2322 lines · plain
1.. SPDX-License-Identifier: GPL-2.02 3====================4The /proc Filesystem5====================6 7===================== ======================================= ================8/proc/sys Terrehon Bowden <terrehon@pacbell.net>, October 7 19999 Bodo Bauer <bb@ricochet.net>102.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 200011move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 200912fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 200913===================== ======================================= ================14 15 16 17.. Table of Contents18 19 0 Preface20 0.1 Introduction/Credits21 0.2 Legal Stuff22 23 1 Collecting System Information24 1.1 Process-Specific Subdirectories25 1.2 Kernel data26 1.3 IDE devices in /proc/ide27 1.4 Networking info in /proc/net28 1.5 SCSI info29 1.6 Parallel port info in /proc/parport30 1.7 TTY info in /proc/tty31 1.8 Miscellaneous kernel statistics in /proc/stat32 1.9 Ext4 file system parameters33 34 2 Modifying System Parameters35 36 3 Per-Process Parameters37 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer38 score39 3.2 /proc/<pid>/oom_score - Display current oom-killer score40 3.3 /proc/<pid>/io - Display the IO accounting fields41 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings42 3.5 /proc/<pid>/mountinfo - Information about mounts43 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm44 3.7 /proc/<pid>/task/<tid>/children - Information about task children45 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file46 3.9 /proc/<pid>/map_files - Information about memory mapped files47 3.10 /proc/<pid>/timerslack_ns - Task timerslack value48 3.11 /proc/<pid>/patch_state - Livepatch patch operation state49 3.12 /proc/<pid>/arch_status - Task architecture specific information50 3.13 /proc/<pid>/fd - List of symlinks to open files51 52 4 Configuring procfs53 4.1 Mount options54 55 5 Filesystem behavior56 57Preface58=======59 600.1 Introduction/Credits61------------------------62 63This documentation is part of a soon (or so we hope) to be released book on64the SuSE Linux distribution. As there is no complete documentation for the65/proc file system and we've used many freely available sources to write these66chapters, it seems only fair to give the work back to the Linux community.67This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm68afraid it's still far from complete, but we hope it will be useful. As far as69we know, it is the first 'all-in-one' document about the /proc file system. It70is focused on the Intel x86 hardware, so if you are looking for PPC, ARM,71SPARC, AXP, etc., features, you probably won't find what you are looking for.72It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But73additions and patches are welcome and will be added to this document if you74mail them to Bodo.75 76We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of77other people for help compiling this documentation. We'd also like to extend a78special thank you to Andi Kleen for documentation, which we relied on heavily79to create this document, as well as the additional information he provided.80Thanks to everybody else who contributed source or docs to the Linux kernel81and helped create a great piece of software... :)82 83If you have any comments, corrections or additions, please don't hesitate to84contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this85document.86 87The latest version of this document is available online at88https://www.kernel.org/doc/html/latest/filesystems/proc.html89 90If the above direction does not works for you, you could try the kernel91mailing list at linux-kernel@vger.kernel.org and/or try to reach me at92comandante@zaralinux.com.93 940.2 Legal Stuff95---------------96 97We don't guarantee the correctness of this document, and if you come to us98complaining about how you screwed up your system because of incorrect99documentation, we won't feel responsible...100 101Chapter 1: Collecting System Information102========================================103 104In This Chapter105---------------106* Investigating the properties of the pseudo file system /proc and its107 ability to provide information on the running Linux system108* Examining /proc's structure109* Uncovering various information about the kernel and the processes running110 on the system111 112------------------------------------------------------------------------------113 114The proc file system acts as an interface to internal data structures in the115kernel. It can be used to obtain information about the system and to change116certain kernel parameters at runtime (sysctl).117 118First, we'll take a look at the read-only parts of /proc. In Chapter 2, we119show you how you can use /proc/sys to change settings.120 1211.1 Process-Specific Subdirectories122-----------------------------------123 124The directory /proc contains (among other things) one subdirectory for each125process running on the system, which is named after the process ID (PID).126 127The link 'self' points to the process reading the file system. Each process128subdirectory has the entries listed in Table 1-1.129 130Note that an open file descriptor to /proc/<pid> or to any of its131contained files or subdirectories does not prevent <pid> being reused132for some other process in the event that <pid> exits. Operations on133open /proc/<pid> file descriptors corresponding to dead processes134never act on any new process that the kernel may, through chance, have135also assigned the process ID <pid>. Instead, operations on these FDs136usually fail with ESRCH.137 138.. table:: Table 1-1: Process specific entries in /proc139 140 ============= ===============================================================141 File Content142 ============= ===============================================================143 clear_refs Clears page referenced bits shown in smaps output144 cmdline Command line arguments145 cpu Current and last cpu in which it was executed (2.4)(smp)146 cwd Link to the current working directory147 environ Values of environment variables148 exe Link to the executable of this process149 fd Directory, which contains all file descriptors150 maps Memory maps to executables and library files (2.4)151 mem Memory held by this process152 root Link to the root directory of this process153 stat Process status154 statm Process memory status information155 status Process status in human readable form156 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function157 symbol the task is blocked in - or "0" if not blocked.158 pagemap Page table159 stack Report full stack trace, enable via CONFIG_STACKTRACE160 smaps An extension based on maps, showing the memory consumption of161 each mapping and flags associated with it162 smaps_rollup Accumulated smaps stats for all mappings of the process. This163 can be derived from smaps, but is faster and more convenient164 numa_maps An extension based on maps, showing the memory locality and165 binding policy as well as mem usage (in pages) of each mapping.166 ============= ===============================================================167 168For example, to get the status information of a process, all you have to do is169read the file /proc/PID/status::170 171 >cat /proc/self/status172 Name: cat173 State: R (running)174 Tgid: 5452175 Pid: 5452176 PPid: 743177 TracerPid: 0 (2.4)178 Uid: 501 501 501 501179 Gid: 100 100 100 100180 FDSize: 256181 Groups: 100 14 16182 Kthread: 0183 VmPeak: 5004 kB184 VmSize: 5004 kB185 VmLck: 0 kB186 VmHWM: 476 kB187 VmRSS: 476 kB188 RssAnon: 352 kB189 RssFile: 120 kB190 RssShmem: 4 kB191 VmData: 156 kB192 VmStk: 88 kB193 VmExe: 68 kB194 VmLib: 1412 kB195 VmPTE: 20 kb196 VmSwap: 0 kB197 HugetlbPages: 0 kB198 CoreDumping: 0199 THP_enabled: 1200 Threads: 1201 SigQ: 0/28578202 SigPnd: 0000000000000000203 ShdPnd: 0000000000000000204 SigBlk: 0000000000000000205 SigIgn: 0000000000000000206 SigCgt: 0000000000000000207 CapInh: 00000000fffffeff208 CapPrm: 0000000000000000209 CapEff: 0000000000000000210 CapBnd: ffffffffffffffff211 CapAmb: 0000000000000000212 NoNewPrivs: 0213 Seccomp: 0214 Speculation_Store_Bypass: thread vulnerable215 SpeculationIndirectBranch: conditional enabled216 voluntary_ctxt_switches: 0217 nonvoluntary_ctxt_switches: 1218 219This shows you nearly the same information you would get if you viewed it with220the ps command. In fact, ps uses the proc file system to obtain its221information. But you get a more detailed view of the process by reading the222file /proc/PID/status. It fields are described in table 1-2.223 224The statm file contains more detailed information about the process225memory usage. Its seven fields are explained in Table 1-3. The stat file226contains detailed information about the process itself. Its fields are227explained in Table 1-4.228 229(for SMP CONFIG users)230 231For making accounting scalable, RSS related information are handled in an232asynchronous manner and the value may not be very precise. To see a precise233snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table.234It's slow but very precise.235 236.. table:: Table 1-2: Contents of the status fields (as of 4.19)237 238 ========================== ===================================================239 Field Content240 ========================== ===================================================241 Name filename of the executable242 Umask file mode creation mask243 State state (R is running, S is sleeping, D is sleeping244 in an uninterruptible wait, Z is zombie,245 T is traced or stopped)246 Tgid thread group ID247 Ngid NUMA group ID (0 if none)248 Pid process id249 PPid process id of the parent process250 TracerPid PID of process tracing this process (0 if not, or251 the tracer is outside of the current pid namespace)252 Uid Real, effective, saved set, and file system UIDs253 Gid Real, effective, saved set, and file system GIDs254 FDSize number of file descriptor slots currently allocated255 Groups supplementary group list256 NStgid descendant namespace thread group ID hierarchy257 NSpid descendant namespace process ID hierarchy258 NSpgid descendant namespace process group ID hierarchy259 NSsid descendant namespace session ID hierarchy260 Kthread kernel thread flag, 1 is yes, 0 is no261 VmPeak peak virtual memory size262 VmSize total program size263 VmLck locked memory size264 VmPin pinned memory size265 VmHWM peak resident set size ("high water mark")266 VmRSS size of memory portions. It contains the three267 following parts268 (VmRSS = RssAnon + RssFile + RssShmem)269 RssAnon size of resident anonymous memory270 RssFile size of resident file mappings271 RssShmem size of resident shmem memory (includes SysV shm,272 mapping of tmpfs and shared anonymous mappings)273 VmData size of private data segments274 VmStk size of stack segments275 VmExe size of text segment276 VmLib size of shared library code277 VmPTE size of page table entries278 VmSwap amount of swap used by anonymous private data279 (shmem swap usage is not included)280 HugetlbPages size of hugetlb memory portions281 CoreDumping process's memory is currently being dumped282 (killing the process may lead to a corrupted core)283 THP_enabled process is allowed to use THP (returns 0 when284 PR_SET_THP_DISABLE is set on the process285 Threads number of threads286 SigQ number of signals queued/max. number for queue287 SigPnd bitmap of pending signals for the thread288 ShdPnd bitmap of shared pending signals for the process289 SigBlk bitmap of blocked signals290 SigIgn bitmap of ignored signals291 SigCgt bitmap of caught signals292 CapInh bitmap of inheritable capabilities293 CapPrm bitmap of permitted capabilities294 CapEff bitmap of effective capabilities295 CapBnd bitmap of capabilities bounding set296 CapAmb bitmap