1.. SPDX-License-Identifier: GPL-2.0 2 3==================== 4The /proc Filesystem 5==================== 6 7===================== ======================================= ================ 8/proc/sys Terrehon Bowden <terrehon@pacbell.net>, October 7 1999 9 Bodo Bauer <bb@ricochet.net> 102.4.x update Jorge Nerin <comandante@zaralinux.com> November 14 2000 11move /proc/sys Shen Feng <shen@cn.fujitsu.com> April 1 2009 12fixes/update part 1.1 Stefani Seibold <stefani@seibold.net> June 9 2009 13===================== ======================================= ================ 14 15 16 17.. Table of Contents 18 19 0 Preface 20 0.1 Introduction/Credits 21 0.2 Legal Stuff 22 23 1 Collecting System Information 24 1.1 Process-Specific Subdirectories 25 1.2 Kernel data 26 1.3 IDE devices in /proc/ide 27 1.4 Networking info in /proc/net 28 1.5 SCSI info 29 1.6 Parallel port info in /proc/parport 30 1.7 TTY info in /proc/tty 31 1.8 Miscellaneous kernel statistics in /proc/stat 32 1.9 Ext4 file system parameters 33 34 2 Modifying System Parameters 35 36 3 Per-Process Parameters 37 3.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj - Adjust the oom-killer 38 score 39 3.2 /proc/<pid>/oom_score - Display current oom-killer score 40 3.3 /proc/<pid>/io - Display the IO accounting fields 41 3.4 /proc/<pid>/coredump_filter - Core dump filtering settings 42 3.5 /proc/<pid>/mountinfo - Information about mounts 43 3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 44 3.7 /proc/<pid>/task/<tid>/children - Information about task children 45 3.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 46 3.9 /proc/<pid>/map_files - Information about memory mapped files 47 3.10 /proc/<pid>/timerslack_ns - Task timerslack value 48 3.11 /proc/<pid>/patch_state - Livepatch patch operation state 49 3.12 /proc/<pid>/arch_status - Task architecture specific information 50 51 4 Configuring procfs 52 4.1 Mount options 53 54 5 Filesystem behavior 55 56Preface 57======= 58 590.1 Introduction/Credits 60------------------------ 61 62This documentation is part of a soon (or so we hope) to be released book on 63the SuSE Linux distribution. As there is no complete documentation for the 64/proc file system and we've used many freely available sources to write these 65chapters, it seems only fair to give the work back to the Linux community. 66This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm 67afraid it's still far from complete, but we hope it will be useful. As far as 68we know, it is the first 'all-in-one' document about the /proc file system. It 69is focused on the Intel x86 hardware, so if you are looking for PPC, ARM, 70SPARC, AXP, etc., features, you probably won't find what you are looking for. 71It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But 72additions and patches are welcome and will be added to this document if you 73mail them to Bodo. 74 75We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of 76other people for help compiling this documentation. We'd also like to extend a 77special thank you to Andi Kleen for documentation, which we relied on heavily 78to create this document, as well as the additional information he provided. 79Thanks to everybody else who contributed source or docs to the Linux kernel 80and helped create a great piece of software... :) 81 82If you have any comments, corrections or additions, please don't hesitate to 83contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this 84document. 85 86The latest version of this document is available online at 87http://tldp.org/LDP/Linux-Filesystem-Hierarchy/html/proc.html 88 89If the above direction does not works for you, you could try the kernel 90mailing list at linux-kernel@vger.kernel.org and/or try to reach me at 91comandante@zaralinux.com. 92 930.2 Legal Stuff 94--------------- 95 96We don't guarantee the correctness of this document, and if you come to us 97complaining about how you screwed up your system because of incorrect 98documentation, we won't feel responsible... 99 100Chapter 1: Collecting System Information 101======================================== 102 103In This Chapter 104--------------- 105* Investigating the properties of the pseudo file system /proc and its 106 ability to provide information on the running Linux system 107* Examining /proc's structure 108* Uncovering various information about the kernel and the processes running 109 on the system 110 111------------------------------------------------------------------------------ 112 113The proc file system acts as an interface to internal data structures in the 114kernel. It can be used to obtain information about the system and to change 115certain kernel parameters at runtime (sysctl). 116 117First, we'll take a look at the read-only parts of /proc. In Chapter 2, we 118show you how you can use /proc/sys to change settings. 119 1201.1 Process-Specific Subdirectories 121----------------------------------- 122 123The directory /proc contains (among other things) one subdirectory for each 124process running on the system, which is named after the process ID (PID). 125 126The link 'self' points to the process reading the file system. Each process 127subdirectory has the entries listed in Table 1-1. 128 129Note that an open file descriptor to /proc/<pid> or to any of its 130contained files or subdirectories does not prevent <pid> being reused 131for some other process in the event that <pid> exits. Operations on 132open /proc/<pid> file descriptors corresponding to dead processes 133never act on any new process that the kernel may, through chance, have 134also assigned the process ID <pid>. Instead, operations on these FDs 135usually fail with ESRCH. 136 137.. table:: Table 1-1: Process specific entries in /proc 138 139 ============= =============================================================== 140 File Content 141 ============= =============================================================== 142 clear_refs Clears page referenced bits shown in smaps output 143 cmdline Command line arguments 144 cpu Current and last cpu in which it was executed (2.4)(smp) 145 cwd Link to the current working directory 146 environ Values of environment variables 147 exe Link to the executable of this process 148 fd Directory, which contains all file descriptors 149 maps Memory maps to executables and library files (2.4) 150 mem Memory held by this process 151 root Link to the root directory of this process 152 stat Process status 153 statm Process memory status information 154 status Process status in human readable form 155 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function 156 symbol the task is blocked in - or "0" if not blocked. 157 pagemap Page table 158 stack Report full stack trace, enable via CONFIG_STACKTRACE 159 smaps An extension based on maps, showing the memory consumption of 160 each mapping and flags associated with it 161 smaps_rollup Accumulated smaps stats for all mappings of the process. This 162 can be derived from smaps, but is faster and more convenient 163 numa_maps An extension based on maps, showing the memory locality and 164 binding policy as well as mem usage (in pages) of each mapping. 165 ============= =============================================================== 166 167For example, to get the status information of a process, all you have to do is 168read the file /proc/PID/status:: 169 170 >cat /proc/self/status 171 Name: cat 172 State: R (running) 173 Tgid: 5452 174 Pid: 5452 175 PPid: 743 176 TracerPid: 0 (2.4) 177 Uid: 501 501 501 501 178 Gid: 100 100 100 100 179 FDSize: 256 180 Groups: 100 14 16 181 VmPeak: 5004 kB 182 VmSize: 5004 kB 183 VmLck: 0 kB 184 VmHWM: 476 kB 185 VmRSS: 476 kB 186 RssAnon: 352 kB 187 RssFile: 120 kB 188 RssShmem: 4 kB 189 VmData: 156 kB 190 VmStk: 88 kB 191 VmExe: 68 kB 192 VmLib: 1412 kB 193 VmPTE: 20 kb 194 VmSwap: 0 kB 195 HugetlbPages: 0 kB 196 CoreDumping: 0 197 THP_enabled: 1 198 Threads: 1 199 SigQ: 0/28578 200 SigPnd: 0000000000000000 201 ShdPnd: 0000000000000000 202 SigBlk: 0000000000000000 203 SigIgn: 0000000000000000 204 SigCgt: 0000000000000000 205 CapInh: 00000000fffffeff 206 CapPrm: 0000000000000000 207 CapEff: 0000000000000000 208 CapBnd: ffffffffffffffff 209 CapAmb: 0000000000000000 210 NoNewPrivs: 0 211 Seccomp: 0 212 Speculation_Store_Bypass: thread vulnerable 213 SpeculationIndirectBranch: conditional enabled 214 voluntary_ctxt_switches: 0 215 nonvoluntary_ctxt_switches: 1 216 217This shows you nearly the same information you would get if you viewed it with 218the ps command. In fact, ps uses the proc file system to obtain its 219information. But you get a more detailed view of the process by reading the 220file /proc/PID/status. It fields are described in table 1-2. 221 222The statm file contains more detailed information about the process 223memory usage. Its seven fields are explained in Table 1-3. The stat file 224contains detailed information about the process itself. Its fields are 225explained in Table 1-4. 226 227(for SMP CONFIG users) 228 229For making accounting scalable, RSS related information are handled in an 230asynchronous manner and the value may not be very precise. To see a precise 231snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table. 232It's slow but very precise. 233 234.. table:: Table 1-2: Contents of the status files (as of 4.19) 235 236 ========================== =================================================== 237 Field Content 238 ========================== =================================================== 239 Name filename of the executable 240 Umask file mode creation mask 241 State state (R is running, S is sleeping, D is sleeping 242 in an uninterruptible wait, Z is zombie, 243 T is traced or stopped) 244 Tgid thread group ID 245 Ngid NUMA group ID (0 if none) 246 Pid process id 247 PPid process id of the parent process 248 TracerPid PID of process tracing this process (0 if not) 249 Uid Real, effective, saved set, and file system UIDs 250 Gid Real, effective, saved set, and file system GIDs 251 FDSize number of file descriptor slots currently allocated 252 Groups supplementary group list 253 NStgid descendant namespace thread group ID hierarchy 254 NSpid descendant namespace process ID hierarchy 255 NSpgid descendant namespace process group ID hierarchy 256 NSsid descendant namespace session ID hierarchy 257 VmPeak peak virtual memory size 258 VmSize total program size 259 VmLck locked memory size 260 VmPin pinned memory size 261 VmHWM peak resident set size ("high water mark") 262 VmRSS size of memory portions. It contains the three 263 following parts 264 (VmRSS = RssAnon + RssFile + RssShmem) 265 RssAnon size of resident anonymous memory 266 RssFile size of resident file mappings 267 RssShmem size of resident shmem memory (includes SysV shm, 268 mapping of tmpfs and shared anonymous mappings) 269 VmData size of private data segments 270 VmStk size of stack segments 271 VmExe size of text segment 272 VmLib size of shared library code 273 VmPTE size of page table entries 274 VmSwap amount of swap used by anonymous private data 275 (shmem swap usage is not included) 276 HugetlbPages size of hugetlb memory portions 277 CoreDumping process's memory is currently being dumped 278 (killing the process may lead to a corrupted core) 279 THP_enabled process is allowed to use THP (returns 0 when 280 PR_SET_THP_DISABLE is set on the process 281 Threads number of threads 282 SigQ number of signals queued/max. number for queue 283 SigPnd bitmap of pending signals for the thread 284 ShdPnd bitmap of shared pending signals for the process 285 SigBlk bitmap of blocked signals 286 SigIgn bitmap of ignored signals 287 SigCgt bitmap of caught signals 288 CapInh bitmap of inheritable capabilities 289 CapPrm bitmap of permitted capabilities 290 CapEff bitmap of effective capabilities 291 CapBnd bitmap of capabilities bounding set 292 CapAmb bitmap of ambient capabilities 293 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...) 