------
1502 These file can be used to adjust the badness heuristic used to select which
1503 process gets killed in out of memory conditions.
1505 The badness heuristic assigns a value to each candidate task ranging from 0
1506 (never kill) to 1000 (always kill) to determine which process is targeted. The
1507 units are roughly a proportion along that range of allowed memory the process
1508 may allocate from based on an estimation of its current memory and swap use.
1509 For example, if a task is using all allowed memory, its badness score will be
1510 1000. If it is using half of its allowed memory, its score will be 500.
1512 There is an additional factor included in the badness score: the current memory
1513 and swap usage is discounted by 3% for root processes.
1515 The amount of "allowed" memory depends on the context in which the oom killer
1516 was called. If it is due to the memory assigned to the allocating task's cpuset
1517 being exhausted, the allowed memory represents the set of mems assigned to that
1518 cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed
1519 memory represents the set of mempolicy nodes. If it is due to a memory
1520 limit (or swap limit) being reached, the allowed memory is that configured
1521 limit. Finally, if it is due to the entire system being out of memory, the
1522 allowed memory represents all allocatable resources.
1524 The value of /proc/<pid>/oom_score_adj is added to the badness score before it
1525 is used to determine which task to kill. Acceptable values range from -1000
1526 (OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to
1527 polarize the preference for oom killing either by always preferring a certain
1528 task or completely disabling it. The lowest possible value, -1000, is
1529 equivalent to disabling oom killing entirely for that task since it will always
1530 report a badness score of 0.
1532 Consequently, it is very simple for userspace to define the amount of memory to
1533 consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for
1534 example, is roughly equivalent to allowing the remainder of tasks sharing the
1535 same system, cpuset, mempolicy, or memory controller resources to use at least
1536 50% more memory. A value of -500, on the other hand, would be roughly
1537 equivalent to discounting 50% of the task's allowed memory from being considered
1538 as scoring against the task.
1540 For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also
1541 be used to tune the badness score. Its acceptable values range from -16
1542 (OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17
1543 (OOM_DISABLE) to disable oom killing entirely for that task. Its value is
1544 scaled linearly with /proc/<pid>/oom_score_adj.
1546 The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last
1547 value set by a CAP_SYS_RESOURCE process. To reduce the value any lower
1548 requires CAP_SYS_RESOURCE.
1550 Caveat: when a parent task is selected, the oom killer will sacrifice any first
1551 generation children with separate address spaces instead, if possible. This
1552 avoids servers and important system daemons from being killed and loses the
1553 minimal amount of work.
1556 3.2 /proc/<pid>/oom_score - Display current oom-killer score
1557 -------------------------------------------------------------
1559 This file can be used to check the current score used by the oom-killer is for
1560 any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which
1561 process should be killed in an out-of-memory situation.
1564 3.3 /proc/<pid>/io - Display the IO accounting fields
1565 -------------------------------------------------------
1567 This file contains IO statistics for each running process
1572 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &
1575 test:/tmp # cat /proc/3828/io
1581 write_bytes: 323932160
1582 cancelled_write_bytes: 0
1591 I/O counter: chars read
1592 The number of bytes which this task has caused to be read from storage. This
1593 is simply the sum of bytes which this process passed to read() and pread().
1594 It includes things like tty IO and it is unaffected by whether or not actual
1595 physical disk IO was required (the read might have been satisfied from
1602 I/O counter: chars written
1603 The number of bytes which this task has caused, or shall cause to be written
1604 to disk. Similar caveats apply here as with rchar.
1610 I/O counter: read syscalls
1611 Attempt to count the number of read I/O operations, i.e. syscalls like read()
1618 I/O counter: write syscalls
1619 Attempt to count the number of write I/O operations, i.e. syscalls like
1620 write() and pwrite().
1626 I/O counter: bytes read
1627 Attempt to count the number of bytes which this process really did cause to
1628 be fetched from the storage layer. Done at the submit_bio() level, so it is
1629 accurate for block-backed filesystems. <please add status regarding NFS and
1630 CIFS at a later time>
1636 I/O counter: bytes written
1637 Attempt to count the number of bytes which this process caused to be sent to
1638 the storage layer. This is done at page-dirtying time.
