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 3.13 /proc/<pid>/fd - List of symlinks to open files 51 52 4 Configuring procfs 53 4.1 Mount options 54 55 5 Filesystem behavior 56 57Preface 58======= 59 600.1 Introduction/Credits 61------------------------ 62 63This documentation is part of a soon (or so we hope) to be released book on 64the SuSE Linux distribution. As there is no complete documentation for the 65/proc file system and we've used many freely available sources to write these 66chapters, it seems only fair to give the work back to the Linux community. 67This work is based on the 2.2.* kernel version and the upcoming 2.4.*. I'm 68afraid it's still far from complete, but we hope it will be useful. As far as 69we know, it is the first 'all-in-one' document about the /proc file system. It 70is focused on the Intel x86 hardware, so if you are looking for PPC, ARM, 71SPARC, AXP, etc., features, you probably won't find what you are looking for. 72It also only covers IPv4 networking, not IPv6 nor other protocols - sorry. But 73additions and patches are welcome and will be added to this document if you 74mail them to Bodo. 75 76We'd like to thank Alan Cox, Rik van Riel, and Alexey Kuznetsov and a lot of 77other people for help compiling this documentation. We'd also like to extend a 78special thank you to Andi Kleen for documentation, which we relied on heavily 79to create this document, as well as the additional information he provided. 80Thanks to everybody else who contributed source or docs to the Linux kernel 81and helped create a great piece of software... :) 82 83If you have any comments, corrections or additions, please don't hesitate to 84contact Bodo Bauer at bb@ricochet.net. We'll be happy to add them to this 85document. 86 87The latest version of this document is available online at 88https://www.kernel.org/doc/html/latest/filesystems/proc.html 89 90If the above direction does not works for you, you could try the kernel 91mailing list at linux-kernel@vger.kernel.org and/or try to reach me at 92comandante@zaralinux.com. 93 940.2 Legal Stuff 95--------------- 96 97We don't guarantee the correctness of this document, and if you come to us 98complaining about how you screwed up your system because of incorrect 99documentation, we won't feel responsible... 100 101Chapter 1: Collecting System Information 102======================================== 103 104In This Chapter 105--------------- 106* Investigating the properties of the pseudo file system /proc and its 107 ability to provide information on the running Linux system 108* Examining /proc's structure 109* Uncovering various information about the kernel and the processes running 110 on the system 111 112------------------------------------------------------------------------------ 113 114The proc file system acts as an interface to internal data structures in the 115kernel. It can be used to obtain information about the system and to change 116certain kernel parameters at runtime (sysctl). 117 118First, we'll take a look at the read-only parts of /proc. In Chapter 2, we 119show you how you can use /proc/sys to change settings. 120 1211.1 Process-Specific Subdirectories 122----------------------------------- 123 124The directory /proc contains (among other things) one subdirectory for each 125process running on the system, which is named after the process ID (PID). 126 127The link 'self' points to the process reading the file system. Each process 128subdirectory has the entries listed in Table 1-1. 129 130Note that an open file descriptor to /proc/<pid> or to any of its 131contained files or subdirectories does not prevent <pid> being reused 132for some other process in the event that <pid> exits. Operations on 133open /proc/<pid> file descriptors corresponding to dead processes 134never act on any new process that the kernel may, through chance, have 135also assigned the process ID <pid>. Instead, operations on these FDs 136usually fail with ESRCH. 137 138.. table:: Table 1-1: Process specific entries in /proc 139 140 ============= =============================================================== 141 File Content 142 ============= =============================================================== 143 clear_refs Clears page referenced bits shown in smaps output 144 cmdline Command line arguments 145 cpu Current and last cpu in which it was executed (2.4)(smp) 146 cwd Link to the current working directory 147 environ Values of environment variables 148 exe Link to the executable of this process 149 fd Directory, which contains all file descriptors 150 maps Memory maps to executables and library files (2.4) 151 mem Memory held by this process 152 root Link to the root directory of this process 153 stat Process status 154 statm Process memory status information 155 status Process status in human readable form 156 wchan Present with CONFIG_KALLSYMS=y: it shows the kernel function 157 symbol the task is blocked in - or "0" if not blocked. 158 pagemap Page table 159 stack Report full stack trace, enable via CONFIG_STACKTRACE 160 smaps An extension based on maps, showing the memory consumption of 161 each mapping and flags associated with it 162 smaps_rollup Accumulated smaps stats for all mappings of the process. This 163 can be derived from smaps, but is faster and more convenient 164 numa_maps An extension based on maps, showing the memory locality and 165 binding policy as well as mem usage (in pages) of each mapping. 166 ============= =============================================================== 167 168For example, to get the status information of a process, all you have to do is 169read the file /proc/PID/status:: 170 171 >cat /proc/self/status 172 Name: cat 173 State: R (running) 174 Tgid: 5452 175 Pid: 5452 176 PPid: 743 177 TracerPid: 0 (2.4) 178 Uid: 501 501 501 501 179 Gid: 100 100 100 100 180 FDSize: 256 181 Groups: 100 14 16 182 Kthread: 0 183 VmPeak: 5004 kB 184 VmSize: 5004 kB 185 VmLck: 0 kB 186 VmHWM: 476 kB 187 VmRSS: 476 kB 188 RssAnon: 352 kB 189 RssFile: 120 kB 190 RssShmem: 4 kB 191 VmData: 156 kB 192 VmStk: 88 kB 193 VmExe: 68 kB 194 VmLib: 1412 kB 195 VmPTE: 20 kb 196 VmSwap: 0 kB 197 HugetlbPages: 0 kB 198 CoreDumping: 0 199 THP_enabled: 1 200 Threads: 1 201 SigQ: 0/28578 202 SigPnd: 0000000000000000 203 ShdPnd: 0000000000000000 204 SigBlk: 0000000000000000 205 SigIgn: 0000000000000000 206 SigCgt: 0000000000000000 207 CapInh: 00000000fffffeff 208 CapPrm: 0000000000000000 209 CapEff: 0000000000000000 210 CapBnd: ffffffffffffffff 211 CapAmb: 0000000000000000 212 NoNewPrivs: 0 213 Seccomp: 0 214 Speculation_Store_Bypass: thread vulnerable 215 SpeculationIndirectBranch: conditional enabled 216 voluntary_ctxt_switches: 0 217 nonvoluntary_ctxt_switches: 1 218 219This shows you nearly the same information you would get if you viewed it with 220the ps command. In fact, ps uses the proc file system to obtain its 221information. But you get a more detailed view of the process by reading the 222file /proc/PID/status. It fields are described in table 1-2. 223 224The statm file contains more detailed information about the process 225memory usage. Its seven fields are explained in Table 1-3. The stat file 226contains detailed information about the process itself. Its fields are 227explained in Table 1-4. 228 229(for SMP CONFIG users) 230 231For making accounting scalable, RSS related information are handled in an 232asynchronous manner and the value may not be very precise. To see a precise 233snapshot of a moment, you can see /proc/<pid>/smaps file and scan page table. 234It's slow but very precise. 235 236.. table:: Table 1-2: Contents of the status fields (as of 4.19) 237 238 ========================== =================================================== 239 Field Content 240 ========================== =================================================== 241 Name filename of the executable 242 Umask file mode creation mask 243 State state (R is running, S is sleeping, D is sleeping 244 in an uninterruptible wait, Z is zombie, 245 T is traced or stopped) 246 Tgid thread group ID 247 Ngid NUMA group ID (0 if none) 248 Pid process id 249 PPid process id of the parent process 250 TracerPid PID of process tracing this process (0 if not, or 251 the tracer is outside of the current pid namespace) 252 Uid Real, effective, saved set, and file system UIDs 253 Gid Real, effective, saved set, and file system GIDs 254 FDSize number of file descriptor slots currently allocated 255 Groups supplementary group list 256 NStgid descendant namespace thread group ID hierarchy 257 NSpid descendant namespace process ID hierarchy 258 NSpgid descendant namespace process group ID hierarchy 259 NSsid descendant namespace session ID hierarchy 260 Kthread kernel thread flag, 1 is yes, 0 is no 261 VmPeak peak virtual memory size 262 VmSize total program size 263 VmLck locked memory size 264 VmPin pinned memory size 265 VmHWM peak resident set size ("high water mark") 266 VmRSS size of memory portions. It contains the