294 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...) 295 Speculation_Store_Bypass speculative store bypass mitigation status 296 SpeculationIndirectBranch indirect branch speculation mode 297 Cpus_allowed mask of CPUs on which this process may run 298 Cpus_allowed_list Same as previous, but in "list format" 299 Mems_allowed mask of memory nodes allowed to this process 300 Mems_allowed_list Same as previous, but in "list format" 301 voluntary_ctxt_switches number of voluntary context switches 302 nonvoluntary_ctxt_switches number of non voluntary context switches 303 ========================== =================================================== 304 305 306.. table:: Table 1-3: Contents of the statm files (as of 2.6.8-rc3) 307 308 ======== =============================== ============================== 309 Field Content 310 ======== =============================== ============================== 311 size total program size (pages) (same as VmSize in status) 312 resident size of memory portions (pages) (same as VmRSS in status) 313 shared number of pages that are shared (i.e. backed by a file, same 314 as RssFile+RssShmem in status) 315 trs number of pages that are 'code' (not including libs; broken, 316 includes data segment) 317 lrs number of pages of library (always 0 on 2.6) 318 drs number of pages of data/stack (including libs; broken, 319 includes library text) 320 dt number of dirty pages (always 0 on 2.6) 321 ======== =============================== ============================== 322 323 324.. table:: Table 1-4: Contents of the stat files (as of 2.6.30-rc7) 325 326 ============= =============================================================== 327 Field Content 328 ============= =============================================================== 329 pid process id 330 tcomm filename of the executable 331 state state (R is running, S is sleeping, D is sleeping in an 332 uninterruptible wait, Z is zombie, T is traced or stopped) 333 ppid process id of the parent process 334 pgrp pgrp of the process 335 sid session id 336 tty_nr tty the process uses 337 tty_pgrp pgrp of the tty 338 flags task flags 339 min_flt number of minor faults 340 cmin_flt number of minor faults with child's 341 maj_flt number of major faults 342 cmaj_flt number of major faults with child's 343 utime user mode jiffies 344 stime kernel mode jiffies 345 cutime user mode jiffies with child's 346 cstime kernel mode jiffies with child's 347 priority priority level 348 nice nice level 349 num_threads number of threads 350 it_real_value (obsolete, always 0) 351 start_time time the process started after system boot 352 vsize virtual memory size 353 rss resident set memory size 354 rsslim current limit in bytes on the rss 355 start_code address above which program text can run 356 end_code address below which program text can run 357 start_stack address of the start of the main process stack 358 esp current value of ESP 359 eip current value of EIP 360 pending bitmap of pending signals 361 blocked bitmap of blocked signals 362 sigign bitmap of ignored signals 363 sigcatch bitmap of caught signals 364 0 (place holder, used to be the wchan address, 365 use /proc/PID/wchan instead) 366 0 (place holder) 367 0 (place holder) 368 exit_signal signal to send to parent thread on exit 369 task_cpu which CPU the task is scheduled on 370 rt_priority realtime priority 371 policy scheduling policy (man sched_setscheduler) 372 blkio_ticks time spent waiting for block IO 373 gtime guest time of the task in jiffies 374 cgtime guest time of the task children in jiffies 375 start_data address above which program data+bss is placed 376 end_data address below which program data+bss is placed 377 start_brk address above which program heap can be expanded with brk() 378 arg_start address above which program command line is placed 379 arg_end address below which program command line is placed 380 env_start address above which program environment is placed 381 env_end address below which program environment is placed 382 exit_code the thread's exit_code in the form reported by the waitpid 383 system call 384 ============= =============================================================== 385 386The /proc/PID/maps file contains the currently mapped memory regions and 387their access permissions. 388 389The format is:: 390 391 address perms offset dev inode pathname 392 393 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test 394 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test 395 0804a000-0806b000 rw-p 00000000 00:00 0 [heap] 396 a7cb1000-a7cb2000 ---p 00000000 00:00 0 397 a7cb2000-a7eb2000 rw-p 00000000 00:00 0 398 a7eb2000-a7eb3000 ---p 00000000 00:00 0 399 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 400 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6 401 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6 402 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6 403 a800b000-a800e000 rw-p 00000000 00:00 0 404 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0 405 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0 406 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0 407 a8024000-a8027000 rw-p 00000000 00:00 0 408 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2 409 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2 410 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2 411 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack] 412 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso] 413 414where "address" is the address space in the process that it occupies, "perms" 415is a set of permissions:: 416 417 r = read 418 w = write 419 x = execute 420 s = shared 421 p = private (copy on write) 422 423"offset" is the offset into the mapping, "dev" is the device (major:minor), and 424"inode" is the inode on that device. 0 indicates that no inode is associated 425with the memory region, as the case would be with BSS (uninitialized data). 426The "pathname" shows the name associated file for this mapping. If the mapping 427is not associated with a file: 428 429 ============= ==================================== 430 [heap] the heap of the program 431 [stack] the stack of the main process 432 [vdso] the "virtual dynamic shared object", 433 the kernel system call handler 434 [anon:<name>] an anonymous mapping that has been 435 named by userspace 436 ============= ==================================== 437 438 or if empty, the mapping is anonymous. 439 440The /proc/PID/smaps is an extension based on maps, showing the memory 441consumption for each of the process's mappings. For each mapping (aka Virtual 442Memory Area, or VMA) there is a series of lines such as the following:: 443 444 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash 445 446 Size: 1084 kB 447 KernelPageSize: 4 kB 448 MMUPageSize: 4 kB 449 Rss: 892 kB 450 Pss: 374 kB 451 Pss_Dirty: 0 kB 452 Shared_Clean: 892 kB 453 Shared_Dirty: 0 kB 454 Private_Clean: 0 kB 455 Private_Dirty: 0 kB 456 Referenced: 892 kB 457 Anonymous: 0 kB 458 LazyFree: 0 kB 459 AnonHugePages: 0 kB 460 ShmemPmdMapped: 0 kB 461 Shared_Hugetlb: 0 kB 462 Private_Hugetlb: 0 kB 463 Swap: 0 kB 464 SwapPss: 0 kB 465 KernelPageSize: 4 kB 466 MMUPageSize: 4 kB 467 Locked: 0 kB 468 THPeligible: 0 469 VmFlags: rd ex mr mw me dw 470 471The first of these lines shows the same information as is displayed for the 472mapping in /proc/PID/maps. Following lines show the size of the mapping 473(size); the size of each page allocated when backing a VMA (KernelPageSize), 474which is usually the same as the size in the page table entries; the page size 475used by the MMU when backing a VMA (in most cases, the same as KernelPageSize); 476the amount of the mapping that is currently resident in RAM (RSS); the 477process' proportional share of this mapping (PSS); and the number of clean and 478dirty shared and private pages in the mapping. 479 480The "proportional set size" (PSS) of a process is the count of pages it has 481in memory, where each page is divided by the number of processes sharing it. 482So if a process has 1000 pages all to itself, and 1000 shared with one other 483process, its PSS will be 1500. "Pss_Dirty" is the portion of PSS which 484consists of dirty pages. ("Pss_Clean" is not included, but it can be 485calculated by subtracting "Pss_Dirty" from "Pss".) 486 487Note that even a page which is part of a MAP_SHARED mapping, but has only 488a single pte mapped, i.e. is currently used by only one process, is accounted 489as private and not as shared. 490 491"Referenced" indicates the amount of memory currently marked as referenced or 492accessed. 493 494"Anonymous" shows the amount of memory that does not belong to any file. Even 495a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE 496and a page is modified, the file page is replaced by a private anonymous copy. 497 498"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE). 499The memory isn't freed immediately with madvise(). It's freed in memory 500pressure if the memory is clean. Please note that the printed value might 501be lower than the real value due to optimizations used in the current 502implementation. If this is not desirable please file a bug report. 503 504"AnonHugePages" shows the ammount of memory backed by transparent hugepage. 505 506"ShmemPmdMapped" shows the ammount of shared (shmem/tmpfs) memory backed by 507huge pages. 508 509"Shared_Hugetlb" and "Private_Hugetlb" show the ammounts of memory backed by 510hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical 511reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field. 512 513"Swap" shows how much would-be-anonymous memory is also used, but out on swap. 514 515For shmem mappings, "Swap" includes also the size of the mapped (and not 516replaced by copy-on-write) part of the underlying shmem object out on swap. 517"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this 518does not take into account swapped out page of underlying shmem objects. 519"Locked" indicates whether the mapping is locked in memory or not. 520 521"THPeligible" indicates whether the mapping is eligible for allocating THP 522pages as well as the THP is PMD mappable or not - 1 if true, 0 otherwise. 523It just shows the current status. 