of ambient capabilities297 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...)298 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...)299 Speculation_Store_Bypass speculative store bypass mitigation status300 SpeculationIndirectBranch indirect branch speculation mode301 Cpus_allowed mask of CPUs on which this process may run302 Cpus_allowed_list Same as previous, but in "list format"303 Mems_allowed mask of memory nodes allowed to this process304 Mems_allowed_list Same as previous, but in "list format"305 voluntary_ctxt_switches number of voluntary context switches306 nonvoluntary_ctxt_switches number of non voluntary context switches307 ========================== ===================================================308 309 310.. table:: Table 1-3: Contents of the statm fields (as of 2.6.8-rc3)311 312 ======== =============================== ==============================313 Field Content314 ======== =============================== ==============================315 size total program size (pages) (same as VmSize in status)316 resident size of memory portions (pages) (same as VmRSS in status)317 shared number of pages that are shared (i.e. backed by a file, same318 as RssFile+RssShmem in status)319 trs number of pages that are 'code' (not including libs; broken,320 includes data segment)321 lrs number of pages of library (always 0 on 2.6)322 drs number of pages of data/stack (including libs; broken,323 includes library text)324 dt number of dirty pages (always 0 on 2.6)325 ======== =============================== ==============================326 327 328.. table:: Table 1-4: Contents of the stat fields (as of 2.6.30-rc7)329 330 ============= ===============================================================331 Field Content332 ============= ===============================================================333 pid process id334 tcomm filename of the executable335 state state (R is running, S is sleeping, D is sleeping in an336 uninterruptible wait, Z is zombie, T is traced or stopped)337 ppid process id of the parent process338 pgrp pgrp of the process339 sid session id340 tty_nr tty the process uses341 tty_pgrp pgrp of the tty342 flags task flags343 min_flt number of minor faults344 cmin_flt number of minor faults with child's345 maj_flt number of major faults346 cmaj_flt number of major faults with child's347 utime user mode jiffies348 stime kernel mode jiffies349 cutime user mode jiffies with child's350 cstime kernel mode jiffies with child's351 priority priority level352 nice nice level353 num_threads number of threads354 it_real_value (obsolete, always 0)355 start_time time the process started after system boot356 vsize virtual memory size357 rss resident set memory size358 rsslim current limit in bytes on the rss359 start_code address above which program text can run360 end_code address below which program text can run361 start_stack address of the start of the main process stack362 esp current value of ESP363 eip current value of EIP364 pending bitmap of pending signals365 blocked bitmap of blocked signals366 sigign bitmap of ignored signals367 sigcatch bitmap of caught signals368 0 (place holder, used to be the wchan address,369 use /proc/PID/wchan instead)370 0 (place holder)371 0 (place holder)372 exit_signal signal to send to parent thread on exit373 task_cpu which CPU the task is scheduled on374 rt_priority realtime priority375 policy scheduling policy (man sched_setscheduler)376 blkio_ticks time spent waiting for block IO377 gtime guest time of the task in jiffies378 cgtime guest time of the task children in jiffies379 start_data address above which program data+bss is placed380 end_data address below which program data+bss is placed381 start_brk address above which program heap can be expanded with brk()382 arg_start address above which program command line is placed383 arg_end address below which program command line is placed384 env_start address above which program environment is placed385 env_end address below which program environment is placed386 exit_code the thread's exit_code in the form reported by the waitpid387 system call388 ============= ===============================================================389 390The /proc/PID/maps file contains the currently mapped memory regions and391their access permissions.392 393The format is::394 395 address perms offset dev inode pathname396 397 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test398 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test399 0804a000-0806b000 rw-p 00000000 00:00 0 [heap]400 a7cb1000-a7cb2000 ---p 00000000 00:00 0401 a7cb2000-a7eb2000 rw-p 00000000 00:00 0402 a7eb2000-a7eb3000 ---p 00000000 00:00 0403 a7eb3000-a7ed5000 rw-p 00000000 00:00 0404 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6405 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6406 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6407 a800b000-a800e000 rw-p 00000000 00:00 0408 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0409 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0410 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0411 a8024000-a8027000 rw-p 00000000 00:00 0412 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2413 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2414 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2415 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack]416 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso]417 418where "address" is the address space in the process that it occupies, "perms"419is a set of permissions::420 421 r = read422 w = write423 x = execute424 s = shared425 p = private (copy on write)426 427"offset" is the offset into the mapping, "dev" is the device (major:minor), and428"inode" is the inode on that device. 0 indicates that no inode is associated429with the memory region, as the case would be with BSS (uninitialized data).430The "pathname" shows the name associated file for this mapping. If the mapping431is not associated with a file:432 433 =================== ===========================================434 [heap] the heap of the program435 [stack] the stack of the main process436 [vdso] the "virtual dynamic shared object",437 the kernel system call handler438 [anon:<name>] a private anonymous mapping that has been439 named by userspace440 [anon_shmem:<name>] an anonymous shared memory mapping that has441 been named by userspace442 =================== ===========================================443 444 or if empty, the mapping is anonymous.445 446Starting with 6.11 kernel, /proc/PID/maps provides an alternative447ioctl()-based API that gives ability to flexibly and efficiently query and448filter individual VMAs. This interface is binary and is meant for more449efficient and easy programmatic use. `struct procmap_query`, defined in450linux/fs.h UAPI header, serves as an input/output argument to the451`PROCMAP_QUERY` ioctl() command. See comments in linus/fs.h UAPI header for452details on query semantics, supported flags, data returned, and general API453usage information.454 455The /proc/PID/smaps is an extension based on maps, showing the memory456consumption for each of the process's mappings. For each mapping (aka Virtual457Memory Area, or VMA) there is a series of lines such as the following::458 459 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash460 461 Size: 1084 kB462 KernelPageSize: 4 kB463 MMUPageSize: 4 kB464 Rss: 892 kB465 Pss: 374 kB466 Pss_Dirty: 0 kB467 Shared_Clean: 892 kB468 Shared_Dirty: 0 kB469 Private_Clean: 0 kB470 Private_Dirty: 0 kB471 Referenced: 892 kB472 Anonymous: 0 kB473 KSM: 0 kB474 LazyFree: 0 kB475 AnonHugePages: 0 kB476 ShmemPmdMapped: 0 kB477 Shared_Hugetlb: 0 kB478 Private_Hugetlb: 0 kB479 Swap: 0 kB480 SwapPss: 0 kB481 KernelPageSize: 4 kB482 MMUPageSize: 4 kB483 Locked: 0 kB484 THPeligible: 0485 VmFlags: rd ex mr mw me dw486 487The first of these lines shows the same information as is displayed for the488mapping in /proc/PID/maps. Following lines show the size of the mapping489(size); the size of each page allocated when backing a VMA (KernelPageSize),490which is usually the same as the size in the page table entries; the page size491used by the MMU when backing a VMA (in most cases, the same as KernelPageSize);492the amount of the mapping that is currently resident in RAM (RSS); the493process' proportional share of this mapping (PSS); and the number of clean and494dirty shared and private pages in the mapping.495 496The "proportional set size" (PSS) of a process is the count of pages it has497in memory, where each page is divided by the number of processes sharing it.498So if a process has 1000 pages all to itself, and 1000 shared with one other499process, its PSS will be 1500. "Pss_Dirty" is the portion of PSS which500consists of dirty pages. ("Pss_Clean" is not included, but it can be501calculated by subtracting "Pss_Dirty" from "Pss".)502 503Note that even a page which is part of a MAP_SHARED mapping, but has only504a single pte mapped, i.e. is currently used by only one process, is accounted505as private and not as shared.506 507"Referenced" indicates the amount of memory currently marked as referenced or508accessed.509 510"Anonymous" shows the amount of memory that does not belong to any file. Even511a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE512and a page is modified, the file page is replaced by a private anonymous copy.513 514"KSM" reports how many of the pages are KSM pages. Note that KSM-placed zeropages515are not included, only actual KSM pages.516 517"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE).518The memory isn't freed immediately with madvise(). It's freed in memory519pressure if the memory is clean. Please note that the printed value might520be lower than the real value due to optimizations used in the current521implementation. If this is not desirable please file a bug report.522 523"AnonHugePages" shows the amount of memory backed by transparent hugepage.524 525"ShmemPmdMapped" shows the amount of shared (shmem/tmpfs) memory backed by526huge pages.527 528"Shared_Hugetlb" and "Private_Hugetlb" show the amounts of memory backed by529hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical530reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field.531 532"Swap" shows how much would-be-anonymous memory is also used, but out on swap.533 534For shmem mappings, "Swap" includes also the size of the mapped (and not535replaced by copy-on-write) part of the underlying shmem object out on swap.536"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this537does not take into account swapped out page of underlying shmem objects.538"Locked" indicates whether the mapping is locked in memory or not.539 540"THPeligible" indicates whether the mapping is eligible for allocating541naturally aligned THP pages of any currently enabled size. 1 if true, 0542otherwise.543 544"VmFlags" field deserves a separate description. This member represents the545kernel flags associated with the particular virtual memory area in two letter546encoded manner. The codes are the following:547 548 == =======================================549 rd readable550 wr writeable551 ex executable552 sh shared553 mr may read554 mw may write555 me may execute556 ms may share557 gd stack segment growns down558 pf pure PFN range559 dw disabled write to the mapped file560 lo pages are locked in memory561 io memory mapped I/O area562 sr sequential read advise provided563 rr random read advise provided564 dc do not copy area on fork565 de do not expand area on remapping566 ac area is accountable567 nr swap space is not reserved for the area568 ht area uses huge tlb pages569 sf synchronous page fault570 ar architecture specific flag571 wf wipe on fork572 dd do not include area