three 267 following parts 268 (VmRSS = RssAnon + RssFile + RssShmem) 269 RssAnon size of resident anonymous memory 270 RssFile size of resident file mappings 271 RssShmem size of resident shmem memory (includes SysV shm, 272 mapping of tmpfs and shared anonymous mappings) 273 VmData size of private data segments 274 VmStk size of stack segments 275 VmExe size of text segment 276 VmLib size of shared library code 277 VmPTE size of page table entries 278 VmSwap amount of swap used by anonymous private data 279 (shmem swap usage is not included) 280 HugetlbPages size of hugetlb memory portions 281 CoreDumping process's memory is currently being dumped 282 (killing the process may lead to a corrupted core) 283 THP_enabled process is allowed to use THP (returns 0 when 284 PR_SET_THP_DISABLE is set on the process 285 Threads number of threads 286 SigQ number of signals queued/max. number for queue 287 SigPnd bitmap of pending signals for the thread 288 ShdPnd bitmap of shared pending signals for the process 289 SigBlk bitmap of blocked signals 290 SigIgn bitmap of ignored signals 291 SigCgt bitmap of caught signals 292 CapInh bitmap of inheritable capabilities 293 CapPrm bitmap of permitted capabilities 294 CapEff bitmap of effective capabilities 295 CapBnd bitmap of capabilities bounding set 296 CapAmb bitmap of ambient capabilities 297 NoNewPrivs no_new_privs, like prctl(PR_GET_NO_NEW_PRIV, ...) 298 Seccomp seccomp mode, like prctl(PR_GET_SECCOMP, ...) 299 Speculation_Store_Bypass speculative store bypass mitigation status 300 SpeculationIndirectBranch indirect branch speculation mode 301 Cpus_allowed mask of CPUs on which this process may run 302 Cpus_allowed_list Same as previous, but in "list format" 303 Mems_allowed mask of memory nodes allowed to this process 304 Mems_allowed_list Same as previous, but in "list format" 305 voluntary_ctxt_switches number of voluntary context switches 306 nonvoluntary_ctxt_switches number of non voluntary context switches 307 ========================== =================================================== 308 309 310.. table:: Table 1-3: Contents of the statm fields (as of 2.6.8-rc3) 311 312 ======== =============================== ============================== 313 Field Content 314 ======== =============================== ============================== 315 size total program size (pages) (same as VmSize in status) 316 resident size of memory portions (pages) (same as VmRSS in status) 317 shared number of pages that are shared (i.e. backed by a file, same 318 as RssFile+RssShmem in status) 319 trs number of pages that are 'code' (not including libs; broken, 320 includes data segment) 321 lrs number of pages of library (always 0 on 2.6) 322 drs number of pages of data/stack (including libs; broken, 323 includes library text) 324 dt number of dirty pages (always 0 on 2.6) 325 ======== =============================== ============================== 326 327 328.. table:: Table 1-4: Contents of the stat fields (as of 2.6.30-rc7) 329 330 ============= =============================================================== 331 Field Content 332 ============= =============================================================== 333 pid process id 334 tcomm filename of the executable 335 state state (R is running, S is sleeping, D is sleeping in an 336 uninterruptible wait, Z is zombie, T is traced or stopped) 337 ppid process id of the parent process 338 pgrp pgrp of the process 339 sid session id 340 tty_nr tty the process uses 341 tty_pgrp pgrp of the tty 342 flags task flags 343 min_flt number of minor faults 344 cmin_flt number of minor faults with child's 345 maj_flt number of major faults 346 cmaj_flt number of major faults with child's 347 utime user mode jiffies 348 stime kernel mode jiffies 349 cutime user mode jiffies with child's 350 cstime kernel mode jiffies with child's 351 priority priority level 352 nice nice level 353 num_threads number of threads 354 it_real_value (obsolete, always 0) 355 start_time time the process started after system boot 356 vsize virtual memory size 357 rss resident set memory size 358 rsslim current limit in bytes on the rss 359 start_code address above which program text can run 360 end_code address below which program text can run 361 start_stack address of the start of the main process stack 362 esp current value of ESP 363 eip current value of EIP 364 pending bitmap of pending signals 365 blocked bitmap of blocked signals 366 sigign bitmap of ignored signals 367 sigcatch bitmap of caught signals 368 0 (place holder, used to be the wchan address, 369 use /proc/PID/wchan instead) 370 0 (place holder) 371 0 (place holder) 372 exit_signal signal to send to parent thread on exit 373 task_cpu which CPU the task is scheduled on 374 rt_priority realtime priority 375 policy scheduling policy (man sched_setscheduler) 376 blkio_ticks time spent waiting for block IO 377 gtime guest time of the task in jiffies 378 cgtime guest time of the task children in jiffies 379 start_data address above which program data+bss is placed 380 end_data address below which program data+bss is placed 381 start_brk address above which program heap can be expanded with brk() 382 arg_start address above which program command line is placed 383 arg_end address below which program command line is placed 384 env_start address above which program environment is placed 385 env_end address below which program environment is placed 386 exit_code the thread's exit_code in the form reported by the waitpid 387 system call 388 ============= =============================================================== 389 390The /proc/PID/maps file contains the currently mapped memory regions and 391their access permissions. 392 393The format is:: 394 395 address perms offset dev inode pathname 396 397 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test 398 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test 399 0804a000-0806b000 rw-p 00000000 00:00 0 [heap] 400 a7cb1000-a7cb2000 ---p 00000000 00:00 0 401 a7cb2000-a7eb2000 rw-p 00000000 00:00 0 402 a7eb2000-a7eb3000 ---p 00000000 00:00 0 403 a7eb3000-a7ed5000 rw-p 00000000 00:00 0 404 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6 405 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6 406 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6 407 a800b000-a800e000 rw-p 00000000 00:00 0 408 a800e000-a8022000 r-xp 00000000 03:00 14462 /lib/libpthread.so.0 409 a8022000-a8023000 r--p 00013000 03:00 14462 /lib/libpthread.so.0 410 a8023000-a8024000 rw-p 00014000 03:00 14462 /lib/libpthread.so.0 411 a8024000-a8027000 rw-p 00000000 00:00 0 412 a8027000-a8043000 r-xp 00000000 03:00 8317 /lib/ld-linux.so.2 413 a8043000-a8044000 r--p 0001b000 03:00 8317 /lib/ld-linux.so.2 414 a8044000-a8045000 rw-p 0001c000 03:00 8317 /lib/ld-linux.so.2 415 aff35000-aff4a000 rw-p 00000000 00:00 0 [stack] 416 ffffe000-fffff000 r-xp 00000000 00:00 0 [vdso] 417 418where "address" is the address space in the process that it occupies, "perms" 419is a set of permissions:: 420 421 r = read 422 w = write 423 x = execute 424 s = shared 425 p = private (copy on write) 426 427"offset" is the offset into the mapping, "dev" is the device (major:minor), and 428"inode" is the inode on that device. 0 indicates that no inode is associated 429with the memory region, as the case would be with BSS (uninitialized data). 430The "pathname" shows the name associated file for this mapping. If the mapping 431is not associated with a file: 432 433 =================== =========================================== 434 [heap] the heap of the program 435 [stack] the stack of the main process 436 [vdso] the "virtual dynamic shared object", 437 the kernel system call handler 438 [anon:<name>] a private anonymous mapping that has been 439 named by userspace 440 [anon_shmem:<name>] an anonymous shared memory mapping that has 441 been named by userspace 442 =================== =========================================== 443 444 or if empty, the mapping is anonymous. 445 446The /proc/PID/smaps is an extension based on maps, showing the memory 447consumption for each of the process's mappings. For each mapping (aka Virtual 448Memory Area, or VMA) there is a series of lines such as the following:: 449 450 08048000-080bc000 r-xp 00000000 03:02 13130 /bin/bash 451 452 Size: 1084 kB 453 KernelPageSize: 4 kB 454 MMUPageSize: 4 kB 455 Rss: 892 kB 456 Pss: 374 kB 457 Pss_Dirty: 0 kB 458 Shared_Clean: 892 kB 459 Shared_Dirty: 0 kB 460 Private_Clean: 0 kB 461 Private_Dirty: 0 kB 462 Referenced: 892 kB 463 Anonymous: 0 kB 464 KSM: 0 kB 465 LazyFree: 0 kB 466 AnonHugePages: 0 kB 467 ShmemPmdMapped: 0 kB 468 Shared_Hugetlb: 0 kB 469 Private_Hugetlb: 0 kB 470 Swap: 0 kB 471 SwapPss: 0 kB 472 KernelPageSize: 4 kB 473 MMUPageSize: 4 kB 474 Locked: 0 kB 475 THPeligible: 0 476 VmFlags: rd ex mr mw me dw 477 478The first of these lines shows the same information as is displayed for the 479mapping in /proc/PID/maps. Following lines show the size of the mapping 480(size); the size of each page allocated when backing a VMA (KernelPageSize), 481which is usually the same as the size in the page table entries; the page size 482used by the MMU when backing a VMA (in most cases, the same as KernelPageSize); 483the amount of the mapping that is currently resident in RAM (RSS); the 484process' proportional share of this mapping (PSS); and the number of clean and 485dirty shared and private pages in the mapping. 