524 525"VmFlags" field deserves a separate description. This member represents the 526kernel flags associated with the particular virtual memory area in two letter 527encoded manner. The codes are the following: 528 529 == ======================================= 530 rd readable 531 wr writeable 532 ex executable 533 sh shared 534 mr may read 535 mw may write 536 me may execute 537 ms may share 538 gd stack segment growns down 539 pf pure PFN range 540 dw disabled write to the mapped file 541 lo pages are locked in memory 542 io memory mapped I/O area 543 sr sequential read advise provided 544 rr random read advise provided 545 dc do not copy area on fork 546 de do not expand area on remapping 547 ac area is accountable 548 nr swap space is not reserved for the area 549 ht area uses huge tlb pages 550 sf synchronous page fault 551 ar architecture specific flag 552 wf wipe on fork 553 dd do not include area into core dump 554 sd soft dirty flag 555 mm mixed map area 556 hg huge page advise flag 557 nh no huge page advise flag 558 mg mergable advise flag 559 bt arm64 BTI guarded page 560 mt arm64 MTE allocation tags are enabled 561 um userfaultfd missing tracking 562 uw userfaultfd wr-protect tracking 563 == ======================================= 564 565Note that there is no guarantee that every flag and associated mnemonic will 566be present in all further kernel releases. Things get changed, the flags may 567be vanished or the reverse -- new added. Interpretation of their meaning 568might change in future as well. So each consumer of these flags has to 569follow each specific kernel version for the exact semantic. 570 571This file is only present if the CONFIG_MMU kernel configuration option is 572enabled. 573 574Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent 575output can be achieved only in the single read call). 576 577This typically manifests when doing partial reads of these files while the 578memory map is being modified. Despite the races, we do provide the following 579guarantees: 580 5811) The mapped addresses never go backwards, which implies no two 582 regions will ever overlap. 5832) If there is something at a given vaddr during the entirety of the 584 life of the smaps/maps walk, there will be some output for it. 585 586The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps, 587but their values are the sums of the corresponding values for all mappings of 588the process. Additionally, it contains these fields: 589 590- Pss_Anon 591- Pss_File 592- Pss_Shmem 593 594They represent the proportional shares of anonymous, file, and shmem pages, as 595described for smaps above. These fields are omitted in smaps since each 596mapping identifies the type (anon, file, or shmem) of all pages it contains. 597Thus all information in smaps_rollup can be derived from smaps, but at a 598significantly higher cost. 599 600The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG 601bits on both physical and virtual pages associated with a process, and the 602soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst 603for details). 604To clear the bits for all the pages associated with the process:: 605 606 > echo 1 > /proc/PID/clear_refs 607 608To clear the bits for the anonymous pages associated with the process:: 609 610 > echo 2 > /proc/PID/clear_refs 611 612To clear the bits for the file mapped pages associated with the process:: 613 614 > echo 3 > /proc/PID/clear_refs 615 616To clear the soft-dirty bit:: 617 618 > echo 4 > /proc/PID/clear_refs 619 620To reset the peak resident set size ("high water mark") to the process's 621current value:: 622 623 > echo 5 > /proc/PID/clear_refs 624 625Any other value written to /proc/PID/clear_refs will have no effect. 626 627The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags 628using /proc/kpageflags and number of times a page is mapped using 629/proc/kpagecount. For detailed explanation, see 630Documentation/admin-guide/mm/pagemap.rst. 631 632The /proc/pid/numa_maps is an extension based on maps, showing the memory 633locality and binding policy, as well as the memory usage (in pages) of 634each mapping. The output follows a general format where mapping details get 635summarized separated by blank spaces, one mapping per each file line:: 636 637 address policy mapping details 638 639 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4 640 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4 641 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4 642 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 643 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 644 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4 645 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4 646 320698b000 default file=/lib64/libc-2.12.so 647 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4 648 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 649 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4 650 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4 651 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4 652 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048 653 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4 654 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4 655 656Where: 657 658"address" is the starting address for the mapping; 659 660"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst); 661 662"mapping details" summarizes mapping data such as mapping type, page usage counters, 663node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page 664size, in KB, that is backing the mapping up. 665 6661.2 Kernel data 667--------------- 668 669Similar to the process entries, the kernel data files give information about 670the running kernel. The files used to obtain this information are contained in 671/proc and are listed in Table 1-5. Not all of these will be present in your 672system. It depends on the kernel configuration and the loaded modules, which 673files are there, and which are missing. 674 675.. table:: Table 1-5: Kernel info in /proc 676 677 ============ =============================================================== 678 File Content 679 ============ =============================================================== 680 apm Advanced power management info 681 buddyinfo Kernel memory allocator information (see text) (2.5) 682 bus Directory containing bus specific information 683 cmdline Kernel command line 684 cpuinfo Info about the CPU 685 devices Available devices (block and character) 686 dma Used DMS channels 687 filesystems Supported filesystems 688 driver Various drivers grouped here, currently rtc (2.4) 689 execdomains Execdomains, related to security (2.4) 690 fb Frame Buffer devices (2.4) 691 fs File system parameters, currently nfs/exports (2.4) 692 ide Directory containing info about the IDE subsystem 693 interrupts Interrupt usage 694 iomem Memory map (2.4) 695 ioports I/O port usage 696 irq Masks for irq to cpu affinity (2.4)(smp?) 697 isapnp ISA PnP (Plug&Play) Info (2.4) 698 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4)) 699 kmsg Kernel messages 700 ksyms Kernel symbol table 701 loadavg Load average of last 1, 5 & 15 minutes; 702 number of processes currently runnable (running or on ready queue); 703 total number of processes in system; 704 last pid created. 705 All fields are separated by one space except "number of 706 processes currently runnable" and "total number of processes 707 in system", which are separated by a slash ('/'). Example: 708 0.61 0.61 0.55 3/828 22084 709 locks Kernel locks 710 meminfo Memory info 711 misc Miscellaneous 712 modules List of loaded modules 713 mounts Mounted filesystems 714 net Networking info (see text) 715 pagetypeinfo Additional page allocator information (see text) (2.5) 716 partitions Table of partitions known to the system 717 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/, 718 decoupled by lspci (2.4) 719 rtc Real time clock 720 scsi SCSI info (see text) 721 slabinfo Slab pool info 722 softirqs softirq usage 723 stat Overall statistics 724 swaps Swap space utilization 725 sys See chapter 2 726 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4) 727 tty Info of tty drivers 728 uptime Wall clock since boot, combined idle time of all cpus 729 version Kernel version 730 video bttv info of video resources (2.4) 731 vmallocinfo Show vmalloced areas 732 ============ =============================================================== 733 734You can, for example, check which interrupts are currently in use and what 735they are used for by looking in the file /proc/interrupts:: 736 737 > cat /proc/interrupts 738 CPU0 739 0: 8728810 XT-PIC timer 740 1: 895 XT-PIC keyboard 741 2: 0 XT-PIC cascade 742 3: 531695 XT-PIC aha152x 743 4: 2014133 XT-PIC serial 744 5: 44401 XT-PIC pcnet_cs 745 8: 2 XT-PIC rtc 746 11: 8 XT-PIC i82365 747 12: 182918 XT-PIC PS/2 Mouse 748 13: 1 XT-PIC fpu 749 14: 1232265 XT-PIC ide0 750 15: 7 XT-PIC ide1 751 NMI: 0 752 753In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the 754output of a SMP machine):: 755 756 > cat /proc/interrupts 757 758 CPU0 CPU1 759 0: 1243498 1214548 IO-APIC-edge timer 760 1: 8949 8958 IO-APIC-edge keyboard 761 2: 0 0 XT-PIC cascade 762 5: 11286 10161 IO-APIC-edge soundblaster 763 8: 1 0 IO-APIC-edge rtc 764 9: 27422 27407 IO-APIC-edge 3c503 765 12: 113645 113873 IO-APIC-edge PS/2 Mouse 766 13: 0 0 XT-PIC fpu 767 14: 22491 24012 IO-APIC-edge ide0 768 15: 2183 2415 IO-APIC-edge ide1 769 17: 30564 30414 IO-APIC-level eth0 770 18: 177 164 IO-APIC-level bttv 771 NMI: 2457961 2457959 772 LOC: 2457882 2457881 773 ERR: 2155 774 775NMI is incremented in this case because every timer interrupt generates a NMI 776(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups. 777 778LOC is the local interrupt counter of the internal APIC of every CPU. 779 780ERR is incremented in the case of errors in the IO-APIC bus (the bus that 781connects the CPUs in a SMP system. This means that an error has been detected, 782the IO-APIC automatically retry the transmission, so it should not be a big 783problem, but you should read the SMP-FAQ. 784 785In 2.6.2* /proc/interrupts was expanded again. This time the goal was for 786/proc/interrupts to display every IRQ vector in use by the system, not 787just those considered 'most important'. The new vectors are: 788 789THR 790 interrupt raised when a machine check threshold counter 791 (typically counting ECC corrected errors of memory or cache) exceeds 792 a configurable threshold. Only available on some systems. 793 794TRM 795 a thermal event interrupt occurs when a temperature threshold 796 has been exceeded for the CPU. This interrupt may also be generated 797 when the temperature drops back to normal. 798 799SPU 800 a spurious interrupt is some interrupt that was raised then lowered 801 by some IO device before it could be fully processed by the APIC. Hence 802 the APIC sees the interrupt but does not know what device it came from. 803 For this case the APIC will generate the interrupt with a IRQ vector 804 of 0xff. This might also be generated by chipset bugs. 805 806RES, CAL, TLB 807 rescheduling, call and TLB flush interrupts are 808 sent from one CPU to another per the needs of the OS. Typically, 809 their statistics are used by kernel developers and interested users to 810 determine the occurrence of interrupts of the given type. 811 812The above IRQ vectors are displayed only when relevant. For example, 813the threshold vector does not exist on x86_64 platforms. Others are 814suppressed when the system is a uniprocessor. As of this writing, only 815i386 and x86_64 platforms support the new IRQ vector displays. 816 817Of some interest is the introduction of the /proc/irq directory to 2.4. 