into core dump573 sd soft dirty flag574 mm mixed map area575 hg huge page advise flag576 nh no huge page advise flag577 mg mergeable advise flag578 bt arm64 BTI guarded page579 mt arm64 MTE allocation tags are enabled580 um userfaultfd missing tracking581 uw userfaultfd wr-protect tracking582 ss shadow stack page583 sl sealed584 == =======================================585 586Note that there is no guarantee that every flag and associated mnemonic will587be present in all further kernel releases. Things get changed, the flags may588be vanished or the reverse -- new added. Interpretation of their meaning589might change in future as well. So each consumer of these flags has to590follow each specific kernel version for the exact semantic.591 592This file is only present if the CONFIG_MMU kernel configuration option is593enabled.594 595Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent596output can be achieved only in the single read call).597 598This typically manifests when doing partial reads of these files while the599memory map is being modified. Despite the races, we do provide the following600guarantees:601 6021) The mapped addresses never go backwards, which implies no two603 regions will ever overlap.6042) If there is something at a given vaddr during the entirety of the605 life of the smaps/maps walk, there will be some output for it.606 607The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps,608but their values are the sums of the corresponding values for all mappings of609the process. Additionally, it contains these fields:610 611- Pss_Anon612- Pss_File613- Pss_Shmem614 615They represent the proportional shares of anonymous, file, and shmem pages, as616described for smaps above. These fields are omitted in smaps since each617mapping identifies the type (anon, file, or shmem) of all pages it contains.618Thus all information in smaps_rollup can be derived from smaps, but at a619significantly higher cost.620 621The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG622bits on both physical and virtual pages associated with a process, and the623soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst624for details).625To clear the bits for all the pages associated with the process::626 627 > echo 1 > /proc/PID/clear_refs628 629To clear the bits for the anonymous pages associated with the process::630 631 > echo 2 > /proc/PID/clear_refs632 633To clear the bits for the file mapped pages associated with the process::634 635 > echo 3 > /proc/PID/clear_refs636 637To clear the soft-dirty bit::638 639 > echo 4 > /proc/PID/clear_refs640 641To reset the peak resident set size ("high water mark") to the process's642current value::643 644 > echo 5 > /proc/PID/clear_refs645 646Any other value written to /proc/PID/clear_refs will have no effect.647 648The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags649using /proc/kpageflags and number of times a page is mapped using650/proc/kpagecount. For detailed explanation, see651Documentation/admin-guide/mm/pagemap.rst.652 653The /proc/pid/numa_maps is an extension based on maps, showing the memory654locality and binding policy, as well as the memory usage (in pages) of655each mapping. The output follows a general format where mapping details get656summarized separated by blank spaces, one mapping per each file line::657 658 address policy mapping details659 660 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4661 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4662 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4663 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4664 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4665 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4666 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4667 320698b000 default file=/lib64/libc-2.12.so668 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4669 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4670 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4671 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4672 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4673 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048674 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4675 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4676 677Where:678 679"address" is the starting address for the mapping;680 681"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst);682 683"mapping details" summarizes mapping data such as mapping type, page usage counters,684node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page685size, in KB, that is backing the mapping up.686 6871.2 Kernel data688---------------689 690Similar to the process entries, the kernel data files give information about691the running kernel. The files used to obtain this information are contained in692/proc and are listed in Table 1-5. Not all of these will be present in your693system. It depends on the kernel configuration and the loaded modules, which694files are there, and which are missing.695 696.. table:: Table 1-5: Kernel info in /proc697 698 ============ ===============================================================699 File Content700 ============ ===============================================================701 allocinfo Memory allocations profiling information702 apm Advanced power management info703 bootconfig Kernel command line obtained from boot config,704 and, if there were kernel parameters from the705 boot loader, a "# Parameters from bootloader:"706 line followed by a line containing those707 parameters prefixed by "# ". (5.5)708 buddyinfo Kernel memory allocator information (see text) (2.5)709 bus Directory containing bus specific information710 cmdline Kernel command line, both from bootloader and embedded711 in the kernel image712 cpuinfo Info about the CPU713 devices Available devices (block and character)714 dma Used DMS channels715 filesystems Supported filesystems716 driver Various drivers grouped here, currently rtc (2.4)717 execdomains Execdomains, related to security (2.4)718 fb Frame Buffer devices (2.4)719 fs File system parameters, currently nfs/exports (2.4)720 ide Directory containing info about the IDE subsystem721 interrupts Interrupt usage722 iomem Memory map (2.4)723 ioports I/O port usage724 irq Masks for irq to cpu affinity (2.4)(smp?)725 isapnp ISA PnP (Plug&Play) Info (2.4)726 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4))727 kmsg Kernel messages728 ksyms Kernel symbol table729 loadavg Load average of last 1, 5 & 15 minutes;730 number of processes currently runnable (running or on ready queue);731 total number of processes in system;732 last pid created.733 All fields are separated by one space except "number of734 processes currently runnable" and "total number of processes735 in system", which are separated by a slash ('/'). Example:736 0.61 0.61 0.55 3/828 22084737 locks Kernel locks738 meminfo Memory info739 misc Miscellaneous740 modules List of loaded modules741 mounts Mounted filesystems742 net Networking info (see text)743 pagetypeinfo Additional page allocator information (see text) (2.5)744 partitions Table of partitions known to the system745 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/,746 decoupled by lspci (2.4)747 rtc Real time clock748 scsi SCSI info (see text)749 slabinfo Slab pool info750 softirqs softirq usage751 stat Overall statistics752 swaps Swap space utilization753 sys See chapter 2754 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4)755 tty Info of tty drivers756 uptime Wall clock since boot, combined idle time of all cpus757 version Kernel version758 video bttv info of video resources (2.4)759 vmallocinfo Show vmalloced areas760 ============ ===============================================================761 762You can, for example, check which interrupts are currently in use and what763they are used for by looking in the file /proc/interrupts::764 765 > cat /proc/interrupts766 CPU0767 0: 8728810 XT-PIC timer768 1: 895 XT-PIC keyboard769 2: 0 XT-PIC cascade770 3: 531695 XT-PIC aha152x771 4: 2014133 XT-PIC serial772 5: 44401 XT-PIC pcnet_cs773 8: 2 XT-PIC rtc774 11: 8 XT-PIC i82365775 12: 182918 XT-PIC PS/2 Mouse776 13: 1 XT-PIC fpu777 14: 1232265 XT-PIC ide0778 15: 7 XT-PIC ide1779 NMI: 0780 781In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the782output of a SMP machine)::783 784 > cat /proc/interrupts785 786 CPU0 CPU1787 0: 1243498 1214548 IO-APIC-edge timer788 1: 8949 8958 IO-APIC-edge keyboard789 2: 0 0 XT-PIC cascade790 5: 11286 10161 IO-APIC-edge soundblaster791 8: 1 0 IO-APIC-edge rtc792 9: 27422 27407 IO-APIC-edge 3c503793 12: 113645 113873 IO-APIC-edge PS/2 Mouse794 13: 0 0 XT-PIC fpu795 14: 22491 24012 IO-APIC-edge ide0796 15: 2183 2415 IO-APIC-edge ide1797 17: 30564 30414 IO-APIC-level eth0798 18: 177 164 IO-APIC-level bttv799 NMI: 2457961 2457959800 LOC: 2457882 2457881801 ERR: 2155802 803NMI is incremented in this case because every timer interrupt generates a NMI804(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups.805 806LOC is the local interrupt counter of the internal APIC of every CPU.807 808ERR is incremented in the case of errors in the IO-APIC bus (the bus that809connects the CPUs in a SMP system. This means that an error has been detected,810the IO-APIC automatically retry the transmission, so it should not be a big811problem, but you should read the SMP-FAQ.812 813In 2.6.2* /proc/interrupts was expanded again. This time the goal was for814/proc/interrupts to display every IRQ vector in use by the system, not815just those considered 'most important'. The new vectors are:816 817THR818 interrupt raised when a machine check threshold counter819 (typically counting ECC corrected errors of memory or cache) exceeds820 a configurable threshold. Only available on some systems.821 822TRM823 a thermal event interrupt occurs when a temperature threshold824 has been exceeded for the CPU. This interrupt may also be generated825 when the temperature drops back to normal.826 827SPU828 a spurious interrupt is some interrupt that was raised then lowered829 by some IO device before it could be fully processed by the APIC. Hence830 the APIC sees the interrupt but does not know what device it came from.831 For this case the APIC will generate the interrupt with a IRQ vector832 of 0xff. This might also be generated by chipset bugs.833 834RES, CAL, TLB835 rescheduling, call and TLB flush interrupts are836 sent from one CPU to another per the needs of the OS. Typically,837 their statistics are used by kernel developers and interested users to838 determine the occurrence of interrupts of the given type.839 840The above IRQ vectors are displayed only when relevant. For example,841the threshold vector does not exist on x86_64 platforms. Others are842suppressed when the system is a uniprocessor. As of this writing, only843i386 and x86_64 platforms support the new IRQ vector displays.844 845Of some interest is the introduction of the /proc/irq directory to 2.4.846It could be used to set IRQ to CPU affinity. This means that you can "hook" an847IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the848irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and849prof_cpu_mask.850 851For example::852 853 > ls /proc/irq/854 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask855 1 11 13 15 17 19 3 5 7 9 default_smp_affinity856 > ls /proc/irq/0/857 smp_affinity858 859smp_affinity is a bitmask, in which you can specify which CPUs can handle the860IRQ. You can set it by doing::861 862 > echo 1 > /proc/irq/10/smp_affinity863 864This means that only the first CPU will handle the IRQ, but you can also echo8655 which means that only the first and third CPU can handle the IRQ.866 867The contents of each smp_affinity file is the same by