1641 cancelled_write_bytes
1642 ---------------------
1644 The big inaccuracy here is truncate. If a process writes 1MB to a file and
1645 then deletes the file, it will in fact perform no writeout. But it will have
1646 been accounted as having caused 1MB of write.
1647 In other words: The number of bytes which this process caused to not happen,
1648 by truncating pagecache. A task can cause "negative" IO too. If this task
1649 truncates some dirty pagecache, some IO which another task has been accounted
1650 for (in its write_bytes) will not be happening. We _could_ just subtract that
1651 from the truncating task's write_bytes, but there is information loss in doing
1658 At its current implementation state, this is a bit racy on 32-bit machines: if
1659 process A reads process B's /proc/pid/io while process B is updating one of
1660 those 64-bit counters, process A could see an intermediate result.
1663 More information about this can be found within the taskstats documentation in
1664 Documentation/accounting.
1666 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
1667 ---------------------------------------------------------------
1668 When a process is dumped, all anonymous memory is written to a core file as
1669 long as the size of the core file isn't limited. But sometimes we don't want
1670 to dump some memory segments, for example, huge shared memory or DAX.
1671 Conversely, sometimes we want to save file-backed memory segments into a core
1672 file, not only the individual files.
1674 /proc/<pid>/coredump_filter allows you to customize which memory segments
1675 will be dumped when the <pid> process is dumped. coredump_filter is a bitmask
1676 of memory types. If a bit of the bitmask is set, memory segments of the
1677 corresponding memory type are dumped, otherwise they are not dumped.
1679 The following 9 memory types are supported:
1680 - (bit 0) anonymous private memory
1681 - (bit 1) anonymous shared memory
1682 - (bit 2) file-backed private memory
1683 - (bit 3) file-backed shared memory
1684 - (bit 4) ELF header pages in file-backed private memory areas (it is
1685 effective only if the bit 2 is cleared)
1686 - (bit 5) hugetlb private memory
1687 - (bit 6) hugetlb shared memory
1688 - (bit 7) DAX private memory
1689 - (bit 8) DAX shared memory
1691 Note that MMIO pages such as frame buffer are never dumped and vDSO pages
1692 are always dumped regardless of the bitmask status.
1694 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is
1695 only affected by bit 5-6, and DAX is only affected by bits 7-8.
1697 The default value of coredump_filter is 0x33; this means all anonymous memory
1698 segments, ELF header pages and hugetlb private memory are dumped.
1700 If you don't want to dump all shared memory segments attached to pid 1234,
1701 write 0x31 to the process's proc file.
1703 $ echo 0x31 > /proc/1234/coredump_filter
1705 When a new process is created, the process inherits the bitmask status from its
1706 parent. It is useful to set up coredump_filter before the program runs.
1709 $ echo 0x7 > /proc/self/coredump_filter
1712 3.5 /proc/<pid>/mountinfo - Information about mounts
1713 --------------------------------------------------------
1715 This file contains lines of the form:
1717 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue
1718 (1)(2)(3) (4) (5) (6) (7) (8) (9) (10) (11)
1720 (1) mount ID: unique identifier of the mount (may be reused after umount)
1721 (2) parent ID: ID of parent (or of self for the top of the mount tree)
1722 (3) major:minor: value of st_dev for files on filesystem
1723 (4) root: root of the mount within the filesystem
1724 (5) mount point: mount point relative to the process's root
1725 (6) mount options: per mount options
1726 (7) optional fields: zero or more fields of the form "tag[:value]"
1727 (8) separator: marks the end of the optional fields
1728 (9) filesystem type: name of filesystem of the form "type[.subtype]"
1729 (10) mount source: filesystem specific information or "none"
1730 (11) super options: per super block options
1732 Parsers should ignore all unrecognised optional fields. Currently the
1733 possible optional fields are:
1735 shared:X mount is shared in peer group X
1736 master:X mount is slave to peer group X
1737 propagate_from:X mount is slave and receives propagation from peer group X (*)
1738 unbindable mount is unbindable
1740 (*) X is the closest dominant peer group under the process's root. If
1741 X is the immediate master of the mount, or if there's no dominant peer
1742 group under the same root, then only the "master:X" field is present
1743 and not the "propagate_from:X" field.