818It could be used to set IRQ to CPU affinity. This means that you can "hook" an 819IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the 820irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and 821prof_cpu_mask. 822 823For example:: 824 825 > ls /proc/irq/ 826 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask 827 1 11 13 15 17 19 3 5 7 9 default_smp_affinity 828 > ls /proc/irq/0/ 829 smp_affinity 830 831smp_affinity is a bitmask, in which you can specify which CPUs can handle the 832IRQ. You can set it by doing:: 833 834 > echo 1 > /proc/irq/10/smp_affinity 835 836This means that only the first CPU will handle the IRQ, but you can also echo 8375 which means that only the first and third CPU can handle the IRQ. 838 839The contents of each smp_affinity file is the same by default:: 840 841 > cat /proc/irq/0/smp_affinity 842 ffffffff 843 844There is an alternate interface, smp_affinity_list which allows specifying 845a CPU range instead of a bitmask:: 846 847 > cat /proc/irq/0/smp_affinity_list 848 1024-1031 849 850The default_smp_affinity mask applies to all non-active IRQs, which are the 851IRQs which have not yet been allocated/activated, and hence which lack a 852/proc/irq/[0-9]* directory. 853 854The node file on an SMP system shows the node to which the device using the IRQ 855reports itself as being attached. This hardware locality information does not 856include information about any possible driver locality preference. 857 858prof_cpu_mask specifies which CPUs are to be profiled by the system wide 859profiler. Default value is ffffffff (all CPUs if there are only 32 of them). 860 861The way IRQs are routed is handled by the IO-APIC, and it's Round Robin 862between all the CPUs which are allowed to handle it. As usual the kernel has 863more info than you and does a better job than you, so the defaults are the 864best choice for almost everyone. [Note this applies only to those IO-APIC's 865that support "Round Robin" interrupt distribution.] 866 867There are three more important subdirectories in /proc: net, scsi, and sys. 868The general rule is that the contents, or even the existence of these 869directories, depend on your kernel configuration. If SCSI is not enabled, the 870directory scsi may not exist. The same is true with the net, which is there 871only when networking support is present in the running kernel. 872 873The slabinfo file gives information about memory usage at the slab level. 874Linux uses slab pools for memory management above page level in version 2.2. 875Commonly used objects have their own slab pool (such as network buffers, 876directory cache, and so on). 877 878:: 879 880 > cat /proc/buddyinfo 881 882 Node 0, zone DMA 0 4 5 4 4 3 ... 883 Node 0, zone Normal 1 0 0 1 101 8 ... 884 Node 0, zone HighMem 2 0 0 1 1 0 ... 885 886External fragmentation is a problem under some workloads, and buddyinfo is a 887useful tool for helping diagnose these problems. Buddyinfo will give you a 888clue as to how big an area you can safely allocate, or why a previous 889allocation failed. 890 891Each column represents the number of pages of a certain order which are 892available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 893ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 894available in ZONE_NORMAL, etc... 895 896More information relevant to external fragmentation can be found in 897pagetypeinfo:: 898 899 > cat /proc/pagetypeinfo 900 Page block order: 9 901 Pages per block: 512 902 903 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 904 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0 905 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0 906 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2 907 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0 908 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0 909 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9 910 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0 911 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452 912 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0 913 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0 914 915 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate 916 Node 0, zone DMA 2 0 5 1 0 917 Node 0, zone DMA32 41 6 967 2 0 918 919Fragmentation avoidance in the kernel works by grouping pages of different 920migrate types into the same contiguous regions of memory called page blocks. 921A page block is typically the size of the default hugepage size, e.g. 2MB on 922X86-64. By keeping pages grouped based on their ability to move, the kernel 923can reclaim pages within a page block to satisfy a high-order allocation. 924 925The pagetypinfo begins with information on the size of a page block. It 926then gives the same type of information as buddyinfo except broken down 927by migrate-type and finishes with details on how many page blocks of each 928type exist. 929 930If min_free_kbytes has been tuned correctly (recommendations made by hugeadm 931from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can 932make an estimate of the likely number of huge pages that can be allocated 933at a given point in time. All the "Movable" blocks should be allocatable 934unless memory has been mlock()'d. Some of the Reclaimable blocks should 935also be allocatable although a lot of filesystem metadata may have to be 936reclaimed to achieve this. 937 938 939meminfo 940~~~~~~~ 941 942Provides information about distribution and utilization of memory. This 943varies by architecture and compile options. Some of the counters reported 944here overlap. The memory reported by the non overlapping counters may not 945add up to the overall memory usage and the difference for some workloads 946can be substantial. In many cases there are other means to find out 947additional memory using subsystem specific interfaces, for instance 948/proc/net/sockstat for TCP memory allocations. 949 950Example output. You may not have all of these fields. 951 952:: 953 954 > cat /proc/meminfo 955 956 MemTotal: 32858820 kB 957 MemFree: 21001236 kB 958 MemAvailable: 27214312 kB 959 Buffers: 581092 kB 960 Cached: 5587612 kB 961 SwapCached: 0 kB 962 Active: 3237152 kB 963 Inactive: 7586256 kB 964 Active(anon): 94064 kB 965 Inactive(anon): 4570616 kB 966 Active(file): 3143088 kB 967 Inactive(file): 3015640 kB 968 Unevictable: 0 kB 969 Mlocked: 0 kB 970 SwapTotal: 0 kB 971 SwapFree: 0 kB 972 Zswap: 1904 kB 973 Zswapped: 7792 kB 974 Dirty: 12 kB 975 Writeback: 0 kB 976 AnonPages: 4654780 kB 977 Mapped: 266244 kB 978 Shmem: 9976 kB 979 KReclaimable: 517708 kB 980 Slab: 660044 kB 981 SReclaimable: 517708 kB 982 SUnreclaim: 142336 kB 983 KernelStack: 11168 kB 984 PageTables: 20540 kB 985 NFS_Unstable: 0 kB 986 Bounce: 0 kB 987 WritebackTmp: 0 kB 988 CommitLimit: 16429408 kB 989 Committed_AS: 7715148 kB 990 VmallocTotal: 34359738367 kB 991 VmallocUsed: 40444 kB 992 VmallocChunk: 0 kB 993 Percpu: 29312 kB 994 HardwareCorrupted: 0 kB 995 AnonHugePages: 4149248 kB 996 ShmemHugePages: 0 kB 997 ShmemPmdMapped: 0 kB 998 FileHugePages: 0 kB 999 FilePmdMapped: 0 kB 1000 CmaTotal: 0 kB 1001 CmaFree: 0 kB 1002 HugePages_Total: 0 1003 HugePages_Free: 0 1004 HugePages_Rsvd: 0 1005 HugePages_Surp: 0 1006 Hugepagesize: 2048 kB 1007 Hugetlb: 0 kB 1008 DirectMap4k: 401152 kB 1009 DirectMap2M: 10008576 kB 1010 DirectMap1G: 24117248 kB 1011 1012MemTotal 1013 Total usable RAM (i.e. physical RAM minus a few reserved 1014 bits and the kernel binary code) 1015MemFree 1016 Total free RAM. On highmem systems, the sum of LowFree+HighFree 1017MemAvailable 1018 An estimate of how much memory is available for starting new 1019 applications, without swapping. Calculated from MemFree, 1020 SReclaimable, the size of the file LRU lists, and the low 1021 watermarks in each zone. 1022 The estimate takes into account that the system needs some 1023 page cache to function well, and that not all reclaimable 1024 slab will be reclaimable, due to items being in use. The 1025 impact of those factors will vary from system to system. 1026Buffers 1027 Relatively temporary storage for raw disk blocks 1028 shouldn't get tremendously large (20MB or so) 1029Cached 1030 In-memory cache for files read from the disk (the 1031 pagecache) as well as tmpfs & shmem. 1032 Doesn't include SwapCached. 1033SwapCached 1034 Memory that once was swapped out, is swapped back in but 1035 still also is in the swapfile (if memory is needed it 1036 doesn't need to be swapped out AGAIN because it is already 1037 in the swapfile. This saves I/O) 1038Active 1039 Memory that has been used more recently and usually not 1040 reclaimed unless absolutely necessary. 1041Inactive 1042 Memory which has been less recently used. It is more 1043 eligible to be reclaimed for other purposes 1044Unevictable 1045 Memory allocated for userspace which cannot be reclaimed, such 1046 as mlocked pages, ramfs backing pages, secret memfd pages etc. 1047Mlocked 1048 Memory locked with mlock(). 1049HighTotal, HighFree 1050 Highmem is all memory above ~860MB of physical memory. 1051 Highmem areas are for use by userspace programs, or 1052 for the pagecache. The kernel must use tricks to access 1053 this memory, making it slower to access than lowmem. 1054LowTotal, LowFree 1055 Lowmem is memory which can be used for everything that 1056 highmem can be used for, but it is also available for the 1057 kernel's use for its own data structures. Among many 1058 other things, it is where everything from the Slab is 1059 allocated. Bad things happen when you're out of lowmem. 1060SwapTotal 1061 total amount of swap space available 1062SwapFree 1063 Memory which has been evicted from RAM, and is temporarily 1064 on the disk 1065Zswap 1066 Memory consumed by the zswap backend (compressed size) 1067Zswapped 1068 Amount of anonymous memory stored in zswap (original size) 1069Dirty 1070 Memory which is waiting to get written back to the disk 1071Writeback 1072 Memory which is actively being written back to the disk 1073AnonPages 1074 Non-file backed pages mapped into userspace page tables 1075Mapped 1076 files which have been mmaped, such as libraries 1077Shmem 1078 Total memory used by shared memory (shmem) and tmpfs 1079KReclaimable 1080 Kernel allocations that the kernel will attempt to reclaim 1081 under memory pressure. Includes SReclaimable (below), and other 1082 direct allocations with a shrinker. 1083Slab 1084 in-kernel data structures cache 1085SReclaimable 1086 Part of Slab, that might be reclaimed, such as caches 1087SUnreclaim 1088 Part of Slab, that cannot be reclaimed on memory pressure 1089KernelStack 1090 Memory consumed by the kernel stacks of all tasks 1091PageTables 1092 Memory consumed by userspace page tables 1093NFS_Unstable 1094 Always zero. Previous counted pages which had been written to 1095 the server, but has not been committed to stable storage. 1096Bounce 1097 Memory used for block device "bounce buffers" 1098WritebackTmp 1099 Memory used by FUSE for temporary writeback buffers 1100CommitLimit 1101 Based on the overcommit ratio ('vm.overcommit_ratio'), 1102 this is the total amount of memory currently available to 1103 be allocated on the system. This limit is only adhered to 1104 if strict overcommit accounting is enabled (mode 2 in 1105 'vm.overcommit_memory'). 1106 1107 The CommitLimit is calculated with the following formula:: 1108 1109 CommitLimit = ([total RAM pages] - [total huge TLB pages]) * 1110 overcommit_ratio / 100 + [total swap pages] 1111 1112 For example, on a system with 1G of physical RAM and 7G 1113 of swap with a `vm.overcommit_ratio` of 30 it would 1114 yield a CommitLimit of 7.3G. 1115 1116 For more details, see the memory overcommit documentation 1117 in mm/overcommit-accounting. 1118Committed_AS 1119 The amount of memory presently allocated on the system. 