486 487The "proportional set size" (PSS) of a process is the count of pages it has 488in memory, where each page is divided by the number of processes sharing it. 489So if a process has 1000 pages all to itself, and 1000 shared with one other 490process, its PSS will be 1500. "Pss_Dirty" is the portion of PSS which 491consists of dirty pages. ("Pss_Clean" is not included, but it can be 492calculated by subtracting "Pss_Dirty" from "Pss".) 493 494Note that even a page which is part of a MAP_SHARED mapping, but has only 495a single pte mapped, i.e. is currently used by only one process, is accounted 496as private and not as shared. 497 498"Referenced" indicates the amount of memory currently marked as referenced or 499accessed. 500 501"Anonymous" shows the amount of memory that does not belong to any file. Even 502a mapping associated with a file may contain anonymous pages: when MAP_PRIVATE 503and a page is modified, the file page is replaced by a private anonymous copy. 504 505"KSM" reports how many of the pages are KSM pages. Note that KSM-placed zeropages 506are not included, only actual KSM pages. 507 508"LazyFree" shows the amount of memory which is marked by madvise(MADV_FREE). 509The memory isn't freed immediately with madvise(). It's freed in memory 510pressure if the memory is clean. Please note that the printed value might 511be lower than the real value due to optimizations used in the current 512implementation. If this is not desirable please file a bug report. 513 514"AnonHugePages" shows the amount of memory backed by transparent hugepage. 515 516"ShmemPmdMapped" shows the amount of shared (shmem/tmpfs) memory backed by 517huge pages. 518 519"Shared_Hugetlb" and "Private_Hugetlb" show the amounts of memory backed by 520hugetlbfs page which is *not* counted in "RSS" or "PSS" field for historical 521reasons. And these are not included in {Shared,Private}_{Clean,Dirty} field. 522 523"Swap" shows how much would-be-anonymous memory is also used, but out on swap. 524 525For shmem mappings, "Swap" includes also the size of the mapped (and not 526replaced by copy-on-write) part of the underlying shmem object out on swap. 527"SwapPss" shows proportional swap share of this mapping. Unlike "Swap", this 528does not take into account swapped out page of underlying shmem objects. 529"Locked" indicates whether the mapping is locked in memory or not. 530 531"THPeligible" indicates whether the mapping is eligible for allocating 532naturally aligned THP pages of any currently enabled size. 1 if true, 0 533otherwise. 534 535"VmFlags" field deserves a separate description. This member represents the 536kernel flags associated with the particular virtual memory area in two letter 537encoded manner. The codes are the following: 538 539 == ======================================= 540 rd readable 541 wr writeable 542 ex executable 543 sh shared 544 mr may read 545 mw may write 546 me may execute 547 ms may share 548 gd stack segment growns down 549 pf pure PFN range 550 dw disabled write to the mapped file 551 lo pages are locked in memory 552 io memory mapped I/O area 553 sr sequential read advise provided 554 rr random read advise provided 555 dc do not copy area on fork 556 de do not expand area on remapping 557 ac area is accountable 558 nr swap space is not reserved for the area 559 ht area uses huge tlb pages 560 sf synchronous page fault 561 ar architecture specific flag 562 wf wipe on fork 563 dd do not include area into core dump 564 sd soft dirty flag 565 mm mixed map area 566 hg huge page advise flag 567 nh no huge page advise flag 568 mg mergeable advise flag 569 bt arm64 BTI guarded page 570 mt arm64 MTE allocation tags are enabled 571 um userfaultfd missing tracking 572 uw userfaultfd wr-protect tracking 573 ss shadow stack page 574 == ======================================= 575 576Note that there is no guarantee that every flag and associated mnemonic will 577be present in all further kernel releases. Things get changed, the flags may 578be vanished or the reverse -- new added. Interpretation of their meaning 579might change in future as well. So each consumer of these flags has to 580follow each specific kernel version for the exact semantic. 581 582This file is only present if the CONFIG_MMU kernel configuration option is 583enabled. 584 585Note: reading /proc/PID/maps or /proc/PID/smaps is inherently racy (consistent 586output can be achieved only in the single read call). 587 588This typically manifests when doing partial reads of these files while the 589memory map is being modified. Despite the races, we do provide the following 590guarantees: 591 5921) The mapped addresses never go backwards, which implies no two 593 regions will ever overlap. 5942) If there is something at a given vaddr during the entirety of the 595 life of the smaps/maps walk, there will be some output for it. 596 597The /proc/PID/smaps_rollup file includes the same fields as /proc/PID/smaps, 598but their values are the sums of the corresponding values for all mappings of 599the process. Additionally, it contains these fields: 600 601- Pss_Anon 602- Pss_File 603- Pss_Shmem 604 605They represent the proportional shares of anonymous, file, and shmem pages, as 606described for smaps above. These fields are omitted in smaps since each 607mapping identifies the type (anon, file, or shmem) of all pages it contains. 608Thus all information in smaps_rollup can be derived from smaps, but at a 609significantly higher cost. 610 611The /proc/PID/clear_refs is used to reset the PG_Referenced and ACCESSED/YOUNG 612bits on both physical and virtual pages associated with a process, and the 613soft-dirty bit on pte (see Documentation/admin-guide/mm/soft-dirty.rst 614for details). 615To clear the bits for all the pages associated with the process:: 616 617 > echo 1 > /proc/PID/clear_refs 618 619To clear the bits for the anonymous pages associated with the process:: 620 621 > echo 2 > /proc/PID/clear_refs 622 623To clear the bits for the file mapped pages associated with the process:: 624 625 > echo 3 > /proc/PID/clear_refs 626 627To clear the soft-dirty bit:: 628 629 > echo 4 > /proc/PID/clear_refs 630 631To reset the peak resident set size ("high water mark") to the process's 632current value:: 633 634 > echo 5 > /proc/PID/clear_refs 635 636Any other value written to /proc/PID/clear_refs will have no effect. 637 638The /proc/pid/pagemap gives the PFN, which can be used to find the pageflags 639using /proc/kpageflags and number of times a page is mapped using 640/proc/kpagecount. For detailed explanation, see 641Documentation/admin-guide/mm/pagemap.rst. 642 643The /proc/pid/numa_maps is an extension based on maps, showing the memory 644locality and binding policy, as well as the memory usage (in pages) of 645each mapping. The output follows a general format where mapping details get 646summarized separated by blank spaces, one mapping per each file line:: 647 648 address policy mapping details 649 650 00400000 default file=/usr/local/bin/app mapped=1 active=0 N3=1 kernelpagesize_kB=4 651 00600000 default file=/usr/local/bin/app anon=1 dirty=1 N3=1 kernelpagesize_kB=4 652 3206000000 default file=/lib64/ld-2.12.so mapped=26 mapmax=6 N0=24 N3=2 kernelpagesize_kB=4 653 320621f000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 654 3206220000 default file=/lib64/ld-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 655 3206221000 default anon=1 dirty=1 N3=1 kernelpagesize_kB=4 656 3206800000 default file=/lib64/libc-2.12.so mapped=59 mapmax=21 active=55 N0=41 N3=18 kernelpagesize_kB=4 657 320698b000 default file=/lib64/libc-2.12.so 658 3206b8a000 default file=/lib64/libc-2.12.so anon=2 dirty=2 N3=2 kernelpagesize_kB=4 659 3206b8e000 default file=/lib64/libc-2.12.so anon=1 dirty=1 N3=1 kernelpagesize_kB=4 660 3206b8f000 default anon=3 dirty=3 active=1 N3=3 kernelpagesize_kB=4 661 7f4dc10a2000 default anon=3 dirty=3 N3=3 kernelpagesize_kB=4 662 7f4dc10b4000 default anon=2 dirty=2 active=1 N3=2 kernelpagesize_kB=4 663 7f4dc1200000 default file=/anon_hugepage\040(deleted) huge anon=1 dirty=1 N3=1 kernelpagesize_kB=2048 664 7fff335f0000 default stack anon=3 dirty=3 N3=3 kernelpagesize_kB=4 665 7fff3369d000 default mapped=1 mapmax=35 active=0 N3=1 kernelpagesize_kB=4 666 667Where: 668 669"address" is the starting address for the mapping; 670 671"policy" reports the NUMA memory policy set for the mapping (see Documentation/admin-guide/mm/numa_memory_policy.rst); 672 673"mapping details" summarizes mapping data such as mapping type, page usage counters, 674node locality page counters (N0 == node0, N1 == node1, ...) and the kernel page 675size, in KB, that is backing the mapping up. 676 6771.2 Kernel data 678--------------- 679 680Similar to the process entries, the kernel data files give information about 681the running kernel. The files used to obtain this information are contained in 682/proc and are listed in Table 1-5. Not all of these will be present in your 683system. It depends on the kernel configuration and the loaded modules, which 684files are there, and which are missing. 