default::868 869 > cat /proc/irq/0/smp_affinity870 ffffffff871 872There is an alternate interface, smp_affinity_list which allows specifying873a CPU range instead of a bitmask::874 875 > cat /proc/irq/0/smp_affinity_list876 1024-1031877 878The default_smp_affinity mask applies to all non-active IRQs, which are the879IRQs which have not yet been allocated/activated, and hence which lack a880/proc/irq/[0-9]* directory.881 882The node file on an SMP system shows the node to which the device using the IRQ883reports itself as being attached. This hardware locality information does not884include information about any possible driver locality preference.885 886prof_cpu_mask specifies which CPUs are to be profiled by the system wide887profiler. Default value is ffffffff (all CPUs if there are only 32 of them).888 889The way IRQs are routed is handled by the IO-APIC, and it's Round Robin890between all the CPUs which are allowed to handle it. As usual the kernel has891more info than you and does a better job than you, so the defaults are the892best choice for almost everyone. [Note this applies only to those IO-APIC's893that support "Round Robin" interrupt distribution.]894 895There are three more important subdirectories in /proc: net, scsi, and sys.896The general rule is that the contents, or even the existence of these897directories, depend on your kernel configuration. If SCSI is not enabled, the898directory scsi may not exist. The same is true with the net, which is there899only when networking support is present in the running kernel.900 901The slabinfo file gives information about memory usage at the slab level.902Linux uses slab pools for memory management above page level in version 2.2.903Commonly used objects have their own slab pool (such as network buffers,904directory cache, and so on).905 906::907 908 > cat /proc/buddyinfo909 910 Node 0, zone DMA 0 4 5 4 4 3 ...911 Node 0, zone Normal 1 0 0 1 101 8 ...912 Node 0, zone HighMem 2 0 0 1 1 0 ...913 914External fragmentation is a problem under some workloads, and buddyinfo is a915useful tool for helping diagnose these problems. Buddyinfo will give you a916clue as to how big an area you can safely allocate, or why a previous917allocation failed.918 919Each column represents the number of pages of a certain order which are920available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in921ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE922available in ZONE_NORMAL, etc...923 924More information relevant to external fragmentation can be found in925pagetypeinfo::926 927 > cat /proc/pagetypeinfo928 Page block order: 9929 Pages per block: 512930 931 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10932 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0933 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0934 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2935 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0936 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0937 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9938 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0939 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452940 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0941 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0942 943 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate944 Node 0, zone DMA 2 0 5 1 0945 Node 0, zone DMA32 41 6 967 2 0946 947Fragmentation avoidance in the kernel works by grouping pages of different948migrate types into the same contiguous regions of memory called page blocks.949A page block is typically the size of the default hugepage size, e.g. 2MB on950X86-64. By keeping pages grouped based on their ability to move, the kernel951can reclaim pages within a page block to satisfy a high-order allocation.952 953The pagetypinfo begins with information on the size of a page block. It954then gives the same type of information as buddyinfo except broken down955by migrate-type and finishes with details on how many page blocks of each956type exist.957 958If min_free_kbytes has been tuned correctly (recommendations made by hugeadm959from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can960make an estimate of the likely number of huge pages that can be allocated961at a given point in time. All the "Movable" blocks should be allocatable962unless memory has been mlock()'d. Some of the Reclaimable blocks should963also be allocatable although a lot of filesystem metadata may have to be964reclaimed to achieve this.965 966 967allocinfo968~~~~~~~~~969 970Provides information about memory allocations at all locations in the code971base. Each allocation in the code is identified by its source file, line972number, module (if originates from a loadable module) and the function calling973the allocation. The number of bytes allocated and number of calls at each974location are reported. The first line indicates the version of the file, the975second line is the header listing fields in the file.976 977Example output.978 979::980 981 > tail -n +3 /proc/allocinfo | sort -rn982 127664128 31168 mm/page_ext.c:270 func:alloc_page_ext983 56373248 4737 mm/slub.c:2259 func:alloc_slab_page984 14880768 3633 mm/readahead.c:247 func:page_cache_ra_unbounded985 14417920 3520 mm/mm_init.c:2530 func:alloc_large_system_hash986 13377536 234 block/blk-mq.c:3421 func:blk_mq_alloc_rqs987 11718656 2861 mm/filemap.c:1919 func:__filemap_get_folio988 9192960 2800 kernel/fork.c:307 func:alloc_thread_stack_node989 4206592 4 net/netfilter/nf_conntrack_core.c:2567 func:nf_ct_alloc_hashtable990 4136960 1010 drivers/staging/ctagmod/ctagmod.c:20 [ctagmod] func:ctagmod_start991 3940352 962 mm/memory.c:4214 func:alloc_anon_folio992 2894464 22613 fs/kernfs/dir.c:615 func:__kernfs_new_node993 ...994 995 996meminfo997~~~~~~~998 999Provides information about distribution and utilization of memory. This1000varies by architecture and compile options. Some of the counters reported1001here overlap. The memory reported by the non overlapping counters may not1002add up to the overall memory usage and the difference for some workloads1003can be substantial. In many cases there are other means to find out1004additional memory using subsystem specific interfaces, for instance1005/proc/net/sockstat for TCP memory allocations.1006 1007Example output. You may not have all of these fields.1008 1009::1010 1011 > cat /proc/meminfo1012 1013 MemTotal: 32858820 kB1014 MemFree: 21001236 kB1015 MemAvailable: 27214312 kB1016 Buffers: 581092 kB1017 Cached: 5587612 kB1018 SwapCached: 0 kB1019 Active: 3237152 kB1020 Inactive: 7586256 kB1021 Active(anon): 94064 kB1022 Inactive(anon): 4570616 kB1023 Active(file): 3143088 kB1024 Inactive(file): 3015640 kB1025 Unevictable: 0 kB1026 Mlocked: 0 kB1027 SwapTotal: 0 kB1028 SwapFree: 0 kB1029 Zswap: 1904 kB1030 Zswapped: 7792 kB1031 Dirty: 12 kB1032 Writeback: 0 kB1033 AnonPages: 4654780 kB1034 Mapped: 266244 kB1035 Shmem: 9976 kB1036 KReclaimable: 517708 kB1037 Slab: 660044 kB1038 SReclaimable: 517708 kB1039 SUnreclaim: 142336 kB1040 KernelStack: 11168 kB1041 PageTables: 20540 kB1042 SecPageTables: 0 kB1043 NFS_Unstable: 0 kB1044 Bounce: 0 kB1045 WritebackTmp: 0 kB1046 CommitLimit: 16429408 kB1047 Committed_AS: 7715148 kB1048 VmallocTotal: 34359738367 kB1049 VmallocUsed: 40444 kB1050 VmallocChunk: 0 kB1051 Percpu: 29312 kB1052 EarlyMemtestBad: 0 kB1053 HardwareCorrupted: 0 kB1054 AnonHugePages: 4149248 kB1055 ShmemHugePages: 0 kB1056 ShmemPmdMapped: 0 kB1057 FileHugePages: 0 kB1058 FilePmdMapped: 0 kB1059 CmaTotal: 0 kB1060 CmaFree: 0 kB1061 HugePages_Total: 01062 HugePages_Free: 01063 HugePages_Rsvd: 01064 HugePages_Surp: 01065 Hugepagesize: 2048 kB1066 Hugetlb: 0 kB1067 DirectMap4k: 401152 kB1068 DirectMap2M: 10008576 kB1069 DirectMap1G: 24117248 kB1070 1071MemTotal1072 Total usable RAM (i.e. physical RAM minus a few reserved1073 bits and the kernel binary code)1074MemFree1075 Total free RAM. On highmem systems, the sum of LowFree+HighFree1076MemAvailable1077 An estimate of how much memory is available for starting new1078 applications, without swapping. Calculated from MemFree,1079 SReclaimable, the size of the file LRU lists, and the low1080 watermarks in each zone.1081 The estimate takes into account that the system needs some1082 page cache to function well, and that not all reclaimable1083 slab will be reclaimable, due to items being in use. The1084 impact of those factors will vary from system to system.1085Buffers1086 Relatively temporary storage for raw disk blocks1087 shouldn't get tremendously large (20MB or so)1088Cached1089 In-memory cache for files read from the disk (the1090 pagecache) as well as tmpfs & shmem.1091 Doesn't include SwapCached.1092SwapCached1093 Memory that once was swapped out, is swapped back in but1094 still also is in the swapfile (if memory is needed it1095 doesn't need to be swapped out AGAIN because it is already1096 in the swapfile. This saves I/O)1097Active1098 Memory that has been used more recently and usually not1099 reclaimed unless absolutely necessary.1100Inactive1101 Memory which has been less recently used. It is more1102 eligible to be reclaimed for other purposes1103Unevictable1104 Memory allocated for userspace which cannot be reclaimed, such1105 as mlocked pages, ramfs backing pages, secret memfd pages etc.1106Mlocked1107 Memory locked with mlock().1108HighTotal, HighFree1109 Highmem is all memory above ~860MB of physical memory.1110 Highmem areas are for use by userspace programs, or1111 for the pagecache. The kernel must use tricks to access1112 this memory, making it slower to access than lowmem.1113LowTotal, LowFree1114 Lowmem is memory which can be used for everything that1115 highmem can be used for, but it is also available for the1116 kernel's use for its own data structures. Among many1117 other things, it is where everything from the Slab is1118 allocated. Bad things happen when you're out of lowmem.1119SwapTotal1120 total amount of swap space available1121SwapFree1122 Memory which has been evicted from RAM, and is temporarily1123 on the disk1124Zswap1125 Memory consumed by the zswap backend (compressed size)1126Zswapped1127 Amount of anonymous memory stored in zswap (original size)1128Dirty1129 Memory which is waiting to get written back to the disk1130Writeback1131 Memory which is actively being written back to the disk1132AnonPages1133 Non-file backed pages mapped into userspace page tables1134Mapped1135 files which have been mmapped, such as libraries1136Shmem1137 Total memory used by shared memory (shmem) and tmpfs1138KReclaimable1139 Kernel allocations that the kernel will attempt to reclaim1140 under memory pressure. Includes SReclaimable (below), and other1141 direct allocations with a shrinker.1142Slab1143 in-kernel data structures cache1144SReclaimable1145 Part of Slab, that might be reclaimed, such as caches1146SUnreclaim1147 Part of Slab, that cannot be reclaimed on memory pressure1148KernelStack1149 Memory consumed by the kernel stacks of all tasks1150PageTables1151 Memory consumed by userspace page tables1152SecPageTables1153 Memory consumed by secondary page tables, this currently includes1154 KVM mmu and IOMMU allocations on x86 and arm64.1155NFS_Unstable1156 Always zero. Previous counted pages which had been written to1157 the server, but has not been committed to stable storage.1158Bounce1159 Memory used for block device "bounce buffers"1160WritebackTmp1161 Memory used by FUSE for temporary writeback buffers1162CommitLimit1163 Based on the overcommit ratio ('vm.overcommit_ratio'),1164 this is the total amount of memory currently available to1165 be allocated on the system. This limit is only adhered to1166 if strict overcommit accounting is enabled (mode 2 in1167 'vm.overcommit_memory').1168 1169 The