1120 The committed memory is a sum of all of the memory which 1121 has been allocated by processes, even if it has not been 1122 "used" by them as of yet. A process which malloc()'s 1G 1123 of memory, but only touches 300M of it will show up as 1124 using 1G. This 1G is memory which has been "committed" to 1125 by the VM and can be used at any time by the allocating 1126 application. With strict overcommit enabled on the system 1127 (mode 2 in 'vm.overcommit_memory'), allocations which would 1128 exceed the CommitLimit (detailed above) will not be permitted. 1129 This is useful if one needs to guarantee that processes will 1130 not fail due to lack of memory once that memory has been 1131 successfully allocated. 1132VmallocTotal 1133 total size of vmalloc virtual address space 1134VmallocUsed 1135 amount of vmalloc area which is used 1136VmallocChunk 1137 largest contiguous block of vmalloc area which is free 1138Percpu 1139 Memory allocated to the percpu allocator used to back percpu 1140 allocations. This stat excludes the cost of metadata. 1141HardwareCorrupted 1142 The amount of RAM/memory in KB, the kernel identifies as 1143 corrupted. 1144AnonHugePages 1145 Non-file backed huge pages mapped into userspace page tables 1146ShmemHugePages 1147 Memory used by shared memory (shmem) and tmpfs allocated 1148 with huge pages 1149ShmemPmdMapped 1150 Shared memory mapped into userspace with huge pages 1151FileHugePages 1152 Memory used for filesystem data (page cache) allocated 1153 with huge pages 1154FilePmdMapped 1155 Page cache mapped into userspace with huge pages 1156CmaTotal 1157 Memory reserved for the Contiguous Memory Allocator (CMA) 1158CmaFree 1159 Free remaining memory in the CMA reserves 1160HugePages_Total, HugePages_Free, HugePages_Rsvd, HugePages_Surp, Hugepagesize, Hugetlb 1161 See Documentation/admin-guide/mm/hugetlbpage.rst. 1162DirectMap4k, DirectMap2M, DirectMap1G 1163 Breakdown of page table sizes used in the kernel's 1164 identity mapping of RAM 1165 1166vmallocinfo 1167~~~~~~~~~~~ 1168 1169Provides information about vmalloced/vmaped areas. One line per area, 1170containing the virtual address range of the area, size in bytes, 1171caller information of the creator, and optional information depending 1172on the kind of area: 1173 1174 ========== =================================================== 1175 pages=nr number of pages 1176 phys=addr if a physical address was specified 1177 ioremap I/O mapping (ioremap() and friends) 1178 vmalloc vmalloc() area 1179 vmap vmap()ed pages 1180 user VM_USERMAP area 1181 vpages buffer for pages pointers was vmalloced (huge area) 1182 N<node>=nr (Only on NUMA kernels) 1183 Number of pages allocated on memory node <node> 1184 ========== =================================================== 1185 1186:: 1187 1188 > cat /proc/vmallocinfo 1189 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ... 1190 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128 1191 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ... 1192 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64 1193 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f... 1194 phys=7fee8000 ioremap 1195 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f... 1196 phys=7fee7000 ioremap 1197 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210 1198 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ... 1199 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3 1200 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ... 1201 pages=2 vmalloc N1=2 1202 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ... 1203 /0x130 [x_tables] pages=4 vmalloc N0=4 1204 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ... 1205 pages=14 vmalloc N2=14 1206 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ... 1207 pages=4 vmalloc N1=4 1208 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ... 1209 pages=2 vmalloc N1=2 1210 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... 1211 pages=10 vmalloc N0=10 1212 1213 1214softirqs 1215~~~~~~~~ 1216 1217Provides counts of softirq handlers serviced since boot time, for each CPU. 1218 1219:: 1220 1221 > cat /proc/softirqs 1222 CPU0 CPU1 CPU2 CPU3 1223 HI: 0 0 0 0 1224 TIMER: 27166 27120 27097 27034 1225 NET_TX: 0 0 0 17 1226 NET_RX: 42 0 0 39 1227 BLOCK: 0 0 107 1121 1228 TASKLET: 0 0 0 290 1229 SCHED: 27035 26983 26971 26746 1230 HRTIMER: 0 0 0 0 1231 RCU: 1678 1769 2178 2250 1232 12331.3 Networking info in /proc/net 1234-------------------------------- 1235 1236The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the 1237additional values you get for IP version 6 if you configure the kernel to 1238support this. Table 1-9 lists the files and their meaning. 1239 1240 1241.. table:: Table 1-8: IPv6 info in /proc/net 1242 1243 ========== ===================================================== 1244 File Content 1245 ========== ===================================================== 1246 udp6 UDP sockets (IPv6) 1247 tcp6 TCP sockets (IPv6) 1248 raw6 Raw device statistics (IPv6) 1249 igmp6 IP multicast addresses, which this host joined (IPv6) 1250 if_inet6 List of IPv6 interface addresses 1251 ipv6_route Kernel routing table for IPv6 1252 rt6_stats Global IPv6 routing tables statistics 1253 sockstat6 Socket statistics (IPv6) 1254 snmp6 Snmp data (IPv6) 1255 ========== ===================================================== 1256 1257.. table:: Table 1-9: Network info in /proc/net 1258 1259 ============= ================================================================ 1260 File Content 1261 ============= ================================================================ 1262 arp Kernel ARP table 1263 dev network devices with statistics 1264 dev_mcast the Layer2 multicast groups a device is listening too 1265 (interface index, label, number of references, number of bound 1266 addresses). 1267 dev_stat network device status 1268 ip_fwchains Firewall chain linkage 1269 ip_fwnames Firewall chain names 1270 ip_masq Directory containing the masquerading tables 1271 ip_masquerade Major masquerading table 1272 netstat Network statistics 1273 raw raw device statistics 1274 route Kernel routing table 1275 rpc Directory containing rpc info 1276 rt_cache Routing cache 1277 snmp SNMP data 1278 sockstat Socket statistics 1279 tcp TCP sockets 1280 udp UDP sockets 1281 unix UNIX domain sockets 1282 wireless Wireless interface data (Wavelan etc) 1283 igmp IP multicast addresses, which this host joined 1284 psched Global packet scheduler parameters. 1285 netlink List of PF_NETLINK sockets 1286 ip_mr_vifs List of multicast virtual interfaces 1287 ip_mr_cache List of multicast routing cache 1288 ============= ================================================================ 1289 1290You can use this information to see which network devices are available in 1291your system and how much traffic was routed over those devices:: 1292 1293 > cat /proc/net/dev 1294 Inter-|Receive |[... 1295 face |bytes packets errs drop fifo frame compressed multicast|[... 1296 lo: 908188 5596 0 0 0 0 0 0 [... 1297 ppp0:15475140 20721 410 0 0 410 0 0 [... 1298 eth0: 614530 7085 0 0 0 0 0 1 [... 1299 1300 ...] Transmit 1301 ...] bytes packets errs drop fifo colls carrier compressed 1302 ...] 908188 5596 0 0 0 0 0 0 1303 ...] 1375103 17405 0 0 0 0 0 0 1304 ...] 1703981 5535 0 0 0 3 0 0 1305 1306In addition, each Channel Bond interface has its own directory. For 1307example, the bond0 device will have a directory called /proc/net/bond0/. 1308It will contain information that is specific to that bond, such as the 1309current slaves of the bond, the link status of the slaves, and how 1310many times the slaves link has failed. 1311 13121.4 SCSI info 1313------------- 1314 1315If you have a SCSI host adapter in your system, you'll find a subdirectory 1316named after the driver for this adapter in /proc/scsi. You'll also see a list 1317of all recognized SCSI devices in /proc/scsi:: 1318 1319 >cat /proc/scsi/scsi 1320 Attached devices: 1321 Host: scsi0 Channel: 00 Id: 00 Lun: 00 1322 Vendor: IBM Model: DGHS09U Rev: 03E0 1323 Type: Direct-Access ANSI SCSI revision: 03 1324 Host: scsi0 Channel: 00 Id: 06 Lun: 00 1325 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04 1326 Type: CD-ROM ANSI SCSI revision: 02 1327 1328 1329The directory named after the driver has one file for each adapter found in 1330the system. These files contain information about the controller, including 1331the used IRQ and the IO address range. The amount of information shown is 1332dependent on the adapter you use. The example shows the output for an Adaptec 1333AHA-2940 SCSI adapter:: 1334 1335 > cat /proc/scsi/aic7xxx/0 1336 1337 Adaptec AIC7xxx driver version: 5.1.19/3.2.4 1338 Compile Options: 1339 TCQ Enabled By Default : Disabled 1340 AIC7XXX_PROC_STATS : Disabled 1341 AIC7XXX_RESET_DELAY : 5 1342 Adapter Configuration: 1343 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 1344 Ultra Wide Controller 1345 PCI MMAPed I/O Base: 0xeb001000 1346 Adapter SEEPROM Config: SEEPROM found and used. 1347 Adaptec SCSI BIOS: Enabled 1348 IRQ: 10 1349 SCBs: Active 0, Max Active 2, 1350 Allocated 15, HW 16, Page 255 1351 Interrupts: 160328 1352 BIOS Control Word: 0x18b6 1353 Adapter Control Word: 0x005b 1354 Extended Translation: Enabled 1355 Disconnect Enable Flags: 0xffff 1356 Ultra Enable Flags: 0x0001 1357 Tag Queue Enable Flags: 0x0000 1358 Ordered Queue Tag Flags: 0x0000 1359 Default Tag Queue Depth: 8 1360 Tagged Queue By Device array for aic7xxx host instance 0: 1361 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 1362 Actual queue depth per device for aic7xxx host instance 0: 1363 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 1364 Statistics: 1365 (scsi0:0:0:0) 1366 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 1367 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 1368 Total transfers 160151 (74577 reads and 85574 writes) 1369 (scsi0:0:6:0) 1370 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 1371 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 1372 Total transfers 0 (0 reads and 0 writes) 1373 1374 13751.5 Parallel port info in /proc/parport 1376--------------------------------------- 1377 1378The directory /proc/parport contains information about the parallel ports of 1379your system. It has one subdirectory for each port, named after the port 1380number (0,1,2,...). 1381 1382These directories contain the four files shown in Table 1-10. 1383 1384 1385.. table:: Table 1-10: Files in /proc/parport 1386 1387 ========= ==================================================================== 1388 File Content 1389 ========= ==================================================================== 1390 autoprobe Any IEEE-1284 device ID information that has been acquired. 1391 devices list of the device drivers using that port. A + will appear by the 1392 name of the device currently using the port (it might not appear 1393 against any). 1394 hardware Parallel port's base address, IRQ line and DMA channel. 1395 irq IRQ that parport is using for that port. This is in a separate 1396 file to allow you to alter it by writing a new value in (IRQ 1397 number or none). 1398 ========= ==================================================================== 1399 14001.6 TTY info in /proc/tty 1401------------------------- 1402 1403Information about the available and actually used tty's can be found in the 1404directory /proc/tty. You'll find entries for drivers and line disciplines in 1405this directory, as shown in Table 1-11. 