1745 For more information on mount propagation see:
1747 Documentation/filesystems/sharedsubtree.txt
1750 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
1751 --------------------------------------------------------
1752 These files provide a method to access a tasks comm value. It also allows for
1753 a task to set its own or one of its thread siblings comm value. The comm value
1754 is limited in size compared to the cmdline value, so writing anything longer
1755 then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
1759 3.7 /proc/<pid>/task/<tid>/children - Information about task children
1760 -------------------------------------------------------------------------
1761 This file provides a fast way to retrieve first level children pids
1762 of a task pointed by <pid>/<tid> pair. The format is a space separated
1765 Note the "first level" here -- if a child has own children they will
1766 not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
1767 to obtain the descendants.
1769 Since this interface is intended to be fast and cheap it doesn't
1770 guarantee to provide precise results and some children might be
1771 skipped, especially if they've exited right after we printed their
1772 pids, so one need to either stop or freeze processes being inspected
1773 if precise results are needed.
1776 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file
1777 ---------------------------------------------------------------
1778 This file provides information associated with an opened file. The regular
1779 files have at least three fields -- 'pos', 'flags' and mnt_id. The 'pos'
1780 represents the current offset of the opened file in decimal form [see lseek(2)
1781 for details], 'flags' denotes the octal O_xxx mask the file has been
1782 created with [see open(2) for details] and 'mnt_id' represents mount ID of
1783 the file system containing the opened file [see 3.5 /proc/<pid>/mountinfo
1792 All locks associated with a file descriptor are shown in its fdinfo too.
1794 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF
1796 The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags
1797 pair provide additional information particular to the objects they represent.
1806 where 'eventfd-count' is hex value of a counter.
1813 sigmask: 0000000000000200
1815 where 'sigmask' is hex value of the signal mask associated
1823 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7
1825 where 'tfd' is a target file descriptor number in decimal form,
1826 'events' is events mask being watched and the 'data' is data
1827 associated with a target [see epoll(7) for more details].
1829 The 'pos' is current offset of the target file in decimal form
1830 [see lseek(2)], 'ino' and 'sdev' are inode and device numbers
1831 where target file resides, all in hex format.
1835 For inotify files the format is the following
1839 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d
1841 where 'wd' is a watch descriptor in decimal form, ie a target file
1842 descriptor number, 'ino' and 'sdev' are inode and device where the
1843 target file resides and the 'mask' is the mask of events, all in hex
1844 form [see inotify(7) for more details].
1846 If the kernel was built with exportfs support, the path to the target
1847 file is encoded as a file handle. The file handle is provided by three
1848 fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex
1851 If the kernel is built without exportfs support the file handle won't be
1854 If there is no inotify mark attached yet the 'inotify' line will be omitted.
1856 For fanotify files the format is
1861 fanotify flags:10 event-flags:0
1862 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003
1863 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4
1865 where fanotify 'flags' and 'event-flags' are values used in fanotify_init
1866 call, 'mnt_id' is the mount point identifier, 'mflags' is the value of
1867 flags associated with mark which are tracked separately from events
1868 mask. 'ino', 'sdev' are target inode and device, 'mask' is the events
1869 mask and 'ignored_mask' is the mask of events which are to be ignored.
1870 All in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'
1871 does provide information about flags and mask used in fanotify_mark
1872 call [see fsnotify manpage for details].
1874 While the first three lines are mandatory and always printed, the rest is
1875 optional and may be omitted if no marks created yet.
1886 it_value: (0, 49406829)
1889 where 'clockid' is the clock type and 'ticks' is the number of the timer expirations
1890 that have occurred [see timerfd_create(2) for details]. 'settime flags' are
1891 flags in octal form been used to setup the timer [see timerfd_settime(2) for
1892 details]. 'it_value' is remaining time until the timer exiration.
1893 'it_interval' is the interval for the timer. Note the timer might be set up
1894 with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'
1895 still exhibits timer's remaining time.
1897 3.9 /proc/<pid>/map_files - Information about memory mapped files
1898 ---------------------------------------------------------------------
1899 This directory contains symbolic links which represent memory mapped files
1900 the process is maintaining. Example output:
1902 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so
1903 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so
1904 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so
1906 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1
1907 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls
1909 The name of a link represents the virtual memory bounds of a mapping, i.e.
1910 vm_area_struct::vm_start-vm_area_struct::vm_end.