CommitLimit is calculated with the following formula::1170 1171 CommitLimit = ([total RAM pages] - [total huge TLB pages]) *1172 overcommit_ratio / 100 + [total swap pages]1173 1174 For example, on a system with 1G of physical RAM and 7G1175 of swap with a `vm.overcommit_ratio` of 30 it would1176 yield a CommitLimit of 7.3G.1177 1178 For more details, see the memory overcommit documentation1179 in mm/overcommit-accounting.1180Committed_AS1181 The amount of memory presently allocated on the system.1182 The committed memory is a sum of all of the memory which1183 has been allocated by processes, even if it has not been1184 "used" by them as of yet. A process which malloc()'s 1G1185 of memory, but only touches 300M of it will show up as1186 using 1G. This 1G is memory which has been "committed" to1187 by the VM and can be used at any time by the allocating1188 application. With strict overcommit enabled on the system1189 (mode 2 in 'vm.overcommit_memory'), allocations which would1190 exceed the CommitLimit (detailed above) will not be permitted.1191 This is useful if one needs to guarantee that processes will1192 not fail due to lack of memory once that memory has been1193 successfully allocated.1194VmallocTotal1195 total size of vmalloc virtual address space1196VmallocUsed1197 amount of vmalloc area which is used1198VmallocChunk1199 largest contiguous block of vmalloc area which is free1200Percpu1201 Memory allocated to the percpu allocator used to back percpu1202 allocations. This stat excludes the cost of metadata.1203EarlyMemtestBad1204 The amount of RAM/memory in kB, that was identified as corrupted1205 by early memtest. If memtest was not run, this field will not1206 be displayed at all. Size is never rounded down to 0 kB.1207 That means if 0 kB is reported, you can safely assume1208 there was at least one pass of memtest and none of the passes1209 found a single faulty byte of RAM.1210HardwareCorrupted1211 The amount of RAM/memory in KB, the kernel identifies as1212 corrupted.1213AnonHugePages1214 Non-file backed huge pages mapped into userspace page tables1215ShmemHugePages1216 Memory used by shared memory (shmem) and tmpfs allocated1217 with huge pages1218ShmemPmdMapped1219 Shared memory mapped into userspace with huge pages1220FileHugePages1221 Memory used for filesystem data (page cache) allocated1222 with huge pages1223FilePmdMapped1224 Page cache mapped into userspace with huge pages1225CmaTotal1226 Memory reserved for the Contiguous Memory Allocator (CMA)1227CmaFree1228 Free remaining memory in the CMA reserves1229HugePages_Total, HugePages_Free, HugePages_Rsvd, HugePages_Surp, Hugepagesize, Hugetlb1230 See Documentation/admin-guide/mm/hugetlbpage.rst.1231DirectMap4k, DirectMap2M, DirectMap1G1232 Breakdown of page table sizes used in the kernel's1233 identity mapping of RAM1234 1235vmallocinfo1236~~~~~~~~~~~1237 1238Provides information about vmalloced/vmaped areas. One line per area,1239containing the virtual address range of the area, size in bytes,1240caller information of the creator, and optional information depending1241on the kind of area:1242 1243 ========== ===================================================1244 pages=nr number of pages1245 phys=addr if a physical address was specified1246 ioremap I/O mapping (ioremap() and friends)1247 vmalloc vmalloc() area1248 vmap vmap()ed pages1249 user VM_USERMAP area1250 vpages buffer for pages pointers was vmalloced (huge area)1251 N<node>=nr (Only on NUMA kernels)1252 Number of pages allocated on memory node <node>1253 ========== ===================================================1254 1255::1256 1257 > cat /proc/vmallocinfo1258 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ...1259 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=1281260 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ...1261 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=641262 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f...1263 phys=7fee8000 ioremap1264 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f...1265 phys=7fee7000 ioremap1266 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x2101267 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ...1268 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=31269 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ...1270 pages=2 vmalloc N1=21271 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ...1272 /0x130 [x_tables] pages=4 vmalloc N0=41273 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ...1274 pages=14 vmalloc N2=141275 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ...1276 pages=4 vmalloc N1=41277 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ...1278 pages=2 vmalloc N1=21279 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ...1280 pages=10 vmalloc N0=101281 1282 1283softirqs1284~~~~~~~~1285 1286Provides counts of softirq handlers serviced since boot time, for each CPU.1287 1288::1289 1290 > cat /proc/softirqs1291 CPU0 CPU1 CPU2 CPU31292 HI: 0 0 0 01293 TIMER: 27166 27120 27097 270341294 NET_TX: 0 0 0 171295 NET_RX: 42 0 0 391296 BLOCK: 0 0 107 11211297 TASKLET: 0 0 0 2901298 SCHED: 27035 26983 26971 267461299 HRTIMER: 0 0 0 01300 RCU: 1678 1769 2178 22501301 13021.3 Networking info in /proc/net1303--------------------------------1304 1305The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the1306additional values you get for IP version 6 if you configure the kernel to1307support this. Table 1-9 lists the files and their meaning.1308 1309 1310.. table:: Table 1-8: IPv6 info in /proc/net1311 1312 ========== =====================================================1313 File Content1314 ========== =====================================================1315 udp6 UDP sockets (IPv6)1316 tcp6 TCP sockets (IPv6)1317 raw6 Raw device statistics (IPv6)1318 igmp6 IP multicast addresses, which this host joined (IPv6)1319 if_inet6 List of IPv6 interface addresses1320 ipv6_route Kernel routing table for IPv61321 rt6_stats Global IPv6 routing tables statistics1322 sockstat6 Socket statistics (IPv6)1323 snmp6 Snmp data (IPv6)1324 ========== =====================================================1325 1326.. table:: Table 1-9: Network info in /proc/net1327 1328 ============= ================================================================1329 File Content1330 ============= ================================================================1331 arp Kernel ARP table1332 dev network devices with statistics1333 dev_mcast the Layer2 multicast groups a device is listening too1334 (interface index, label, number of references, number of bound1335 addresses).1336 dev_stat network device status1337 ip_fwchains Firewall chain linkage1338 ip_fwnames Firewall chain names1339 ip_masq Directory containing the masquerading tables1340 ip_masquerade Major masquerading table1341 netstat Network statistics1342 raw raw device statistics1343 route Kernel routing table1344 rpc Directory containing rpc info1345 rt_cache Routing cache1346 snmp SNMP data1347 sockstat Socket statistics1348 softnet_stat Per-CPU incoming packets queues statistics of online CPUs1349 tcp TCP sockets1350 udp UDP sockets1351 unix UNIX domain sockets1352 wireless Wireless interface data (Wavelan etc)1353 igmp IP multicast addresses, which this host joined1354 psched Global packet scheduler parameters.1355 netlink List of PF_NETLINK sockets1356 ip_mr_vifs List of multicast virtual interfaces1357 ip_mr_cache List of multicast routing cache1358 ============= ================================================================1359 1360You can use this information to see which network devices are available in1361your system and how much traffic was routed over those devices::1362 1363 > cat /proc/net/dev1364 Inter-|Receive |[...1365 face |bytes packets errs drop fifo frame compressed multicast|[...1366 lo: 908188 5596 0 0 0 0 0 0 [...1367 ppp0:15475140 20721 410 0 0 410 0 0 [...1368 eth0: 614530 7085 0 0 0 0 0 1 [...1369 1370 ...] Transmit1371 ...] bytes packets errs drop fifo colls carrier compressed1372 ...] 908188 5596 0 0 0 0 0 01373 ...] 1375103 17405 0 0 0 0 0 01374 ...] 1703981 5535 0 0 0 3 0 01375 1376In addition, each Channel Bond interface has its own directory. For1377example, the bond0 device will have a directory called /proc/net/bond0/.1378It will contain information that is specific to that bond, such as the1379current slaves of the bond, the link status of the slaves, and how1380many times the slaves link has failed.1381 13821.4 SCSI info1383-------------1384 1385If you have a SCSI or ATA host adapter in your system, you'll find a1386subdirectory named after the driver for this adapter in /proc/scsi.1387You'll also see a list of all recognized SCSI devices in /proc/scsi::1388 1389 >cat /proc/scsi/scsi1390 Attached devices:1391 Host: scsi0 Channel: 00 Id: 00 Lun: 001392 Vendor: IBM Model: DGHS09U Rev: 03E01393 Type: Direct-Access ANSI SCSI revision: 031394 Host: scsi0 Channel: 00 Id: 06 Lun: 001395 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.041396 Type: CD-ROM ANSI SCSI revision: 021397 1398 1399The directory named after the driver has one file for each adapter found in1400the system. These files contain information about the controller, including1401the used IRQ and the IO address range. The amount of information shown is1402dependent on the adapter you use. The example shows the output for an Adaptec1403AHA-2940 SCSI adapter::1404 1405 > cat /proc/scsi/aic7xxx/01406 1407 Adaptec AIC7xxx driver version: 5.1.19/3.2.41408 Compile Options:1409 TCQ Enabled By Default : Disabled1410 AIC7XXX_PROC_STATS : Disabled1411 AIC7XXX_RESET_DELAY : 51412 Adapter Configuration:1413 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter1414 Ultra Wide Controller1415 PCI MMAPed I/O Base: 0xeb0010001416 Adapter SEEPROM Config: SEEPROM found and used.1417 Adaptec SCSI BIOS: Enabled1418 IRQ: 101419 SCBs: Active 0, Max Active 2,1420 Allocated 15, HW 16, Page 2551421 Interrupts: 1603281422 BIOS Control Word: 0x18b61423 Adapter Control Word: 0x005b1424 Extended Translation: Enabled1425 Disconnect Enable Flags: 0xffff1426 Ultra Enable Flags: 0x00011427 Tag Queue Enable Flags: 0x00001428 Ordered Queue Tag Flags: 0x00001429 Default Tag Queue Depth: 81430 Tagged Queue By Device array for aic7xxx host instance 0:1431 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}1432 Actual queue depth per device for aic7xxx host instance 0:1433 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}1434 Statistics:1435 (scsi0:0:0:0)1436 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 81437 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0)1438 Total transfers 160151 (74577 reads and 85574 writes)1439 (scsi0:0:6:0)1440 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 151441 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0)1442 Total transfers 0 (0 reads and 0 writes)1443 1444 14451.5 Parallel port info in /proc/parport1446---------------------------------------1447 1448The directory /proc/parport contains information about the parallel ports of1449your system. It has one subdirectory for each port, named after the port1450number (0,1,2,...).1451 1452These directories contain the four files shown in Table 1-10.1453 1454 1455.. table:: Table 1-10: Files in /proc/parport1456 1457 ========= ====================================================================1458 File Content1459 ========= ====================================================================1460 autoprobe Any IEEE-1284 device ID information that has been acquired.1461 devices list of the device drivers using that port. A + will appear by the1462 name of the device currently using the port (it might not appear1463 against any).1464 hardware Parallel port's base address, IRQ line and DMA channel.1465 irq IRQ that parport is using for that port. This