1406 1407 1408.. table:: Table 1-11: Files in /proc/tty 1409 1410 ============= ============================================== 1411 File Content 1412 ============= ============================================== 1413 drivers list of drivers and their usage 1414 ldiscs registered line disciplines 1415 driver/serial usage statistic and status of single tty lines 1416 ============= ============================================== 1417 1418To see which tty's are currently in use, you can simply look into the file 1419/proc/tty/drivers:: 1420 1421 > cat /proc/tty/drivers 1422 pty_slave /dev/pts 136 0-255 pty:slave 1423 pty_master /dev/ptm 128 0-255 pty:master 1424 pty_slave /dev/ttyp 3 0-255 pty:slave 1425 pty_master /dev/pty 2 0-255 pty:master 1426 serial /dev/cua 5 64-67 serial:callout 1427 serial /dev/ttyS 4 64-67 serial 1428 /dev/tty0 /dev/tty0 4 0 system:vtmaster 1429 /dev/ptmx /dev/ptmx 5 2 system 1430 /dev/console /dev/console 5 1 system:console 1431 /dev/tty /dev/tty 5 0 system:/dev/tty 1432 unknown /dev/tty 4 1-63 console 1433 1434 14351.7 Miscellaneous kernel statistics in /proc/stat 1436------------------------------------------------- 1437 1438Various pieces of information about kernel activity are available in the 1439/proc/stat file. All of the numbers reported in this file are aggregates 1440since the system first booted. For a quick look, simply cat the file:: 1441 1442 > cat /proc/stat 1443 cpu 2255 34 2290 22625563 6290 127 456 0 0 0 1444 cpu0 1132 34 1441 11311718 3675 127 438 0 0 0 1445 cpu1 1123 0 849 11313845 2614 0 18 0 0 0 1446 intr 114930548 113199788 3 0 5 263 0 4 [... lots more numbers ...] 1447 ctxt 1990473 1448 btime 1062191376 1449 processes 2915 1450 procs_running 1 1451 procs_blocked 0 1452 softirq 183433 0 21755 12 39 1137 231 21459 2263 1453 1454The very first "cpu" line aggregates the numbers in all of the other "cpuN" 1455lines. These numbers identify the amount of time the CPU has spent performing 1456different kinds of work. Time units are in USER_HZ (typically hundredths of a 1457second). The meanings of the columns are as follows, from left to right: 1458 1459- user: normal processes executing in user mode 1460- nice: niced processes executing in user mode 1461- system: processes executing in kernel mode 1462- idle: twiddling thumbs 1463- iowait: In a word, iowait stands for waiting for I/O to complete. But there 1464 are several problems: 1465 1466 1. CPU will not wait for I/O to complete, iowait is the time that a task is 1467 waiting for I/O to complete. When CPU goes into idle state for 1468 outstanding task I/O, another task will be scheduled on this CPU. 1469 2. In a multi-core CPU, the task waiting for I/O to complete is not running 1470 on any CPU, so the iowait of each CPU is difficult to calculate. 1471 3. The value of iowait field in /proc/stat will decrease in certain 1472 conditions. 1473 1474 So, the iowait is not reliable by reading from /proc/stat. 1475- irq: servicing interrupts 1476- softirq: servicing softirqs 1477- steal: involuntary wait 1478- guest: running a normal guest 1479- guest_nice: running a niced guest 1480 1481The "intr" line gives counts of interrupts serviced since boot time, for each 1482of the possible system interrupts. The first column is the total of all 1483interrupts serviced including unnumbered architecture specific interrupts; 1484each subsequent column is the total for that particular numbered interrupt. 1485Unnumbered interrupts are not shown, only summed into the total. 1486 1487The "ctxt" line gives the total number of context switches across all CPUs. 1488 1489The "btime" line gives the time at which the system booted, in seconds since 1490the Unix epoch. 1491 1492The "processes" line gives the number of processes and threads created, which 1493includes (but is not limited to) those created by calls to the fork() and 1494clone() system calls. 1495 1496The "procs_running" line gives the total number of threads that are 1497running or ready to run (i.e., the total number of runnable threads). 1498 1499The "procs_blocked" line gives the number of processes currently blocked, 1500waiting for I/O to complete. 1501 1502The "softirq" line gives counts of softirqs serviced since boot time, for each 1503of the possible system softirqs. The first column is the total of all 1504softirqs serviced; each subsequent column is the total for that particular 1505softirq. 1506 1507 15081.8 Ext4 file system parameters 1509------------------------------- 1510 1511Information about mounted ext4 file systems can be found in 1512/proc/fs/ext4. Each mounted filesystem will have a directory in 1513/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or 1514/proc/fs/ext4/dm-0). The files in each per-device directory are shown 1515in Table 1-12, below. 1516 1517.. table:: Table 1-12: Files in /proc/fs/ext4/<devname> 1518 1519 ============== ========================================================== 1520 File Content 1521 mb_groups details of multiblock allocator buddy cache of free blocks 1522 ============== ========================================================== 1523 15241.9 /proc/consoles 1525------------------- 1526Shows registered system console lines. 1527 1528To see which character device lines are currently used for the system console 1529/dev/console, you may simply look into the file /proc/consoles:: 1530 1531 > cat /proc/consoles 1532 tty0 -WU (ECp) 4:7 1533 ttyS0 -W- (Ep) 4:64 1534 1535The columns are: 1536 1537+--------------------+-------------------------------------------------------+ 1538| device | name of the device | 1539+====================+=======================================================+ 1540| operations | * R = can do read operations | 1541| | * W = can do write operations | 1542| | * U = can do unblank | 1543+--------------------+-------------------------------------------------------+ 1544| flags | * E = it is enabled | 1545| | * C = it is preferred console | 1546| | * B = it is primary boot console | 1547| | * p = it is used for printk buffer | 1548| | * b = it is not a TTY but a Braille device | 1549| | * a = it is safe to use when cpu is offline | 1550+--------------------+-------------------------------------------------------+ 1551| major:minor | major and minor number of the device separated by a | 1552| | colon | 1553+--------------------+-------------------------------------------------------+ 1554 1555Summary 1556------- 1557 1558The /proc file system serves information about the running system. It not only 1559allows access to process data but also allows you to request the kernel status 1560by reading files in the hierarchy. 1561 1562The directory structure of /proc reflects the types of information and makes 1563it easy, if not obvious, where to look for specific data. 1564 1565Chapter 2: Modifying System Parameters 1566====================================== 1567 1568In This Chapter 1569--------------- 1570 1571* Modifying kernel parameters by writing into files found in /proc/sys 1572* Exploring the files which modify certain parameters 1573* Review of the /proc/sys file tree 1574 1575------------------------------------------------------------------------------ 1576 1577A very interesting part of /proc is the directory /proc/sys. This is not only 1578a source of information, it also allows you to change parameters within the 1579kernel. Be very careful when attempting this. You can optimize your system, 1580but you can also cause it to crash. Never alter kernel parameters on a 1581production system. Set up a development machine and test to make sure that 1582everything works the way you want it to. You may have no alternative but to 1583reboot the machine once an error has been made. 1584 1585To change a value, simply echo the new value into the file. 1586You need to be root to do this. You can create your own boot script 1587to perform this every time your system boots. 1588 1589The files in /proc/sys can be used to fine tune and monitor miscellaneous and 1590general things in the operation of the Linux kernel. Since some of the files 1591can inadvertently disrupt your system, it is advisable to read both 1592documentation and source before actually making adjustments. In any case, be 1593very careful when writing to any of these files. The entries in /proc may 1594change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt 1595review the kernel documentation in the directory /usr/src/linux/Documentation. 1596This chapter is heavily based on the documentation included in the pre 2.2 1597kernels, and became part of it in version 2.2.1 of the Linux kernel. 1598 1599Please see: Documentation/admin-guide/sysctl/ directory for descriptions of these 1600entries. 1601 1602Summary 1603------- 1604 1605Certain aspects of kernel behavior can be modified at runtime, without the 1606need to recompile the kernel, or even to reboot the system. The files in the 1607/proc/sys tree can not only be read, but also modified. You can use the echo 1608command to write value into these files, thereby changing the default settings 1609of the kernel. 1610 1611 1612Chapter 3: Per-process Parameters 1613================================= 1614 16153.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score 1616-------------------------------------------------------------------------------- 1617 1618These files can be used to adjust the badness heuristic used to select which 1619process gets killed in out of memory (oom) conditions. 1620 1621The badness heuristic assigns a value to each candidate task ranging from 0 1622(never kill) to 1000 (always kill) to determine which process is targeted. The 1623units are roughly a proportion along that range of allowed memory the process 1624may allocate from based on an estimation of its current memory and swap use. 1625For example, if a task is using all allowed memory, its badness score will be 16261000. If it is using half of its allowed memory, its score will be 500. 1627 1628The amount of "allowed" memory depends on the context in which the oom killer 1629was called. If it is due to the memory assigned to the allocating task's cpuset 1630being exhausted, the allowed memory represents the set of mems assigned to that 1631cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed 1632memory represents the set of mempolicy nodes. If it is due to a memory 1633limit (or swap limit) being reached, the allowed memory is that configured 1634limit. Finally, if it is due to the entire system being out of memory, the 1635allowed memory represents all allocatable resources. 1636 1637The value of /proc/<pid>/oom_score_adj is added to the badness score before it 1638is used to determine which task to kill. Acceptable values range from -1000 1639(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to 1640polarize the preference for oom killing either by always preferring a certain 1641task or completely disabling it. The lowest possible value, -1000, is 1642equivalent to disabling oom killing entirely for that task since it will always 1643report a badness score of 0. 