685 686.. table:: Table 1-5: Kernel info in /proc 687 688 ============ =============================================================== 689 File Content 690 ============ =============================================================== 691 allocinfo Memory allocations profiling information 692 apm Advanced power management info 693 bootconfig Kernel command line obtained from boot config, 694 and, if there were kernel parameters from the 695 boot loader, a "# Parameters from bootloader:" 696 line followed by a line containing those 697 parameters prefixed by "# ". (5.5) 698 buddyinfo Kernel memory allocator information (see text) (2.5) 699 bus Directory containing bus specific information 700 cmdline Kernel command line, both from bootloader and embedded 701 in the kernel image 702 cpuinfo Info about the CPU 703 devices Available devices (block and character) 704 dma Used DMS channels 705 filesystems Supported filesystems 706 driver Various drivers grouped here, currently rtc (2.4) 707 execdomains Execdomains, related to security (2.4) 708 fb Frame Buffer devices (2.4) 709 fs File system parameters, currently nfs/exports (2.4) 710 ide Directory containing info about the IDE subsystem 711 interrupts Interrupt usage 712 iomem Memory map (2.4) 713 ioports I/O port usage 714 irq Masks for irq to cpu affinity (2.4)(smp?) 715 isapnp ISA PnP (Plug&Play) Info (2.4) 716 kcore Kernel core image (can be ELF or A.OUT(deprecated in 2.4)) 717 kmsg Kernel messages 718 ksyms Kernel symbol table 719 loadavg Load average of last 1, 5 & 15 minutes; 720 number of processes currently runnable (running or on ready queue); 721 total number of processes in system; 722 last pid created. 723 All fields are separated by one space except "number of 724 processes currently runnable" and "total number of processes 725 in system", which are separated by a slash ('/'). Example: 726 0.61 0.61 0.55 3/828 22084 727 locks Kernel locks 728 meminfo Memory info 729 misc Miscellaneous 730 modules List of loaded modules 731 mounts Mounted filesystems 732 net Networking info (see text) 733 pagetypeinfo Additional page allocator information (see text) (2.5) 734 partitions Table of partitions known to the system 735 pci Deprecated info of PCI bus (new way -> /proc/bus/pci/, 736 decoupled by lspci (2.4) 737 rtc Real time clock 738 scsi SCSI info (see text) 739 slabinfo Slab pool info 740 softirqs softirq usage 741 stat Overall statistics 742 swaps Swap space utilization 743 sys See chapter 2 744 sysvipc Info of SysVIPC Resources (msg, sem, shm) (2.4) 745 tty Info of tty drivers 746 uptime Wall clock since boot, combined idle time of all cpus 747 version Kernel version 748 video bttv info of video resources (2.4) 749 vmallocinfo Show vmalloced areas 750 ============ =============================================================== 751 752You can, for example, check which interrupts are currently in use and what 753they are used for by looking in the file /proc/interrupts:: 754 755 > cat /proc/interrupts 756 CPU0 757 0: 8728810 XT-PIC timer 758 1: 895 XT-PIC keyboard 759 2: 0 XT-PIC cascade 760 3: 531695 XT-PIC aha152x 761 4: 2014133 XT-PIC serial 762 5: 44401 XT-PIC pcnet_cs 763 8: 2 XT-PIC rtc 764 11: 8 XT-PIC i82365 765 12: 182918 XT-PIC PS/2 Mouse 766 13: 1 XT-PIC fpu 767 14: 1232265 XT-PIC ide0 768 15: 7 XT-PIC ide1 769 NMI: 0 770 771In 2.4.* a couple of lines where added to this file LOC & ERR (this time is the 772output of a SMP machine):: 773 774 > cat /proc/interrupts 775 776 CPU0 CPU1 777 0: 1243498 1214548 IO-APIC-edge timer 778 1: 8949 8958 IO-APIC-edge keyboard 779 2: 0 0 XT-PIC cascade 780 5: 11286 10161 IO-APIC-edge soundblaster 781 8: 1 0 IO-APIC-edge rtc 782 9: 27422 27407 IO-APIC-edge 3c503 783 12: 113645 113873 IO-APIC-edge PS/2 Mouse 784 13: 0 0 XT-PIC fpu 785 14: 22491 24012 IO-APIC-edge ide0 786 15: 2183 2415 IO-APIC-edge ide1 787 17: 30564 30414 IO-APIC-level eth0 788 18: 177 164 IO-APIC-level bttv 789 NMI: 2457961 2457959 790 LOC: 2457882 2457881 791 ERR: 2155 792 793NMI is incremented in this case because every timer interrupt generates a NMI 794(Non Maskable Interrupt) which is used by the NMI Watchdog to detect lockups. 795 796LOC is the local interrupt counter of the internal APIC of every CPU. 797 798ERR is incremented in the case of errors in the IO-APIC bus (the bus that 799connects the CPUs in a SMP system. This means that an error has been detected, 800the IO-APIC automatically retry the transmission, so it should not be a big 801problem, but you should read the SMP-FAQ. 802 803In 2.6.2* /proc/interrupts was expanded again. This time the goal was for 804/proc/interrupts to display every IRQ vector in use by the system, not 805just those considered 'most important'. The new vectors are: 806 807THR 808 interrupt raised when a machine check threshold counter 809 (typically counting ECC corrected errors of memory or cache) exceeds 810 a configurable threshold. Only available on some systems. 811 812TRM 813 a thermal event interrupt occurs when a temperature threshold 814 has been exceeded for the CPU. This interrupt may also be generated 815 when the temperature drops back to normal. 816 817SPU 818 a spurious interrupt is some interrupt that was raised then lowered 819 by some IO device before it could be fully processed by the APIC. Hence 820 the APIC sees the interrupt but does not know what device it came from. 821 For this case the APIC will generate the interrupt with a IRQ vector 822 of 0xff. This might also be generated by chipset bugs. 823 824RES, CAL, TLB 825 rescheduling, call and TLB flush interrupts are 826 sent from one CPU to another per the needs of the OS. Typically, 827 their statistics are used by kernel developers and interested users to 828 determine the occurrence of interrupts of the given type. 829 830The above IRQ vectors are displayed only when relevant. For example, 831the threshold vector does not exist on x86_64 platforms. Others are 832suppressed when the system is a uniprocessor. As of this writing, only 833i386 and x86_64 platforms support the new IRQ vector displays. 834 835Of some interest is the introduction of the /proc/irq directory to 2.4. 836It could be used to set IRQ to CPU affinity. This means that you can "hook" an 837IRQ to only one CPU, or to exclude a CPU of handling IRQs. The contents of the 838irq subdir is one subdir for each IRQ, and two files; default_smp_affinity and 839prof_cpu_mask. 840 841For example:: 842 843 > ls /proc/irq/ 844 0 10 12 14 16 18 2 4 6 8 prof_cpu_mask 845 1 11 13 15 17 19 3 5 7 9 default_smp_affinity 846 > ls /proc/irq/0/ 847 smp_affinity 848 849smp_affinity is a bitmask, in which you can specify which CPUs can handle the 850IRQ. You can set it by doing:: 851 852 > echo 1 > /proc/irq/10/smp_affinity 853 854This means that only the first CPU will handle the IRQ, but you can also echo 8555 which means that only the first and third CPU can handle the IRQ. 856 857The contents of each smp_affinity file is the same by default:: 858 859 > cat /proc/irq/0/smp_affinity 860 ffffffff 861 862There is an alternate interface, smp_affinity_list which allows specifying 863a CPU range instead of a bitmask:: 864 865 > cat /proc/irq/0/smp_affinity_list 866 1024-1031 867 868The default_smp_affinity mask applies to all non-active IRQs, which are the 869IRQs which have not yet been allocated/activated, and hence which lack a 870/proc/irq/[0-9]* directory. 871 872The node file on an SMP system shows the node to which the device using the IRQ 873reports itself as being attached. This hardware locality information does not 874include information about any possible driver locality preference. 875 876prof_cpu_mask specifies which CPUs are to be profiled by the system wide 877profiler. Default value is ffffffff (all CPUs if there are only 32 of them). 878 879The way IRQs are routed is handled by the IO-APIC, and it's Round Robin 880between all the CPUs which are allowed to handle it. As usual the kernel has 881more info than you and does a better job than you, so the defaults are the 882best choice for almost everyone. [Note this applies only to those IO-APIC's 883that support "Round Robin" interrupt distribution.] 884 885There are three more important subdirectories in /proc: net, scsi, and sys. 886The general rule is that the contents, or even the existence of these 887directories, depend on your kernel configuration. If SCSI is not enabled, the 888directory scsi may not exist. The same is true with the net, which is there 889only when networking support is present in the running kernel. 