is in a separate1466 file to allow you to alter it by writing a new value in (IRQ1467 number or none).1468 ========= ====================================================================1469 14701.6 TTY info in /proc/tty1471-------------------------1472 1473Information about the available and actually used tty's can be found in the1474directory /proc/tty. You'll find entries for drivers and line disciplines in1475this directory, as shown in Table 1-11.1476 1477 1478.. table:: Table 1-11: Files in /proc/tty1479 1480 ============= ==============================================1481 File Content1482 ============= ==============================================1483 drivers list of drivers and their usage1484 ldiscs registered line disciplines1485 driver/serial usage statistic and status of single tty lines1486 ============= ==============================================1487 1488To see which tty's are currently in use, you can simply look into the file1489/proc/tty/drivers::1490 1491 > cat /proc/tty/drivers1492 pty_slave /dev/pts 136 0-255 pty:slave1493 pty_master /dev/ptm 128 0-255 pty:master1494 pty_slave /dev/ttyp 3 0-255 pty:slave1495 pty_master /dev/pty 2 0-255 pty:master1496 serial /dev/cua 5 64-67 serial:callout1497 serial /dev/ttyS 4 64-67 serial1498 /dev/tty0 /dev/tty0 4 0 system:vtmaster1499 /dev/ptmx /dev/ptmx 5 2 system1500 /dev/console /dev/console 5 1 system:console1501 /dev/tty /dev/tty 5 0 system:/dev/tty1502 unknown /dev/tty 4 1-63 console1503 1504 15051.7 Miscellaneous kernel statistics in /proc/stat1506-------------------------------------------------1507 1508Various pieces of information about kernel activity are available in the1509/proc/stat file. All of the numbers reported in this file are aggregates1510since the system first booted. For a quick look, simply cat the file::1511 1512 > cat /proc/stat1513 cpu 237902850 368826709 106375398 1873517540 1135548 0 14507935 0 0 01514 cpu0 60045249 91891769 26331539 468411416 495718 0 5739640 0 0 01515 cpu1 59746288 91759249 26609887 468860630 312281 0 4384817 0 0 01516 cpu2 59489247 92985423 26904446 467808813 171668 0 2268998 0 0 01517 cpu3 58622065 92190267 26529524 468436680 155879 0 2114478 0 0 01518 intr 8688370575 8 3373 0 0 0 0 0 0 1 40791 0 0 353317 0 0 0 0 224789828 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 190974333 41958554 123983334 43 0 224593 0 0 0 <more 0's deleted>1519 ctxt 228482210621520 btime 16053169991521 processes 7467871471522 procs_running 21523 procs_blocked 01524 softirq 12121874454 100099120 3938138295 127375644 2795979 187870761 0 173808342 3072582055 52608 2241843541525 1526The very first "cpu" line aggregates the numbers in all of the other "cpuN"1527lines. These numbers identify the amount of time the CPU has spent performing1528different kinds of work. Time units are in USER_HZ (typically hundredths of a1529second). The meanings of the columns are as follows, from left to right:1530 1531- user: normal processes executing in user mode1532- nice: niced processes executing in user mode1533- system: processes executing in kernel mode1534- idle: twiddling thumbs1535- iowait: In a word, iowait stands for waiting for I/O to complete. But there1536 are several problems:1537 1538 1. CPU will not wait for I/O to complete, iowait is the time that a task is1539 waiting for I/O to complete. When CPU goes into idle state for1540 outstanding task I/O, another task will be scheduled on this CPU.1541 2. In a multi-core CPU, the task waiting for I/O to complete is not running1542 on any CPU, so the iowait of each CPU is difficult to calculate.1543 3. The value of iowait field in /proc/stat will decrease in certain1544 conditions.1545 1546 So, the iowait is not reliable by reading from /proc/stat.1547- irq: servicing interrupts1548- softirq: servicing softirqs1549- steal: involuntary wait1550- guest: running a normal guest1551- guest_nice: running a niced guest1552 1553The "intr" line gives counts of interrupts serviced since boot time, for each1554of the possible system interrupts. The first column is the total of all1555interrupts serviced including unnumbered architecture specific interrupts;1556each subsequent column is the total for that particular numbered interrupt.1557Unnumbered interrupts are not shown, only summed into the total.1558 1559The "ctxt" line gives the total number of context switches across all CPUs.1560 1561The "btime" line gives the time at which the system booted, in seconds since1562the Unix epoch.1563 1564The "processes" line gives the number of processes and threads created, which1565includes (but is not limited to) those created by calls to the fork() and1566clone() system calls.1567 1568The "procs_running" line gives the total number of threads that are1569running or ready to run (i.e., the total number of runnable threads).1570 1571The "procs_blocked" line gives the number of processes currently blocked,1572waiting for I/O to complete.1573 1574The "softirq" line gives counts of softirqs serviced since boot time, for each1575of the possible system softirqs. The first column is the total of all1576softirqs serviced; each subsequent column is the total for that particular1577softirq.1578 1579 15801.8 Ext4 file system parameters1581-------------------------------1582 1583Information about mounted ext4 file systems can be found in1584/proc/fs/ext4. Each mounted filesystem will have a directory in1585/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or1586/proc/fs/ext4/sda9 or /proc/fs/ext4/dm-0). The files in each per-device1587directory are shown in Table 1-12, below.1588 1589.. table:: Table 1-12: Files in /proc/fs/ext4/<devname>1590 1591 ============== ==========================================================1592 File Content1593 mb_groups details of multiblock allocator buddy cache of free blocks1594 ============== ==========================================================1595 15961.9 /proc/consoles1597-------------------1598Shows registered system console lines.1599 1600To see which character device lines are currently used for the system console1601/dev/console, you may simply look into the file /proc/consoles::1602 1603 > cat /proc/consoles1604 tty0 -WU (ECp) 4:71605 ttyS0 -W- (Ep) 4:641606 1607The columns are:1608 1609+--------------------+-------------------------------------------------------+1610| device | name of the device |1611+====================+=======================================================+1612| operations | * R = can do read operations |1613| | * W = can do write operations |1614| | * U = can do unblank |1615+--------------------+-------------------------------------------------------+1616| flags | * E = it is enabled |1617| | * C = it is preferred console |1618| | * B = it is primary boot console |1619| | * p = it is used for printk buffer |1620| | * b = it is not a TTY but a Braille device |1621| | * a = it is safe to use when cpu is offline |1622+--------------------+-------------------------------------------------------+1623| major:minor | major and minor number of the device separated by a |1624| | colon |1625+--------------------+-------------------------------------------------------+1626 1627Summary1628-------1629 1630The /proc file system serves information about the running system. It not only1631allows access to process data but also allows you to request the kernel status1632by reading files in the hierarchy.1633 1634The directory structure of /proc reflects the types of information and makes1635it easy, if not obvious, where to look for specific data.1636 1637Chapter 2: Modifying System Parameters1638======================================1639 1640In This Chapter1641---------------1642 1643* Modifying kernel parameters by writing into files found in /proc/sys1644* Exploring the files which modify certain parameters1645* Review of the /proc/sys file tree1646 1647------------------------------------------------------------------------------1648 1649A very interesting part of /proc is the directory /proc/sys. This is not only1650a source of information, it also allows you to change parameters within the1651kernel. Be very careful when attempting this. You can optimize your system,1652but you can also cause it to crash. Never alter kernel parameters on a1653production system. Set up a development machine and test to make sure that1654everything works the way you want it to. You may have no alternative but to1655reboot the machine once an error has been made.1656 1657To change a value, simply echo the new value into the file.1658You need to be root to do this. You can create your own boot script1659to perform this every time your system boots.1660 1661The files in /proc/sys can be used to fine tune and monitor miscellaneous and1662general things in the operation of the Linux kernel. Since some of the files1663can inadvertently disrupt your system, it is advisable to read both1664documentation and source before actually making adjustments. In any case, be1665very careful when writing to any of these files. The entries in /proc may1666change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt1667review the kernel documentation in the directory linux/Documentation.1668This chapter is heavily based on the documentation included in the pre 2.21669kernels, and became part of it in version 2.2.1 of the Linux kernel.1670 1671Please see: Documentation/admin-guide/sysctl/ directory for descriptions of1672these entries.1673 1674Summary1675-------1676 1677Certain aspects of kernel behavior can be modified at runtime, without the1678need to recompile the kernel, or even to reboot the system. The files in the1679/proc/sys tree can not only be read, but also modified. You can use the echo1680command to write value into these files, thereby changing the default settings1681of the kernel.1682 1683 1684Chapter 3: Per-process Parameters1685=================================1686 16873.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score1688--------------------------------------------------------------------------------1689 1690These files can be used to adjust the badness heuristic used to select which1691process gets killed in out of memory (oom) conditions.1692 1693The badness heuristic assigns a value to each candidate task ranging from 01694(never kill) to 1000 (always kill) to determine which process is targeted. The1695units are roughly a proportion along that range of allowed memory the process1696may allocate from based on an estimation of its current memory and swap use.1697For example, if a task is using all allowed memory, its badness score will be16981000. If it is using half of its allowed memory, its score will be 500.1699 1700The amount of "allowed" memory depends on the context in which the oom killer1701was called. If it is due to the memory assigned to the allocating task's cpuset1702being exhausted, the allowed memory represents the set of mems assigned to that1703cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed1704memory represents the set of mempolicy nodes. If it is due to a memory1705limit (or swap limit) being reached, the allowed memory is that configured1706limit. Finally, if it is due to the entire system being out of memory, the1707allowed memory represents all allocatable resources.1708 1709The value of /proc/<pid>/oom_score_adj is added to the badness score before it1710is used to determine which task to kill. Acceptable values range from -10001711(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to1712polarize the preference for oom killing either by always preferring a certain1713task or completely disabling it. The lowest possible value, -1000, is1714equivalent to disabling oom killing entirely for that task since it will always1715report a badness score of 0.1716 1717Consequently, it is very