1644 1645Consequently, it is very simple for userspace to define the amount of memory to 1646consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for 1647example, is roughly equivalent to allowing the remainder of tasks sharing the 1648same system, cpuset, mempolicy, or memory controller resources to use at least 164950% more memory. A value of -500, on the other hand, would be roughly 1650equivalent to discounting 50% of the task's allowed memory from being considered 1651as scoring against the task. 1652 1653For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also 1654be used to tune the badness score. Its acceptable values range from -16 1655(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17 1656(OOM_DISABLE) to disable oom killing entirely for that task. Its value is 1657scaled linearly with /proc/<pid>/oom_score_adj. 1658 1659The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last 1660value set by a CAP_SYS_RESOURCE process. To reduce the value any lower 1661requires CAP_SYS_RESOURCE. 1662 1663 16643.2 /proc/<pid>/oom_score - Display current oom-killer score 1665------------------------------------------------------------- 1666 1667This file can be used to check the current score used by the oom-killer for 1668any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which 1669process should be killed in an out-of-memory situation. 1670 1671Please note that the exported value includes oom_score_adj so it is 1672effectively in range [0,2000]. 1673 1674 16753.3 /proc/<pid>/io - Display the IO accounting fields 1676------------------------------------------------------- 1677 1678This file contains IO statistics for each running process. 1679 1680Example 1681~~~~~~~ 1682 1683:: 1684 1685 test:/tmp # dd if=/dev/zero of=/tmp/test.dat & 1686 [1] 3828 1687 1688 test:/tmp # cat /proc/3828/io 1689 rchar: 323934931 1690 wchar: 323929600 1691 syscr: 632687 1692 syscw: 632675 1693 read_bytes: 0 1694 write_bytes: 323932160 1695 cancelled_write_bytes: 0 1696 1697 1698Description 1699~~~~~~~~~~~ 1700 1701rchar 1702^^^^^ 1703 1704I/O counter: chars read 1705The number of bytes which this task has caused to be read from storage. This 1706is simply the sum of bytes which this process passed to read() and pread(). 1707It includes things like tty IO and it is unaffected by whether or not actual 1708physical disk IO was required (the read might have been satisfied from 1709pagecache). 1710 1711 1712wchar 1713^^^^^ 1714 1715I/O counter: chars written 1716The number of bytes which this task has caused, or shall cause to be written 1717to disk. Similar caveats apply here as with rchar. 1718 1719 1720syscr 1721^^^^^ 1722 1723I/O counter: read syscalls 1724Attempt to count the number of read I/O operations, i.e. syscalls like read() 1725and pread(). 1726 1727 1728syscw 1729^^^^^ 1730 1731I/O counter: write syscalls 1732Attempt to count the number of write I/O operations, i.e. syscalls like 1733write() and pwrite(). 1734 1735 1736read_bytes 1737^^^^^^^^^^ 1738 1739I/O counter: bytes read 1740Attempt to count the number of bytes which this process really did cause to 1741be fetched from the storage layer. Done at the submit_bio() level, so it is 1742accurate for block-backed filesystems. <please add status regarding NFS and 1743CIFS at a later time> 1744 1745 1746write_bytes 1747^^^^^^^^^^^ 1748 1749I/O counter: bytes written 1750Attempt to count the number of bytes which this process caused to be sent to 1751the storage layer. This is done at page-dirtying time. 1752 1753 1754cancelled_write_bytes 1755^^^^^^^^^^^^^^^^^^^^^ 1756 1757The big inaccuracy here is truncate. If a process writes 1MB to a file and 1758then deletes the file, it will in fact perform no writeout. But it will have 1759been accounted as having caused 1MB of write. 1760In other words: The number of bytes which this process caused to not happen, 1761by truncating pagecache. A task can cause "negative" IO too. If this task 1762truncates some dirty pagecache, some IO which another task has been accounted 1763for (in its write_bytes) will not be happening. We _could_ just subtract that 1764from the truncating task's write_bytes, but there is information loss in doing 1765that. 1766 1767 1768.. Note:: 1769 1770 At its current implementation state, this is a bit racy on 32-bit machines: 1771 if process A reads process B's /proc/pid/io while process B is updating one 1772 of those 64-bit counters, process A could see an intermediate result. 1773 1774 1775More information about this can be found within the taskstats documentation in 1776Documentation/accounting. 1777 17783.4 /proc/<pid>/coredump_filter - Core dump filtering settings 1779--------------------------------------------------------------- 1780When a process is dumped, all anonymous memory is written to a core file as 1781long as the size of the core file isn't limited. But sometimes we don't want 1782to dump some memory segments, for example, huge shared memory or DAX. 1783Conversely, sometimes we want to save file-backed memory segments into a core 1784file, not only the individual files. 1785 1786/proc/<pid>/coredump_filter allows you to customize which memory segments 1787will be dumped when the <pid> process is dumped. coredump_filter is a bitmask 1788of memory types. If a bit of the bitmask is set, memory segments of the 1789corresponding memory type are dumped, otherwise they are not dumped. 1790 1791The following 9 memory types are supported: 1792 1793 - (bit 0) anonymous private memory 1794 - (bit 1) anonymous shared memory 1795 - (bit 2) file-backed private memory 1796 - (bit 3) file-backed shared memory 1797 - (bit 4) ELF header pages in file-backed private memory areas (it is 1798 effective only if the bit 2 is cleared) 1799 - (bit 5) hugetlb private memory 1800 - (bit 6) hugetlb shared memory 1801 - (bit 7) DAX private memory 1802 - (bit 8) DAX shared memory 1803 1804 Note that MMIO pages such as frame buffer are never dumped and vDSO pages 1805 are always dumped regardless of the bitmask status. 1806 1807 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is 1808 only affected by bit 5-6, and DAX is only affected by bits 7-8. 1809 1810The default value of coredump_filter is 0x33; this means all anonymous memory 1811segments, ELF header pages and hugetlb private memory are dumped. 1812 1813If you don't want to dump all shared memory segments attached to pid 1234, 1814write 0x31 to the process's proc file:: 1815 1816 $ echo 0x31 > /proc/1234/coredump_filter 1817 1818When a new process is created, the process inherits the bitmask status from its 1819parent. It is useful to set up coredump_filter before the program runs. 1820For example:: 1821 1822 $ echo 0x7 > /proc/self/coredump_filter 1823 $ ./some_program 1824 18253.5 /proc/<pid>/mountinfo - Information about mounts 1826-------------------------------------------------------- 1827 1828This file contains lines of the form:: 1829 1830 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue 1831 (1)(2)(3) (4) (5) (6) (n…m) (m+1)(m+2) (m+3) (m+4) 1832 1833 (1) mount ID: unique identifier of the mount (may be reused after umount) 1834 (2) parent ID: ID of parent (or of self for the top of the mount tree) 1835 (3) major:minor: value of st_dev for files on filesystem 1836 (4) root: root of the mount within the filesystem 1837 (5) mount point: mount point relative to the process's root 1838 (6) mount options: per mount options 1839 (n…m) optional fields: zero or more fields of the form "tag[:value]" 1840 (m+1) separator: marks the end of the optional fields 1841 (m+2) filesystem type: name of filesystem of the form "type[.subtype]" 1842 (m+3) mount source: filesystem specific information or "none" 1843 (m+4) super options: per super block options 1844 1845Parsers should ignore all unrecognised optional fields. Currently the 1846possible optional fields are: 1847 1848================ ============================================================== 1849shared:X mount is shared in peer group X 1850master:X mount is slave to peer group X 1851propagate_from:X mount is slave and receives propagation from peer group X [#]_ 1852unbindable mount is unbindable 1853================ ============================================================== 1854 1855.. [#] X is the closest dominant peer group under the process's root. If 1856 X is the immediate master of the mount, or if there's no dominant peer 1857 group under the same root, then only the "master:X" field is present 1858 and not the "propagate_from:X" field. 1859 1860For more information on mount propagation see: 1861 1862 Documentation/filesystems/sharedsubtree.rst 1863 1864 18653.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 1866-------------------------------------------------------- 1867These files provide a method to access a task's comm value. It also allows for 1868a task to set its own or one of its thread siblings comm value. The comm value 1869is limited in size compared to the cmdline value, so writing anything longer 1870then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated 1871comm value. 1872 1873 18743.7 /proc/<pid>/task/<tid>/children - Information about task children 1875------------------------------------------------------------------------- 1876This file provides a fast way to retrieve first level children pids 1877of a task pointed by <pid>/<tid> pair. The format is a space separated 1878stream of pids. 1879 1880Note the "first level" here -- if a child has its own children they will 1881not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children 1882to obtain the descendants. 1883 1884Since this interface is intended to be fast and cheap it doesn't 1885guarantee to provide precise results and some children might be 1886skipped, especially if they've exited right after we printed their 1887pids, so one needs to either stop or freeze processes being inspected 1888if precise results are needed. 1889 1890 18913.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 1892--------------------------------------------------------------- 1893This file provides information associated with an opened file. The regular 1894files have at least four fields -- 'pos', 'flags', 'mnt_id' and 'ino'. 1895The 'pos' represents the current offset of the opened file in decimal 1896form [see lseek(2) for details], 'flags' denotes the octal O_xxx mask the 1897file has been created with [see open(2) for details] and 'mnt_id' represents 1898mount ID of the file system containing the opened file [see 3.5 1899/proc/<pid>/mountinfo for details]. 'ino' represents the inode number of 1900the file. 1901 1902A typical output is:: 1903 1904 pos: 0 1905 flags: 0100002 1906 mnt_id: 19 1907 ino: 63107 1908 1909All locks associated with a file descriptor are shown in its fdinfo too:: 1910 1911 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF 1912 1913The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags 1914pair provide additional information particular to the objects they represent. 1915 1916Eventfd files 1917~~~~~~~~~~~~~ 1918 1919:: 1920 1921 pos: 0 1922 flags: 04002 1923 mnt_id: 9 1924 ino: 63107 1925 eventfd-count: 5a 1926 1927where 'eventfd-count' is hex value of a counter. 1928 1929Signalfd files 1930~~~~~~~~~~~~~~ 1931 1932:: 1933 1934 pos: 0 1935 flags: 04002 1936 mnt_id: 9 1937 ino: 63107 1938 sigmask: 0000000000000200 1939 1940where 'sigmask' is hex value of the signal mask associated 1941with a file. 1942 1943Epoll files 1944~~~~~~~~~~~ 1945 1946:: 1947 1948 pos: 0 1949 flags: 02 1950 mnt_id: 9 1951 ino: 63107 1952 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7 1953 1954where 'tfd' is a target file descriptor number in decimal form, 1955'events' is events mask being watched and the 'data' is data 1956associated with a target [see epoll(7) for more details]. 