890 891The slabinfo file gives information about memory usage at the slab level. 892Linux uses slab pools for memory management above page level in version 2.2. 893Commonly used objects have their own slab pool (such as network buffers, 894directory cache, and so on). 895 896:: 897 898 > cat /proc/buddyinfo 899 900 Node 0, zone DMA 0 4 5 4 4 3 ... 901 Node 0, zone Normal 1 0 0 1 101 8 ... 902 Node 0, zone HighMem 2 0 0 1 1 0 ... 903 904External fragmentation is a problem under some workloads, and buddyinfo is a 905useful tool for helping diagnose these problems. Buddyinfo will give you a 906clue as to how big an area you can safely allocate, or why a previous 907allocation failed. 908 909Each column represents the number of pages of a certain order which are 910available. In this case, there are 0 chunks of 2^0*PAGE_SIZE available in 911ZONE_DMA, 4 chunks of 2^1*PAGE_SIZE in ZONE_DMA, 101 chunks of 2^4*PAGE_SIZE 912available in ZONE_NORMAL, etc... 913 914More information relevant to external fragmentation can be found in 915pagetypeinfo:: 916 917 > cat /proc/pagetypeinfo 918 Page block order: 9 919 Pages per block: 512 920 921 Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 922 Node 0, zone DMA, type Unmovable 0 0 0 1 1 1 1 1 1 1 0 923 Node 0, zone DMA, type Reclaimable 0 0 0 0 0 0 0 0 0 0 0 924 Node 0, zone DMA, type Movable 1 1 2 1 2 1 1 0 1 0 2 925 Node 0, zone DMA, type Reserve 0 0 0 0 0 0 0 0 0 1 0 926 Node 0, zone DMA, type Isolate 0 0 0 0 0 0 0 0 0 0 0 927 Node 0, zone DMA32, type Unmovable 103 54 77 1 1 1 11 8 7 1 9 928 Node 0, zone DMA32, type Reclaimable 0 0 2 1 0 0 0 0 1 0 0 929 Node 0, zone DMA32, type Movable 169 152 113 91 77 54 39 13 6 1 452 930 Node 0, zone DMA32, type Reserve 1 2 2 2 2 0 1 1 1 1 0 931 Node 0, zone DMA32, type Isolate 0 0 0 0 0 0 0 0 0 0 0 932 933 Number of blocks type Unmovable Reclaimable Movable Reserve Isolate 934 Node 0, zone DMA 2 0 5 1 0 935 Node 0, zone DMA32 41 6 967 2 0 936 937Fragmentation avoidance in the kernel works by grouping pages of different 938migrate types into the same contiguous regions of memory called page blocks. 939A page block is typically the size of the default hugepage size, e.g. 2MB on 940X86-64. By keeping pages grouped based on their ability to move, the kernel 941can reclaim pages within a page block to satisfy a high-order allocation. 942 943The pagetypinfo begins with information on the size of a page block. It 944then gives the same type of information as buddyinfo except broken down 945by migrate-type and finishes with details on how many page blocks of each 946type exist. 947 948If min_free_kbytes has been tuned correctly (recommendations made by hugeadm 949from libhugetlbfs https://github.com/libhugetlbfs/libhugetlbfs/), one can 950make an estimate of the likely number of huge pages that can be allocated 951at a given point in time. All the "Movable" blocks should be allocatable 952unless memory has been mlock()'d. Some of the Reclaimable blocks should 953also be allocatable although a lot of filesystem metadata may have to be 954reclaimed to achieve this. 955 956 957allocinfo 958~~~~~~~~~ 959 960Provides information about memory allocations at all locations in the code 961base. Each allocation in the code is identified by its source file, line 962number, module (if originates from a loadable module) and the function calling 963the allocation. The number of bytes allocated and number of calls at each 964location are reported. 965 966Example output. 967 968:: 969 970 > sort -rn /proc/allocinfo 971 127664128 31168 mm/page_ext.c:270 func:alloc_page_ext 972 56373248 4737 mm/slub.c:2259 func:alloc_slab_page 973 14880768 3633 mm/readahead.c:247 func:page_cache_ra_unbounded 974 14417920 3520 mm/mm_init.c:2530 func:alloc_large_system_hash 975 13377536 234 block/blk-mq.c:3421 func:blk_mq_alloc_rqs 976 11718656 2861 mm/filemap.c:1919 func:__filemap_get_folio 977 9192960 2800 kernel/fork.c:307 func:alloc_thread_stack_node 978 4206592 4 net/netfilter/nf_conntrack_core.c:2567 func:nf_ct_alloc_hashtable 979 4136960 1010 drivers/staging/ctagmod/ctagmod.c:20 [ctagmod] func:ctagmod_start 980 3940352 962 mm/memory.c:4214 func:alloc_anon_folio 981 2894464 22613 fs/kernfs/dir.c:615 func:__kernfs_new_node 982 ... 983 984 985meminfo 986~~~~~~~ 987 988Provides information about distribution and utilization of memory. This 989varies by architecture and compile options. Some of the counters reported 990here overlap. The memory reported by the non overlapping counters may not 991add up to the overall memory usage and the difference for some workloads 992can be substantial. In many cases there are other means to find out 993additional memory using subsystem specific interfaces, for instance 994/proc/net/sockstat for TCP memory allocations. 995 996Example output. You may not have all of these fields. 997 998:: 999 1000 > cat /proc/meminfo 1001 1002 MemTotal: 32858820 kB 1003 MemFree: 21001236 kB 1004 MemAvailable: 27214312 kB 1005 Buffers: 581092 kB 1006 Cached: 5587612 kB 1007 SwapCached: 0 kB 1008 Active: 3237152 kB 1009 Inactive: 7586256 kB 1010 Active(anon): 94064 kB 1011 Inactive(anon): 4570616 kB 1012 Active(file): 3143088 kB 1013 Inactive(file): 3015640 kB 1014 Unevictable: 0 kB 1015 Mlocked: 0 kB 1016 SwapTotal: 0 kB 1017 SwapFree: 0 kB 1018 Zswap: 1904 kB 1019 Zswapped: 7792 kB 1020 Dirty: 12 kB 1021 Writeback: 0 kB 1022 AnonPages: 4654780 kB 1023 Mapped: 266244 kB 1024 Shmem: 9976 kB 1025 KReclaimable: 517708 kB 1026 Slab: 660044 kB 1027 SReclaimable: 517708 kB 1028 SUnreclaim: 142336 kB 1029 KernelStack: 11168 kB 1030 PageTables: 20540 kB 1031 SecPageTables: 0 kB 1032 NFS_Unstable: 0 kB 1033 Bounce: 0 kB 1034 WritebackTmp: 0 kB 1035 CommitLimit: 16429408 kB 1036 Committed_AS: 7715148 kB 1037 VmallocTotal: 34359738367 kB 1038 VmallocUsed: 40444 kB 1039 VmallocChunk: 0 kB 1040 Percpu: 29312 kB 1041 EarlyMemtestBad: 0 kB 1042 HardwareCorrupted: 0 kB 1043 AnonHugePages: 4149248 kB 1044 ShmemHugePages: 0 kB 1045 ShmemPmdMapped: 0 kB 1046 FileHugePages: 0 kB 1047 FilePmdMapped: 0 kB 1048 CmaTotal: 0 kB 1049 CmaFree: 0 kB 1050 HugePages_Total: 0 1051 HugePages_Free: 0 1052 HugePages_Rsvd: 0 1053 HugePages_Surp: 0 1054 Hugepagesize: 2048 kB 1055 Hugetlb: 0 kB 1056 DirectMap4k: 401152 kB 1057 DirectMap2M: 10008576 kB 1058 DirectMap1G: 24117248 kB 1059 1060MemTotal 1061 Total usable RAM (i.e. physical RAM minus a few reserved 1062 bits and the kernel binary code) 1063MemFree 1064 Total free RAM. On highmem systems, the sum of LowFree+HighFree 1065MemAvailable 1066 An estimate of how much memory is available for starting new 1067 applications, without swapping. Calculated from MemFree, 1068 SReclaimable, the size of the file LRU lists, and the low 1069 watermarks in each zone. 1070 The estimate takes into account that the system needs some 1071 page cache to function well, and that not all reclaimable 1072 slab will be reclaimable, due to items being in use. The 1073 impact of those factors will vary from system to system. 1074Buffers 1075 Relatively temporary storage for raw disk blocks 1076 shouldn't get tremendously large (20MB or so) 1077Cached 1078 In-memory cache for files read from the disk (the 1079 pagecache) as well as tmpfs & shmem. 1080 Doesn't include SwapCached. 1081SwapCached 1082 Memory that once was swapped out, is swapped back in but 1083 still also is in the swapfile (if memory is needed it 1084 doesn't need to be swapped out AGAIN because it is already 1085 in the swapfile. This saves I/O) 1086Active 1087 Memory that has been used more recently and usually not 1088 reclaimed unless absolutely necessary. 1089Inactive 1090 Memory which has been less recently used. It is more 1091 eligible to be reclaimed for other purposes 1092Unevictable 1093 Memory allocated for userspace which cannot be reclaimed, such 1094 as mlocked pages, ramfs backing pages, secret memfd pages etc. 1095Mlocked 1096 Memory locked with mlock(). 1097HighTotal, HighFree 1098 Highmem is all memory above ~860MB of physical memory. 1099 Highmem areas are for use by userspace programs, or 1100 for the pagecache. The kernel must use tricks to access 1101 this memory, making it slower to access than lowmem. 1102LowTotal, LowFree 1103 Lowmem is memory which can be used for everything that 1104 highmem can be used for, but it is also available for the 1105 kernel's use for its own data structures. Among many 1106 other things, it is where everything from the Slab is 1107 allocated. Bad things happen when you're out of lowmem. 1108SwapTotal 1109 total amount of swap space available 1110SwapFree 1111 Memory which has been evicted from RAM, and is temporarily 1112 on the disk 1113Zswap 1114 Memory consumed by the zswap backend (compressed size) 1115Zswapped 1116 Amount of anonymous memory stored in zswap (original size) 1117Dirty 1118 Memory which is waiting to get written back to the disk 1119Writeback 1120 Memory which is actively being written back to the disk 1121AnonPages 1122 Non-file backed pages mapped into userspace page tables 1123Mapped 1124 files which have been mmapped, such as libraries 1125Shmem 1126 Total memory used by shared memory (shmem) and tmpfs 1127KReclaimable 1128 Kernel allocations that the kernel will attempt to reclaim 1129 under memory pressure. Includes SReclaimable (below), and other 1130 direct allocations with a shrinker. 