simple for userspace to define the amount of memory to1718consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for1719example, is roughly equivalent to allowing the remainder of tasks sharing the1720same system, cpuset, mempolicy, or memory controller resources to use at least172150% more memory. A value of -500, on the other hand, would be roughly1722equivalent to discounting 50% of the task's allowed memory from being considered1723as scoring against the task.1724 1725For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also1726be used to tune the badness score. Its acceptable values range from -161727(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -171728(OOM_DISABLE) to disable oom killing entirely for that task. Its value is1729scaled linearly with /proc/<pid>/oom_score_adj.1730 1731The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last1732value set by a CAP_SYS_RESOURCE process. To reduce the value any lower1733requires CAP_SYS_RESOURCE.1734 1735 17363.2 /proc/<pid>/oom_score - Display current oom-killer score1737-------------------------------------------------------------1738 1739This file can be used to check the current score used by the oom-killer for1740any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which1741process should be killed in an out-of-memory situation.1742 1743Please note that the exported value includes oom_score_adj so it is1744effectively in range [0,2000].1745 1746 17473.3 /proc/<pid>/io - Display the IO accounting fields1748-------------------------------------------------------1749 1750This file contains IO statistics for each running process.1751 1752Example1753~~~~~~~1754 1755::1756 1757 test:/tmp # dd if=/dev/zero of=/tmp/test.dat &1758 [1] 38281759 1760 test:/tmp # cat /proc/3828/io1761 rchar: 3239349311762 wchar: 3239296001763 syscr: 6326871764 syscw: 6326751765 read_bytes: 01766 write_bytes: 3239321601767 cancelled_write_bytes: 01768 1769 1770Description1771~~~~~~~~~~~1772 1773rchar1774^^^^^1775 1776I/O counter: chars read1777The number of bytes which this task has caused to be read from storage. This1778is simply the sum of bytes which this process passed to read() and pread().1779It includes things like tty IO and it is unaffected by whether or not actual1780physical disk IO was required (the read might have been satisfied from1781pagecache).1782 1783 1784wchar1785^^^^^1786 1787I/O counter: chars written1788The number of bytes which this task has caused, or shall cause to be written1789to disk. Similar caveats apply here as with rchar.1790 1791 1792syscr1793^^^^^1794 1795I/O counter: read syscalls1796Attempt to count the number of read I/O operations, i.e. syscalls like read()1797and pread().1798 1799 1800syscw1801^^^^^1802 1803I/O counter: write syscalls1804Attempt to count the number of write I/O operations, i.e. syscalls like1805write() and pwrite().1806 1807 1808read_bytes1809^^^^^^^^^^1810 1811I/O counter: bytes read1812Attempt to count the number of bytes which this process really did cause to1813be fetched from the storage layer. Done at the submit_bio() level, so it is1814accurate for block-backed filesystems. <please add status regarding NFS and1815CIFS at a later time>1816 1817 1818write_bytes1819^^^^^^^^^^^1820 1821I/O counter: bytes written1822Attempt to count the number of bytes which this process caused to be sent to1823the storage layer. This is done at page-dirtying time.1824 1825 1826cancelled_write_bytes1827^^^^^^^^^^^^^^^^^^^^^1828 1829The big inaccuracy here is truncate. If a process writes 1MB to a file and1830then deletes the file, it will in fact perform no writeout. But it will have1831been accounted as having caused 1MB of write.1832In other words: The number of bytes which this process caused to not happen,1833by truncating pagecache. A task can cause "negative" IO too. If this task1834truncates some dirty pagecache, some IO which another task has been accounted1835for (in its write_bytes) will not be happening. We _could_ just subtract that1836from the truncating task's write_bytes, but there is information loss in doing1837that.1838 1839 1840.. Note::1841 1842 At its current implementation state, this is a bit racy on 32-bit machines:1843 if process A reads process B's /proc/pid/io while process B is updating one1844 of those 64-bit counters, process A could see an intermediate result.1845 1846 1847More information about this can be found within the taskstats documentation in1848Documentation/accounting.1849 18503.4 /proc/<pid>/coredump_filter - Core dump filtering settings1851---------------------------------------------------------------1852When a process is dumped, all anonymous memory is written to a core file as1853long as the size of the core file isn't limited. But sometimes we don't want1854to dump some memory segments, for example, huge shared memory or DAX.1855Conversely, sometimes we want to save file-backed memory segments into a core1856file, not only the individual files.1857 1858/proc/<pid>/coredump_filter allows you to customize which memory segments1859will be dumped when the <pid> process is dumped. coredump_filter is a bitmask1860of memory types. If a bit of the bitmask is set, memory segments of the1861corresponding memory type are dumped, otherwise they are not dumped.1862 1863The following 9 memory types are supported:1864 1865 - (bit 0) anonymous private memory1866 - (bit 1) anonymous shared memory1867 - (bit 2) file-backed private memory1868 - (bit 3) file-backed shared memory1869 - (bit 4) ELF header pages in file-backed private memory areas (it is1870 effective only if the bit 2 is cleared)1871 - (bit 5) hugetlb private memory1872 - (bit 6) hugetlb shared memory1873 - (bit 7) DAX private memory1874 - (bit 8) DAX shared memory1875 1876 Note that MMIO pages such as frame buffer are never dumped and vDSO pages1877 are always dumped regardless of the bitmask status.1878 1879 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is1880 only affected by bit 5-6, and DAX is only affected by bits 7-8.1881 1882The default value of coredump_filter is 0x33; this means all anonymous memory1883segments, ELF header pages and hugetlb private memory are dumped.1884 1885If you don't want to dump all shared memory segments attached to pid 1234,1886write 0x31 to the process's proc file::1887 1888 $ echo 0x31 > /proc/1234/coredump_filter1889 1890When a new process is created, the process inherits the bitmask status from its1891parent. It is useful to set up coredump_filter before the program runs.1892For example::1893 1894 $ echo 0x7 > /proc/self/coredump_filter1895 $ ./some_program1896 18973.5 /proc/<pid>/mountinfo - Information about mounts1898--------------------------------------------------------1899 1900This file contains lines of the form::1901 1902 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue1903 (1)(2)(3) (4) (5) (6) (n…m) (m+1)(m+2) (m+3) (m+4)1904 1905 (1) mount ID: unique identifier of the mount (may be reused after umount)1906 (2) parent ID: ID of parent (or of self for the top of the mount tree)1907 (3) major:minor: value of st_dev for files on filesystem1908 (4) root: root of the mount within the filesystem1909 (5) mount point: mount point relative to the process's root1910 (6) mount options: per mount options1911 (n…m) optional fields: zero or more fields of the form "tag[:value]"1912 (m+1) separator: marks the end of the optional fields1913 (m+2) filesystem type: name of filesystem of the form "type[.subtype]"1914 (m+3) mount source: filesystem specific information or "none"1915 (m+4) super options: per super block options1916 1917Parsers should ignore all unrecognised optional fields. Currently the1918possible optional fields are:1919 1920================ ==============================================================1921shared:X mount is shared in peer group X1922master:X mount is slave to peer group X1923propagate_from:X mount is slave and receives propagation from peer group X [#]_1924unbindable mount is unbindable1925================ ==============================================================1926 1927.. [#] X is the closest dominant peer group under the process's root. If1928 X is the immediate master of the mount, or if there's no dominant peer1929 group under the same root, then only the "master:X" field is present1930 and not the "propagate_from:X" field.1931 1932For more information on mount propagation see:1933 1934 Documentation/filesystems/sharedsubtree.rst1935 1936 19373.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm1938--------------------------------------------------------1939These files provide a method to access a task's comm value. It also allows for1940a task to set its own or one of its thread siblings comm value. The comm value1941is limited in size compared to the cmdline value, so writing anything longer1942then the kernel's TASK_COMM_LEN (currently 16 chars, including the NUL1943terminator) will result in a truncated comm value.1944 1945 19463.7 /proc/<pid>/task/<tid>/children - Information about task children1947-------------------------------------------------------------------------1948This file provides a fast way to retrieve first level children pids1949of a task pointed by <pid>/<tid> pair. The format is a space separated1950stream of pids.1951 1952Note the "first level" here -- if a child has its own children they will1953not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children1954to obtain the descendants.1955 1956Since this interface is intended to be fast and cheap it doesn't1957guarantee to provide precise results and some children might be1958skipped, especially if they've exited right after we printed their1959pids, so one needs to either stop or freeze processes being inspected1960if precise results are needed.1961 1962 19633.8 /proc/<pid>/fdinfo/<fd> - Information about opened file1964---------------------------------------------------------------1965This file provides information associated with an opened file. The regular1966files have at least four fields -- 'pos', 'flags', 'mnt_id' and 'ino'.1967The 'pos' represents the current offset of the opened file in decimal1968form [see lseek(2) for details], 'flags' denotes the octal O_xxx mask the1969file has been created with [see open(2) for details] and 'mnt_id' represents1970mount ID of the file system containing the opened file [see 3.51971/proc/<pid>/mountinfo for details]. 