1957 1958The 'pos' is current offset of the target file in decimal form 1959[see lseek(2)], 'ino' and 'sdev' are inode and device numbers 1960where target file resides, all in hex format. 1961 1962Fsnotify files 1963~~~~~~~~~~~~~~ 1964For inotify files the format is the following:: 1965 1966 pos: 0 1967 flags: 02000000 1968 mnt_id: 9 1969 ino: 63107 1970 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d 1971 1972where 'wd' is a watch descriptor in decimal form, i.e. a target file 1973descriptor number, 'ino' and 'sdev' are inode and device where the 1974target file resides and the 'mask' is the mask of events, all in hex 1975form [see inotify(7) for more details]. 1976 1977If the kernel was built with exportfs support, the path to the target 1978file is encoded as a file handle. The file handle is provided by three 1979fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex 1980format. 1981 1982If the kernel is built without exportfs support the file handle won't be 1983printed out. 1984 1985If there is no inotify mark attached yet the 'inotify' line will be omitted. 1986 1987For fanotify files the format is:: 1988 1989 pos: 0 1990 flags: 02 1991 mnt_id: 9 1992 ino: 63107 1993 fanotify flags:10 event-flags:0 1994 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003 1995 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4 1996 1997where fanotify 'flags' and 'event-flags' are values used in fanotify_init 1998call, 'mnt_id' is the mount point identifier, 'mflags' is the value of 1999flags associated with mark which are tracked separately from events 2000mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events 2001mask and 'ignored_mask' is the mask of events which are to be ignored. 2002All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask' 2003provide information about flags and mask used in fanotify_mark 2004call [see fsnotify manpage for details]. 2005 2006While the first three lines are mandatory and always printed, the rest is 2007optional and may be omitted if no marks created yet. 2008 2009Timerfd files 2010~~~~~~~~~~~~~ 2011 2012:: 2013 2014 pos: 0 2015 flags: 02 2016 mnt_id: 9 2017 ino: 63107 2018 clockid: 0 2019 ticks: 0 2020 settime flags: 01 2021 it_value: (0, 49406829) 2022 it_interval: (1, 0) 2023 2024where 'clockid' is the clock type and 'ticks' is the number of the timer expirations 2025that have occurred [see timerfd_create(2) for details]. 'settime flags' are 2026flags in octal form been used to setup the timer [see timerfd_settime(2) for 2027details]. 'it_value' is remaining time until the timer expiration. 2028'it_interval' is the interval for the timer. Note the timer might be set up 2029with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' 2030still exhibits timer's remaining time. 2031 2032DMA Buffer files 2033~~~~~~~~~~~~~~~~ 2034 2035:: 2036 2037 pos: 0 2038 flags: 04002 2039 mnt_id: 9 2040 ino: 63107 2041 size: 32768 2042 count: 2 2043 exp_name: system-heap 2044 2045where 'size' is the size of the DMA buffer in bytes. 'count' is the file count of 2046the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter. 2047 20483.9 /proc/<pid>/map_files - Information about memory mapped files 2049--------------------------------------------------------------------- 2050This directory contains symbolic links which represent memory mapped files 2051the process is maintaining. Example output:: 2052 2053 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so 2054 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so 2055 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so 2056 | ... 2057 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1 2058 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls 2059 2060The name of a link represents the virtual memory bounds of a mapping, i.e. 2061vm_area_struct::vm_start-vm_area_struct::vm_end. 2062 2063The main purpose of the map_files is to retrieve a set of memory mapped 2064files in a fast way instead of parsing /proc/<pid>/maps or 2065/proc/<pid>/smaps, both of which contain many more records. At the same 2066time one can open(2) mappings from the listings of two processes and 2067comparing their inode numbers to figure out which anonymous memory areas 2068are actually shared. 2069 20703.10 /proc/<pid>/timerslack_ns - Task timerslack value 2071--------------------------------------------------------- 2072This file provides the value of the task's timerslack value in nanoseconds. 2073This value specifies an amount of time that normal timers may be deferred 2074in order to coalesce timers and avoid unnecessary wakeups. 2075 2076This allows a task's interactivity vs power consumption tradeoff to be 2077adjusted. 2078 2079Writing 0 to the file will set the task's timerslack to the default value. 2080 2081Valid values are from 0 - ULLONG_MAX 2082 2083An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level 2084permissions on the task specified to change its timerslack_ns value. 2085 20863.11 /proc/<pid>/patch_state - Livepatch patch operation state 2087----------------------------------------------------------------- 2088When CONFIG_LIVEPATCH is enabled, this file displays the value of the 2089patch state for the task. 2090 2091A value of '-1' indicates that no patch is in transition. 2092 2093A value of '0' indicates that a patch is in transition and the task is 2094unpatched. If the patch is being enabled, then the task hasn't been 2095patched yet. If the patch is being disabled, then the task has already 2096been unpatched. 2097 2098A value of '1' indicates that a patch is in transition and the task is 2099patched. If the patch is being enabled, then the task has already been 2100patched. If the patch is being disabled, then the task hasn't been 2101unpatched yet. 2102 21033.12 /proc/<pid>/arch_status - task architecture specific status 2104------------------------------------------------------------------- 2105When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the 2106architecture specific status of the task. 2107 2108Example 2109~~~~~~~ 2110 2111:: 2112 2113 $ cat /proc/6753/arch_status 2114 AVX512_elapsed_ms: 8 2115 2116Description 2117~~~~~~~~~~~ 2118 2119x86 specific entries 2120~~~~~~~~~~~~~~~~~~~~~ 2121 2122AVX512_elapsed_ms 2123^^^^^^^^^^^^^^^^^^ 2124 2125 If AVX512 is supported on the machine, this entry shows the milliseconds 2126 elapsed since the last time AVX512 usage was recorded. The recording 2127 happens on a best effort basis when a task is scheduled out. This means 2128 that the value depends on two factors: 2129 2130 1) The time which the task spent on the CPU without being scheduled 2131 out. With CPU isolation and a single runnable task this can take 2132 several seconds. 2133 2134 2) The time since the task was scheduled out last. Depending on the 2135 reason for being scheduled out (time slice exhausted, syscall ...) 2136 this can be arbitrary long time. 2137 2138 As a consequence the value cannot be considered precise and authoritative 2139 information. The application which uses this information has to be aware 2140 of the overall scenario on the system in order to determine whether a 2141 task is a real AVX512 user or not. Precise information can be obtained 2142 with performance counters. 2143 2144 A special value of '-1' indicates that no AVX512 usage was recorded, thus 2145 the task is unlikely an AVX512 user, but depends on the workload and the 2146 scheduling scenario, it also could be a false negative mentioned above. 2147 2148Chapter 4: Configuring procfs 2149============================= 2150 21514.1 Mount options 2152--------------------- 2153 2154The following mount options are supported: 2155 2156 ========= ======================================================== 2157 hidepid= Set /proc/<pid>/ access mode. 2158 gid= Set the group authorized to learn processes information. 2159 subset= Show only the specified subset of procfs. 2160 ========= ======================================================== 2161 2162hidepid=off or hidepid=0 means classic mode - everybody may access all 2163/proc/<pid>/ directories (default). 2164 2165hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/ 2166directories but their own. Sensitive files like cmdline, sched*, status are now 2167protected against other users. This makes it impossible to learn whether any 2168user runs specific program (given the program doesn't reveal itself by its 2169behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for 2170other users, poorly written programs passing sensitive information via program 2171arguments are now protected against local eavesdroppers. 2172 2173hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be 2174fully invisible to other users. It doesn't mean that it hides a fact whether a 2175process with a specific pid value exists (it can be learned by other means, e.g. 2176by "kill -0 $PID"), but it hides process' uid and gid, which may be learned by 2177stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of 2178gathering information about running processes, whether some daemon runs with 2179elevated privileges, whether other user runs some sensitive program, whether 2180other users run any program at all, etc. 2181 2182hidepid=ptraceable or hidepid=4 means that procfs should only contain 2183/proc/<pid>/ directories that the caller can ptrace. 2184 2185gid= defines a group authorized to learn processes information otherwise 2186prohibited by hidepid=. If you use some daemon like identd which needs to learn 2187information about processes information, just add identd to this group. 2188 2189subset=pid hides all top level files and directories in the procfs that 2190are not related to tasks. 2191 2192Chapter 5: Filesystem behavior 2193============================== 2194 2195Originally, before the advent of pid namepsace, procfs was a global file 2196system. It means that there was only one procfs instance in the system. 2197 2198When pid namespace was added, a separate procfs instance was mounted in 2199each pid namespace. So, procfs mount options are global among all 2200mountpoints within the same namespace:: 2201 2202 # grep ^proc /proc/mounts 2203 proc /proc proc rw,relatime,hidepid=2 0 0 2204 2205 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc 2206 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 0 2207 +++ exited with 0 +++ 2208 2209 # grep ^proc /proc/mounts 2210 proc /proc proc rw,relatime,hidepid=2 0 0 2211 proc /tmp/proc proc rw,relatime,hidepid=2 0 0 2212 2213and only after remounting procfs mount options will change at all 2214mountpoints:: 2215 2216 # mount -o remount,hidepid=1 -t proc proc /tmp/proc 2217 2218 # grep ^proc /proc/mounts 2219 proc /proc proc rw,relatime,hidepid=1 0 0 2220 proc /tmp/proc proc rw,relatime,hidepid=1 0 0 2221 2222This behavior is different from the behavior of other filesystems. 2223 2224The new procfs behavior is more like other filesystems. Each procfs mount 2225creates a new procfs instance. Mount options affect own procfs instance. 2226It means that it became possible to have several procfs instances 2227displaying tasks with different filtering options in one pid namespace:: 2228 2229 # mount -o hidepid=invisible -t proc proc /proc 2230 # mount -o hidepid=noaccess -t proc proc /tmp/proc 2231 # grep ^proc /proc/mounts 2232 proc /proc proc rw,relatime,hidepid=invisible 0 0 2233 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 0 2234