1131Slab 1132 in-kernel data structures cache 1133SReclaimable 1134 Part of Slab, that might be reclaimed, such as caches 1135SUnreclaim 1136 Part of Slab, that cannot be reclaimed on memory pressure 1137KernelStack 1138 Memory consumed by the kernel stacks of all tasks 1139PageTables 1140 Memory consumed by userspace page tables 1141SecPageTables 1142 Memory consumed by secondary page tables, this currently includes 1143 KVM mmu and IOMMU allocations on x86 and arm64. 1144NFS_Unstable 1145 Always zero. Previous counted pages which had been written to 1146 the server, but has not been committed to stable storage. 1147Bounce 1148 Memory used for block device "bounce buffers" 1149WritebackTmp 1150 Memory used by FUSE for temporary writeback buffers 1151CommitLimit 1152 Based on the overcommit ratio ('vm.overcommit_ratio'), 1153 this is the total amount of memory currently available to 1154 be allocated on the system. This limit is only adhered to 1155 if strict overcommit accounting is enabled (mode 2 in 1156 'vm.overcommit_memory'). 1157 1158 The CommitLimit is calculated with the following formula:: 1159 1160 CommitLimit = ([total RAM pages] - [total huge TLB pages]) * 1161 overcommit_ratio / 100 + [total swap pages] 1162 1163 For example, on a system with 1G of physical RAM and 7G 1164 of swap with a `vm.overcommit_ratio` of 30 it would 1165 yield a CommitLimit of 7.3G. 1166 1167 For more details, see the memory overcommit documentation 1168 in mm/overcommit-accounting. 1169Committed_AS 1170 The amount of memory presently allocated on the system. 1171 The committed memory is a sum of all of the memory which 1172 has been allocated by processes, even if it has not been 1173 "used" by them as of yet. A process which malloc()'s 1G 1174 of memory, but only touches 300M of it will show up as 1175 using 1G. This 1G is memory which has been "committed" to 1176 by the VM and can be used at any time by the allocating 1177 application. With strict overcommit enabled on the system 1178 (mode 2 in 'vm.overcommit_memory'), allocations which would 1179 exceed the CommitLimit (detailed above) will not be permitted. 1180 This is useful if one needs to guarantee that processes will 1181 not fail due to lack of memory once that memory has been 1182 successfully allocated. 1183VmallocTotal 1184 total size of vmalloc virtual address space 1185VmallocUsed 1186 amount of vmalloc area which is used 1187VmallocChunk 1188 largest contiguous block of vmalloc area which is free 1189Percpu 1190 Memory allocated to the percpu allocator used to back percpu 1191 allocations. This stat excludes the cost of metadata. 1192EarlyMemtestBad 1193 The amount of RAM/memory in kB, that was identified as corrupted 1194 by early memtest. If memtest was not run, this field will not 1195 be displayed at all. Size is never rounded down to 0 kB. 1196 That means if 0 kB is reported, you can safely assume 1197 there was at least one pass of memtest and none of the passes 1198 found a single faulty byte of RAM. 1199HardwareCorrupted 1200 The amount of RAM/memory in KB, the kernel identifies as 1201 corrupted. 1202AnonHugePages 1203 Non-file backed huge pages mapped into userspace page tables 1204ShmemHugePages 1205 Memory used by shared memory (shmem) and tmpfs allocated 1206 with huge pages 1207ShmemPmdMapped 1208 Shared memory mapped into userspace with huge pages 1209FileHugePages 1210 Memory used for filesystem data (page cache) allocated 1211 with huge pages 1212FilePmdMapped 1213 Page cache mapped into userspace with huge pages 1214CmaTotal 1215 Memory reserved for the Contiguous Memory Allocator (CMA) 1216CmaFree 1217 Free remaining memory in the CMA reserves 1218HugePages_Total, HugePages_Free, HugePages_Rsvd, HugePages_Surp, Hugepagesize, Hugetlb 1219 See Documentation/admin-guide/mm/hugetlbpage.rst. 1220DirectMap4k, DirectMap2M, DirectMap1G 1221 Breakdown of page table sizes used in the kernel's 1222 identity mapping of RAM 1223 1224vmallocinfo 1225~~~~~~~~~~~ 1226 1227Provides information about vmalloced/vmaped areas. One line per area, 1228containing the virtual address range of the area, size in bytes, 1229caller information of the creator, and optional information depending 1230on the kind of area: 1231 1232 ========== =================================================== 1233 pages=nr number of pages 1234 phys=addr if a physical address was specified 1235 ioremap I/O mapping (ioremap() and friends) 1236 vmalloc vmalloc() area 1237 vmap vmap()ed pages 1238 user VM_USERMAP area 1239 vpages buffer for pages pointers was vmalloced (huge area) 1240 N<node>=nr (Only on NUMA kernels) 1241 Number of pages allocated on memory node <node> 1242 ========== =================================================== 1243 1244:: 1245 1246 > cat /proc/vmallocinfo 1247 0xffffc20000000000-0xffffc20000201000 2101248 alloc_large_system_hash+0x204 ... 1248 /0x2c0 pages=512 vmalloc N0=128 N1=128 N2=128 N3=128 1249 0xffffc20000201000-0xffffc20000302000 1052672 alloc_large_system_hash+0x204 ... 1250 /0x2c0 pages=256 vmalloc N0=64 N1=64 N2=64 N3=64 1251 0xffffc20000302000-0xffffc20000304000 8192 acpi_tb_verify_table+0x21/0x4f... 1252 phys=7fee8000 ioremap 1253 0xffffc20000304000-0xffffc20000307000 12288 acpi_tb_verify_table+0x21/0x4f... 1254 phys=7fee7000 ioremap 1255 0xffffc2000031d000-0xffffc2000031f000 8192 init_vdso_vars+0x112/0x210 1256 0xffffc2000031f000-0xffffc2000032b000 49152 cramfs_uncompress_init+0x2e ... 1257 /0x80 pages=11 vmalloc N0=3 N1=3 N2=2 N3=3 1258 0xffffc2000033a000-0xffffc2000033d000 12288 sys_swapon+0x640/0xac0 ... 1259 pages=2 vmalloc N1=2 1260 0xffffc20000347000-0xffffc2000034c000 20480 xt_alloc_table_info+0xfe ... 1261 /0x130 [x_tables] pages=4 vmalloc N0=4 1262 0xffffffffa0000000-0xffffffffa000f000 61440 sys_init_module+0xc27/0x1d00 ... 1263 pages=14 vmalloc N2=14 1264 0xffffffffa000f000-0xffffffffa0014000 20480 sys_init_module+0xc27/0x1d00 ... 1265 pages=4 vmalloc N1=4 1266 0xffffffffa0014000-0xffffffffa0017000 12288 sys_init_module+0xc27/0x1d00 ... 1267 pages=2 vmalloc N1=2 1268 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... 1269 pages=10 vmalloc N0=10 1270 1271 1272softirqs 1273~~~~~~~~ 1274 1275Provides counts of softirq handlers serviced since boot time, for each CPU. 1276 1277:: 1278 1279 > cat /proc/softirqs 1280 CPU0 CPU1 CPU2 CPU3 1281 HI: 0 0 0 0 1282 TIMER: 27166 27120 27097 27034 1283 NET_TX: 0 0 0 17 1284 NET_RX: 42 0 0 39 1285 BLOCK: 0 0 107 1121 1286 TASKLET: 0 0 0 290 1287 SCHED: 27035 26983 26971 26746 1288 HRTIMER: 0 0 0 0 1289 RCU: 1678 1769 2178 2250 1290 12911.3 Networking info in /proc/net 1292-------------------------------- 1293 1294The subdirectory /proc/net follows the usual pattern. Table 1-8 shows the 1295additional values you get for IP version 6 if you configure the kernel to 1296support this. Table 1-9 lists the files and their meaning. 1297 1298 1299.. table:: Table 1-8: IPv6 info in /proc/net 1300 1301 ========== ===================================================== 1302 File Content 1303 ========== ===================================================== 1304 udp6 UDP sockets (IPv6) 1305 tcp6 TCP sockets (IPv6) 1306 raw6 Raw device statistics (IPv6) 1307 igmp6 IP multicast addresses, which this host joined (IPv6) 1308 if_inet6 List of IPv6 interface addresses 1309 ipv6_route Kernel routing table for IPv6 1310 rt6_stats Global IPv6 routing tables statistics 1311 sockstat6 Socket statistics (IPv6) 1312 snmp6 Snmp data (IPv6) 1313 ========== ===================================================== 1314 1315.. table:: Table 1-9: Network info in /proc/net 1316 1317 ============= ================================================================ 1318 File Content 1319 ============= ================================================================ 1320 arp Kernel ARP table 1321 dev network devices with statistics 1322 dev_mcast the Layer2 multicast groups a device is listening too 1323 (interface index, label, number of references, number of bound 1324 addresses). 1325 dev_stat network device status 1326 ip_fwchains Firewall chain linkage 1327 ip_fwnames Firewall chain names 1328 ip_masq Directory containing the masquerading tables 1329 ip_masquerade Major masquerading table 1330 netstat Network statistics 1331 raw raw device statistics 1332 route Kernel routing table 1333 rpc Directory containing rpc info 1334 rt_cache Routing cache 1335 snmp SNMP data 1336 sockstat Socket statistics 1337 softnet_stat Per-CPU incoming packets queues statistics of online CPUs 1338 tcp TCP sockets 1339 udp UDP sockets 1340 unix UNIX domain sockets 1341 wireless Wireless interface data (Wavelan etc) 1342 igmp IP multicast addresses, which this host joined 1343 psched Global packet scheduler parameters. 