'ino' represents the inode number of1972the file.1973 1974A typical output is::1975 1976 pos: 01977 flags: 01000021978 mnt_id: 191979 ino: 631071980 1981All locks associated with a file descriptor are shown in its fdinfo too::1982 1983 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF1984 1985The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags1986pair provide additional information particular to the objects they represent.1987 1988Eventfd files1989~~~~~~~~~~~~~1990 1991::1992 1993 pos: 01994 flags: 040021995 mnt_id: 91996 ino: 631071997 eventfd-count: 5a1998 1999where 'eventfd-count' is hex value of a counter.2000 2001Signalfd files2002~~~~~~~~~~~~~~2003 2004::2005 2006 pos: 02007 flags: 040022008 mnt_id: 92009 ino: 631072010 sigmask: 00000000000002002011 2012where 'sigmask' is hex value of the signal mask associated2013with a file.2014 2015Epoll files2016~~~~~~~~~~~2017 2018::2019 2020 pos: 02021 flags: 022022 mnt_id: 92023 ino: 631072024 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:72025 2026where 'tfd' is a target file descriptor number in decimal form,2027'events' is events mask being watched and the 'data' is data2028associated with a target [see epoll(7) for more details].2029 2030The 'pos' is current offset of the target file in decimal form2031[see lseek(2)], 'ino' and 'sdev' are inode and device numbers2032where target file resides, all in hex format.2033 2034Fsnotify files2035~~~~~~~~~~~~~~2036For inotify files the format is the following::2037 2038 pos: 02039 flags: 020000002040 mnt_id: 92041 ino: 631072042 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d2043 2044where 'wd' is a watch descriptor in decimal form, i.e. a target file2045descriptor number, 'ino' and 'sdev' are inode and device where the2046target file resides and the 'mask' is the mask of events, all in hex2047form [see inotify(7) for more details].2048 2049If the kernel was built with exportfs support, the path to the target2050file is encoded as a file handle. The file handle is provided by three2051fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex2052format.2053 2054If the kernel is built without exportfs support the file handle won't be2055printed out.2056 2057If there is no inotify mark attached yet the 'inotify' line will be omitted.2058 2059For fanotify files the format is::2060 2061 pos: 02062 flags: 022063 mnt_id: 92064 ino: 631072065 fanotify flags:10 event-flags:02066 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:400000032067 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b42068 2069where fanotify 'flags' and 'event-flags' are values used in fanotify_init2070call, 'mnt_id' is the mount point identifier, 'mflags' is the value of2071flags associated with mark which are tracked separately from events2072mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events2073mask and 'ignored_mask' is the mask of events which are to be ignored.2074All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask'2075provide information about flags and mask used in fanotify_mark2076call [see fsnotify manpage for details].2077 2078While the first three lines are mandatory and always printed, the rest is2079optional and may be omitted if no marks created yet.2080 2081Timerfd files2082~~~~~~~~~~~~~2083 2084::2085 2086 pos: 02087 flags: 022088 mnt_id: 92089 ino: 631072090 clockid: 02091 ticks: 02092 settime flags: 012093 it_value: (0, 49406829)2094 it_interval: (1, 0)2095 2096where 'clockid' is the clock type and 'ticks' is the number of the timer expirations2097that have occurred [see timerfd_create(2) for details]. 'settime flags' are2098flags in octal form been used to setup the timer [see timerfd_settime(2) for2099details]. 'it_value' is remaining time until the timer expiration.2100'it_interval' is the interval for the timer. Note the timer might be set up2101with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value'2102still exhibits timer's remaining time.2103 2104DMA Buffer files2105~~~~~~~~~~~~~~~~2106 2107::2108 2109 pos: 02110 flags: 040022111 mnt_id: 92112 ino: 631072113 size: 327682114 count: 22115 exp_name: system-heap2116 2117where 'size' is the size of the DMA buffer in bytes. 'count' is the file count of2118the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter.2119 21203.9 /proc/<pid>/map_files - Information about memory mapped files2121---------------------------------------------------------------------2122This directory contains symbolic links which represent memory mapped files2123the process is maintaining. Example output::2124 2125 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so2126 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so2127 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so2128 | ...2129 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.12130 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls2131 2132The name of a link represents the virtual memory bounds of a mapping, i.e.2133vm_area_struct::vm_start-vm_area_struct::vm_end.2134 2135The main purpose of the map_files is to retrieve a set of memory mapped2136files in a fast way instead of parsing /proc/<pid>/maps or2137/proc/<pid>/smaps, both of which contain many more records. At the same2138time one can open(2) mappings from the listings of two processes and2139comparing their inode numbers to figure out which anonymous memory areas2140are actually shared.2141 21423.10 /proc/<pid>/timerslack_ns - Task timerslack value2143---------------------------------------------------------2144This file provides the value of the task's timerslack value in nanoseconds.2145This value specifies an amount of time that normal timers may be deferred2146in order to coalesce timers and avoid unnecessary wakeups.2147 2148This allows a task's interactivity vs power consumption tradeoff to be2149adjusted.2150 2151Writing 0 to the file will set the task's timerslack to the default value.2152 2153Valid values are from 0 - ULLONG_MAX2154 2155An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level2156permissions on the task specified to change its timerslack_ns value.2157 21583.11 /proc/<pid>/patch_state - Livepatch patch operation state2159-----------------------------------------------------------------2160When CONFIG_LIVEPATCH is enabled, this file displays the value of the2161patch state for the task.2162 2163A value of '-1' indicates that no patch is in transition.2164 2165A value of '0' indicates that a patch is in transition and the task is2166unpatched. If the patch is being enabled, then the task hasn't been2167patched yet. If the patch is being disabled, then the task has already2168been unpatched.2169 2170A value of '1' indicates that a patch is in transition and the task is2171patched. If the patch is being enabled, then the task has already been2172patched. If the patch is being disabled, then the task hasn't been2173unpatched yet.2174 21753.12 /proc/<pid>/arch_status - task architecture specific status2176-------------------------------------------------------------------2177When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the2178architecture specific status of the task.2179 2180Example2181~~~~~~~2182 2183::2184 2185 $ cat /proc/6753/arch_status2186 AVX512_elapsed_ms: 82187 2188Description2189~~~~~~~~~~~2190 2191x86 specific entries2192~~~~~~~~~~~~~~~~~~~~~2193 2194AVX512_elapsed_ms2195^^^^^^^^^^^^^^^^^^2196 2197 If AVX512 is supported on the machine, this entry shows the milliseconds2198 elapsed since the last time AVX512 usage was recorded. The recording2199 happens on a best effort basis when a task is scheduled out. This means2200 that the value depends on two factors:2201 2202 1) The time which the task spent on the CPU without being scheduled2203 out. With CPU isolation and a single runnable task this can take2204 several seconds.2205 2206 2) The time since the task was scheduled out last. Depending on the2207 reason for being scheduled out (time slice exhausted, syscall ...)2208 this can be arbitrary long time.2209 2210 As a consequence the value cannot be considered precise and authoritative2211 information. The application which uses this information has to be aware2212 of the overall scenario on the system in order to determine whether a2213 task is a real AVX512 user or not. Precise information can be obtained2214 with performance counters.2215 2216 A special value of '-1' indicates that no AVX512 usage was recorded, thus2217 the task is unlikely an AVX512 user, but depends on the workload and the2218 scheduling scenario, it also could be a false negative mentioned above.2219 22203.13 /proc/<pid>/fd - List of symlinks to open files2221-------------------------------------------------------2222This directory contains symbolic links which represent open files2223the process is maintaining. Example output::2224 2225 lr-x------ 1 root root 64 Sep 20 17:53 0 -> /dev/null2226 l-wx------ 1 root root 64 Sep 20 17:53 1 -> /dev/null2227 lrwx------ 1 root root 64 Sep 20 17:53 10 -> 'socket:[12539]'2228 lrwx------ 1 root root 64 Sep 20 17:53 11 -> 'socket:[12540]'2229 lrwx------ 1 root root 64 Sep 20 17:53 12 -> 'socket:[12542]'2230 2231The number of open files for the process is stored in 'size' member2232of stat() output for /proc/<pid>/fd for fast access.2233-------------------------------------------------------2234 2235 2236Chapter 4: Configuring procfs2237=============================2238 22394.1 Mount options2240---------------------2241 2242The following mount options are supported:2243 2244 ========= ========================================================2245 hidepid= Set /proc/<pid>/ access mode.2246 gid= Set the group authorized to learn processes information.2247 subset= Show only the specified subset of procfs.2248 ========= ========================================================2249 2250hidepid=off or hidepid=0 means classic mode - everybody may access all2251/proc/<pid>/ directories (default).2252 2253hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/2254directories but their own. Sensitive files like cmdline, sched*, status are now2255protected against other users. This makes it impossible to learn whether any2256user runs specific program (given the program doesn't reveal itself by its2257behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for2258other users, poorly written programs passing sensitive information via program2259arguments are now protected against local eavesdroppers.2260 2261hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be2262fully invisible to other users. It doesn't mean that it hides a fact whether a2263process with a specific pid value exists (it can be learned by other means, e.g.2264by "kill -0 $PID"), but it hides process' uid and gid, which may be learned by2265stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of2266gathering information about running processes, whether some daemon runs with2267elevated privileges, whether other user runs some sensitive program, whether2268other users run any program at all, etc.2269 2270hidepid=ptraceable or hidepid=4 means that procfs should only contain2271/proc/<pid>/ directories that the caller can ptrace.2272 2273gid= defines a group authorized to learn processes information otherwise2274prohibited by hidepid=. If you use some daemon like identd which needs to learn2275information about processes information, just add identd to this group.2276 2277subset=pid hides all top level files and directories in the procfs that2278are not related to tasks.2279 2280Chapter 5: Filesystem behavior2281==============================2282 2283Originally, before the advent of pid namespace, procfs was a global file2284system. It means that there was only one procfs instance in the system.2285 2286When pid namespace was added, a separate procfs instance was mounted in2287each pid namespace. So, procfs mount options are global among all2288mountpoints within the same namespace::2289 2290 # grep ^proc /proc/mounts2291 proc /proc proc rw,relatime,hidepid=2 0 02292 2293 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc2294 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 02295 +++ exited with 0 +++2296 2297 # grep ^proc /proc/mounts2298 proc /proc proc rw,relatime,hidepid=2 0 02299 proc /tmp/proc proc rw,relatime,hidepid=2 0 02300 2301and only after remounting procfs mount options will change at all2302mountpoints::2303 2304 # mount -o remount,hidepid=1 -t proc proc /tmp/proc2305 2306 # grep ^proc /proc/mounts2307 proc /proc proc rw,relatime,hidepid=1 0 02308 proc /tmp/proc proc rw,relatime,hidepid=1 0 02309 2310This behavior is different from the behavior of other filesystems.2311 2312The new procfs behavior is more like other filesystems. Each procfs mount2313creates a new procfs instance. Mount options affect own procfs instance.2314It means that it became possible to have several procfs instances2315displaying tasks with different filtering options in one pid namespace::2316 2317 # mount -o hidepid=invisible -t proc proc /proc2318 # mount -o hidepid=noaccess -t proc proc /tmp/proc2319 # grep ^proc /proc/mounts2320 proc /proc proc rw,relatime,hidepid=invisible 0 02321 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 02322