1912 The main purpose of the map_files is to retrieve a set of memory mapped
1913 files in a fast way instead of parsing /proc/<pid>/maps or
1914 /proc/<pid>/smaps, both of which contain many more records. At the same
1915 time one can open(2) mappings from the listings of two processes and
1916 comparing their inode numbers to figure out which anonymous memory areas
1917 are actually shared.
1919 3.10 /proc/<pid>/timerslack_ns - Task timerslack value
1920 ---------------------------------------------------------
1921 This file provides the value of the task's timerslack value in nanoseconds.
1922 This value specifies a amount of time that normal timers may be deferred
1923 in order to coalesce timers and avoid unnecessary wakeups.
1925 This allows a task's interactivity vs power consumption trade off to be
1928 Writing 0 to the file will set the tasks timerslack to the default value.
1930 Valid values are from 0 - ULLONG_MAX
1932 An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level
1933 permissions on the task specified to change its timerslack_ns value.
1935 3.11 /proc/<pid>/patch_state - Livepatch patch operation state
1936 -----------------------------------------------------------------
1937 When CONFIG_LIVEPATCH is enabled, this file displays the value of the
1938 patch state for the task.
1940 A value of '-1' indicates that no patch is in transition.
1942 A value of '0' indicates that a patch is in transition and the task is
1943 unpatched. If the patch is being enabled, then the task hasn't been
1944 patched yet. If the patch is being disabled, then the task has already
1947 A value of '1' indicates that a patch is in transition and the task is
1948 patched. If the patch is being enabled, then the task has already been
1949 patched. If the patch is being disabled, then the task hasn't been
1952 3.12 /proc/<pid>/arch_status - task architecture specific status
1953 -------------------------------------------------------------------
1954 When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the
1955 architecture specific status of the task.
1959 $ cat /proc/6753/arch_status
1960 AVX512_elapsed_ms: 8
1965 x86 specific entries:
1966 ---------------------
1969 If AVX512 is supported on the machine, this entry shows the milliseconds
1970 elapsed since the last time AVX512 usage was recorded. The recording
1971 happens on a best effort basis when a task is scheduled out. This means
1972 that the value depends on two factors:
1974 1) The time which the task spent on the CPU without being scheduled
1975 out. With CPU isolation and a single runnable task this can take
1978 2) The time since the task was scheduled out last. Depending on the
1979 reason for being scheduled out (time slice exhausted, syscall ...)
1980 this can be arbitrary long time.
1982 As a consequence the value cannot be considered precise and authoritative
1983 information. The application which uses this information has to be aware
1984 of the overall scenario on the system in order to determine whether a
1985 task is a real AVX512 user or not. Precise information can be obtained
1986 with performance counters.
1988 A special value of '-1' indicates that no AVX512 usage was recorded, thus
1989 the task is unlikely an AVX512 user, but depends on the workload and the
1990 scheduling scenario, it also could be a false negative mentioned above.
1992 ------------------------------------------------------------------------------
1994 ------------------------------------------------------------------------------
1997 ---------------------
1999 The following mount options are supported:
2001 hidepid= Set /proc/<pid>/ access mode.
2002 gid= Set the group authorized to learn processes information.
2004 hidepid=0 means classic mode - everybody may access all /proc/<pid>/ directories
2007 hidepid=1 means users may not access any /proc/<pid>/ directories but their
2008 own. Sensitive files like cmdline, sched*, status are now protected against
2009 other users. This makes it impossible to learn whether any user runs
2010 specific program (given the program doesn't reveal itself by its behaviour).
2011 As an additional bonus, as /proc/<pid>/cmdline is unaccessible for other users,
2012 poorly written programs passing sensitive information via program arguments are
2013 now protected against local eavesdroppers.
2015 hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be fully invisible to other
2016 users. It doesn't mean that it hides a fact whether a process with a specific
2017 pid value exists (it can be learned by other means, e.g. by "kill -0 $PID"),
2018 but it hides process' uid and gid, which may be learned by stat()'ing
2019 /proc/<pid>/ otherwise. It greatly complicates an intruder's task of gathering
2020 information about running processes, whether some daemon runs with elevated
2021 privileges, whether other user runs some sensitive program, whether other users
2022 run any program at all, etc.
2024 gid= defines a group authorized to learn processes information otherwise
2025 prohibited by hidepid=. If you use some daemon like identd which needs to learn
2026 information about processes information, just add identd to this group.