1344 netlink List of PF_NETLINK sockets 1345 ip_mr_vifs List of multicast virtual interfaces 1346 ip_mr_cache List of multicast routing cache 1347 ============= ================================================================ 1348 1349You can use this information to see which network devices are available in 1350your system and how much traffic was routed over those devices:: 1351 1352 > cat /proc/net/dev 1353 Inter-|Receive |[... 1354 face |bytes packets errs drop fifo frame compressed multicast|[... 1355 lo: 908188 5596 0 0 0 0 0 0 [... 1356 ppp0:15475140 20721 410 0 0 410 0 0 [... 1357 eth0: 614530 7085 0 0 0 0 0 1 [... 1358 1359 ...] Transmit 1360 ...] bytes packets errs drop fifo colls carrier compressed 1361 ...] 908188 5596 0 0 0 0 0 0 1362 ...] 1375103 17405 0 0 0 0 0 0 1363 ...] 1703981 5535 0 0 0 3 0 0 1364 1365In addition, each Channel Bond interface has its own directory. For 1366example, the bond0 device will have a directory called /proc/net/bond0/. 1367It will contain information that is specific to that bond, such as the 1368current slaves of the bond, the link status of the slaves, and how 1369many times the slaves link has failed. 1370 13711.4 SCSI info 1372------------- 1373 1374If you have a SCSI or ATA host adapter in your system, you'll find a 1375subdirectory named after the driver for this adapter in /proc/scsi. 1376You'll also see a list of all recognized SCSI devices in /proc/scsi:: 1377 1378 >cat /proc/scsi/scsi 1379 Attached devices: 1380 Host: scsi0 Channel: 00 Id: 00 Lun: 00 1381 Vendor: IBM Model: DGHS09U Rev: 03E0 1382 Type: Direct-Access ANSI SCSI revision: 03 1383 Host: scsi0 Channel: 00 Id: 06 Lun: 00 1384 Vendor: PIONEER Model: CD-ROM DR-U06S Rev: 1.04 1385 Type: CD-ROM ANSI SCSI revision: 02 1386 1387 1388The directory named after the driver has one file for each adapter found in 1389the system. These files contain information about the controller, including 1390the used IRQ and the IO address range. The amount of information shown is 1391dependent on the adapter you use. The example shows the output for an Adaptec 1392AHA-2940 SCSI adapter:: 1393 1394 > cat /proc/scsi/aic7xxx/0 1395 1396 Adaptec AIC7xxx driver version: 5.1.19/3.2.4 1397 Compile Options: 1398 TCQ Enabled By Default : Disabled 1399 AIC7XXX_PROC_STATS : Disabled 1400 AIC7XXX_RESET_DELAY : 5 1401 Adapter Configuration: 1402 SCSI Adapter: Adaptec AHA-294X Ultra SCSI host adapter 1403 Ultra Wide Controller 1404 PCI MMAPed I/O Base: 0xeb001000 1405 Adapter SEEPROM Config: SEEPROM found and used. 1406 Adaptec SCSI BIOS: Enabled 1407 IRQ: 10 1408 SCBs: Active 0, Max Active 2, 1409 Allocated 15, HW 16, Page 255 1410 Interrupts: 160328 1411 BIOS Control Word: 0x18b6 1412 Adapter Control Word: 0x005b 1413 Extended Translation: Enabled 1414 Disconnect Enable Flags: 0xffff 1415 Ultra Enable Flags: 0x0001 1416 Tag Queue Enable Flags: 0x0000 1417 Ordered Queue Tag Flags: 0x0000 1418 Default Tag Queue Depth: 8 1419 Tagged Queue By Device array for aic7xxx host instance 0: 1420 {255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255} 1421 Actual queue depth per device for aic7xxx host instance 0: 1422 {1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1} 1423 Statistics: 1424 (scsi0:0:0:0) 1425 Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8 1426 Transinfo settings: current(12/8/1/0), goal(12/8/1/0), user(12/15/1/0) 1427 Total transfers 160151 (74577 reads and 85574 writes) 1428 (scsi0:0:6:0) 1429 Device using Narrow/Sync transfers at 5.0 MByte/sec, offset 15 1430 Transinfo settings: current(50/15/0/0), goal(50/15/0/0), user(50/15/0/0) 1431 Total transfers 0 (0 reads and 0 writes) 1432 1433 14341.5 Parallel port info in /proc/parport 1435--------------------------------------- 1436 1437The directory /proc/parport contains information about the parallel ports of 1438your system. It has one subdirectory for each port, named after the port 1439number (0,1,2,...). 1440 1441These directories contain the four files shown in Table 1-10. 1442 1443 1444.. table:: Table 1-10: Files in /proc/parport 1445 1446 ========= ==================================================================== 1447 File Content 1448 ========= ==================================================================== 1449 autoprobe Any IEEE-1284 device ID information that has been acquired. 1450 devices list of the device drivers using that port. A + will appear by the 1451 name of the device currently using the port (it might not appear 1452 against any). 1453 hardware Parallel port's base address, IRQ line and DMA channel. 1454 irq IRQ that parport is using for that port. This is in a separate 1455 file to allow you to alter it by writing a new value in (IRQ 1456 number or none). 1457 ========= ==================================================================== 1458 14591.6 TTY info in /proc/tty 1460------------------------- 1461 1462Information about the available and actually used tty's can be found in the 1463directory /proc/tty. You'll find entries for drivers and line disciplines in 1464this directory, as shown in Table 1-11. 1465 1466 1467.. table:: Table 1-11: Files in /proc/tty 1468 1469 ============= ============================================== 1470 File Content 1471 ============= ============================================== 1472 drivers list of drivers and their usage 1473 ldiscs registered line disciplines 1474 driver/serial usage statistic and status of single tty lines 1475 ============= ============================================== 1476 1477To see which tty's are currently in use, you can simply look into the file 1478/proc/tty/drivers:: 1479 1480 > cat /proc/tty/drivers 1481 pty_slave /dev/pts 136 0-255 pty:slave 1482 pty_master /dev/ptm 128 0-255 pty:master 1483 pty_slave /dev/ttyp 3 0-255 pty:slave 1484 pty_master /dev/pty 2 0-255 pty:master 1485 serial /dev/cua 5 64-67 serial:callout 1486 serial /dev/ttyS 4 64-67 serial 1487 /dev/tty0 /dev/tty0 4 0 system:vtmaster 1488 /dev/ptmx /dev/ptmx 5 2 system 1489 /dev/console /dev/console 5 1 system:console 1490 /dev/tty /dev/tty 5 0 system:/dev/tty 1491 unknown /dev/tty 4 1-63 console 1492 1493 14941.7 Miscellaneous kernel statistics in /proc/stat 1495------------------------------------------------- 1496 1497Various pieces of information about kernel activity are available in the 1498/proc/stat file. All of the numbers reported in this file are aggregates 1499since the system first booted. For a quick look, simply cat the file:: 1500 1501 > cat /proc/stat 1502 cpu 237902850 368826709 106375398 1873517540 1135548 0 14507935 0 0 0 1503 cpu0 60045249 91891769 26331539 468411416 495718 0 5739640 0 0 0 1504 cpu1 59746288 91759249 26609887 468860630 312281 0 4384817 0 0 0 1505 cpu2 59489247 92985423 26904446 467808813 171668 0 2268998 0 0 0 1506 cpu3 58622065 92190267 26529524 468436680 155879 0 2114478 0 0 0 1507 intr 8688370575 8 3373 0 0 0 0 0 0 1 40791 0 0 353317 0 0 0 0 224789828 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 190974333 41958554 123983334 43 0 224593 0 0 0 <more 0's deleted> 1508 ctxt 22848221062 1509 btime 1605316999 1510 processes 746787147 1511 procs_running 2 1512 procs_blocked 0 1513 softirq 12121874454 100099120 3938138295 127375644 2795979 187870761 0 173808342 3072582055 52608 224184354 1514 1515The very first "cpu" line aggregates the numbers in all of the other "cpuN" 1516lines. These numbers identify the amount of time the CPU has spent performing 1517different kinds of work. Time units are in USER_HZ (typically hundredths of a 1518second). The meanings of the columns are as follows, from left to right: 1519 1520- user: normal processes executing in user mode 1521- nice: niced processes executing in user mode 1522- system: processes executing in kernel mode 1523- idle: twiddling thumbs 1524- iowait: In a word, iowait stands for waiting for I/O to complete. But there 1525 are several problems: 1526 1527 1. CPU will not wait for I/O to complete, iowait is the time that a task is 1528 waiting for I/O to complete. When CPU goes into idle state for 1529 outstanding task I/O, another task will be scheduled on this CPU. 1530 2. In a multi-core CPU, the task waiting for I/O to complete is not running 1531 on any CPU, so the iowait of each CPU is difficult to calculate. 1532 3. The value of iowait field in /proc/stat will decrease in certain 1533 conditions. 1534 1535 So, the iowait is not reliable by reading from /proc/stat. 1536- irq: servicing interrupts 1537- softirq: servicing softirqs 1538- steal: involuntary wait 1539- guest: running a normal guest 1540- guest_nice: running a niced guest 1541 1542The "intr" line gives counts of interrupts serviced since boot time, for each 1543of the possible system interrupts. The first column is the total of all 1544interrupts serviced including unnumbered architecture specific interrupts; 1545each subsequent column is the total for that particular numbered interrupt. 1546Unnumbered interrupts are not shown, only summed into the total. 1547 1548The "ctxt" line gives the total number of context switches across all CPUs. 1549 1550The "btime" line gives the time at which the system booted, in seconds since 1551the Unix epoch. 1552 1553The "processes" line gives the number of processes and threads created, which 1554includes (but is not limited to) those created by calls to the fork() and 1555clone() system calls. 1556 1557The "procs_running" line gives the total number of threads that are 1558running or ready to run (i.e., the total number of runnable threads). 1559 1560The "procs_blocked" line gives the number of processes currently blocked, 1561waiting for I/O to complete. 1562 1563The "softirq" line gives counts of softirqs serviced since boot time, for each 1564of the possible system softirqs. The first column is the total of all 1565softirqs serviced; each subsequent column is the total for that particular 1566softirq. 1567 1568 15691.8 Ext4 file system parameters 1570------------------------------- 1571 1572Information about mounted ext4 file systems can be found in 1573/proc/fs/ext4. Each mounted filesystem will have a directory in 1574/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or 1575/proc/fs/ext4/sda9 or /proc/fs/ext4/dm-0). The files in each per-device 1576directory are shown in Table 1-12, below. 1577 1578.. table:: Table 1-12: Files in /proc/fs/ext4/<devname> 1579 1580 ============== ========================================================== 1581 File Content 1582 mb_groups details of multiblock allocator buddy cache of free blocks 1583 ============== ========================================================== 1584 15851.9 /proc/consoles 1586------------------- 1587Shows registered system console lines. 1588 1589To see which character device lines are currently used for the system console 1590/dev/console, you may simply look into the file /proc/consoles:: 1591 1592 > cat /proc/consoles 1593 tty0 -WU (ECp) 4:7 1594 ttyS0 -W- (Ep) 4:64 1595 1596The columns are: 1597 1598+--------------------+-------------------------------------------------------+ 1599| device | name of the device | 1600+====================+=======================================================+ 1601| operations | * R = can do read operations | 1602| | * W = can do write operations | 1603| | * U = can do unblank | 1604+--------------------+-------------------------------------------------------+ 1605| flags | * E = it is enabled | 1606| | * C = it is preferred console | 1607| | * B = it is primary boot console | 1608| | * p = it is used for printk buffer | 1609| | * b = it is not a TTY but a Braille device | 1610| | * a = it is safe to use when cpu is offline | 1611+--------------------+-------------------------------------------------------+ 1612| major:minor | major and minor number of the device separated by a | 1613| | colon | 1614+--------------------+-------------------------------------------------------+ 1615 1616Summary 1617------- 1618 1619The /proc file system serves information about the running system. It not only 1620allows access to process data but also allows you to request the kernel status 1621by reading files in the hierarchy. 1622 1623The directory structure of /proc reflects the types of information and makes 1624it easy, if not obvious, where to look for specific data. 1625 1626Chapter 2: Modifying System Parameters 1627====================================== 1628 1629In This Chapter 1630--------------- 1631 1632* Modifying kernel parameters by writing into files found in /proc/sys 1633* Exploring the files which modify certain parameters 1634* Review of the /proc/sys file tree 1635 1636------------------------------------------------------------------------------ 1637 1638A very interesting part of /proc is the directory /proc/sys. This is not only 1639a source of information, it also allows you to change parameters within the 1640kernel. Be very careful when attempting this. You can optimize your system, 1641but you can also cause it to crash. Never alter kernel parameters on a 1642production system. Set up a development machine and test to make sure that 1643everything works the way you want it to. You may have no alternative but to 1644reboot the machine once an error has been made. 1645 1646To change a value, simply echo the new value into the file. 1647You need to be root to do this. You can create your own boot script 1648to perform this every time your system boots. 1649 1650The files in /proc/sys can be used to fine tune and monitor miscellaneous and 1651general things in the operation of the Linux kernel. Since some of the files 1652can inadvertently disrupt your system, it is advisable to read both 1653documentation and source before actually making adjustments. In any case, be 1654very careful when writing to any of these files. The entries in /proc may 1655change slightly between the 2.1.* and the 2.2 kernel, so if there is any doubt 1656review the kernel documentation in the directory linux/Documentation. 1657This chapter is heavily based on the documentation included in the pre 2.2 1658kernels, and became part of it in version 2.2.1 of the Linux kernel. 1659 1660Please see: Documentation/admin-guide/sysctl/ directory for descriptions of 1661these entries. 1662 1663Summary 1664------- 1665 1666Certain aspects of kernel behavior can be modified at runtime, without the 1667need to recompile the kernel, or even to reboot the system. The files in the 1668/proc/sys tree can not only be read, but also modified. You can use the echo 1669command to write value into these files, thereby changing the default settings 1670of the kernel. 1671 1672 1673Chapter 3: Per-process Parameters 1674================================= 1675 16763.1 /proc/<pid>/oom_adj & /proc/<pid>/oom_score_adj- Adjust the oom-killer score 1677-------------------------------------------------------------------------------- 1678 1679These files can be used to adjust the badness heuristic used to select which 1680process gets killed in out of memory (oom) conditions. 1681 1682The badness heuristic assigns a value to each candidate task ranging from 0 1683(never kill) to 1000 (always kill) to determine which process is targeted. The 1684units are roughly a proportion along that range of allowed memory the process 1685may allocate from based on an estimation of its current memory and swap use. 1686For example, if a task is using all allowed memory, its badness score will be 16871000. If it is using half of its allowed memory, its score will be 500. 1688 1689The amount of "allowed" memory depends on the context in which the oom killer 1690was called. If it is due to the memory assigned to the allocating task's cpuset 1691being exhausted, the allowed memory represents the set of mems assigned to that 1692cpuset. If it is due to a mempolicy's node(s) being exhausted, the allowed 1693memory represents the set of mempolicy nodes. If it is due to a memory 1694limit (or swap limit) being reached, the allowed memory is that configured 1695limit. Finally, if it is due to the entire system being out of memory, the 1696allowed memory represents all allocatable resources. 1697 1698The value of /proc/<pid>/oom_score_adj is added to the badness score before it 1699is used to determine which task to kill. Acceptable values range from -1000 1700(OOM_SCORE_ADJ_MIN) to +1000 (OOM_SCORE_ADJ_MAX). This allows userspace to 1701polarize the preference for oom killing either by always preferring a certain 1702task or completely disabling it. The lowest possible value, -1000, is 1703equivalent to disabling oom killing entirely for that task since it will always 1704report a badness score of 0. 1705 1706Consequently, it is very simple for userspace to define the amount of memory to 1707consider for each task. Setting a /proc/<pid>/oom_score_adj value of +500, for 1708example, is roughly equivalent to allowing the remainder of tasks sharing the 1709same system, cpuset, mempolicy, or memory controller resources to use at least 171050% more memory. A value of -500, on the other hand, would be roughly 1711equivalent to discounting 50% of the task's allowed memory from being considered 1712as scoring against the task. 1713 1714For backwards compatibility with previous kernels, /proc/<pid>/oom_adj may also 1715be used to tune the badness score. Its acceptable values range from -16 1716(OOM_ADJUST_MIN) to +15 (OOM_ADJUST_MAX) and a special value of -17 1717(OOM_DISABLE) to disable oom killing entirely for that task. Its value is 1718scaled linearly with /proc/<pid>/oom_score_adj. 1719 1720The value of /proc/<pid>/oom_score_adj may be reduced no lower than the last 1721value set by a CAP_SYS_RESOURCE process. To reduce the value any lower 1722requires CAP_SYS_RESOURCE. 1723 1724 17253.2 /proc/<pid>/oom_score - Display current oom-killer score 1726------------------------------------------------------------- 1727 1728This file can be used to check the current score used by the oom-killer for 1729any given <pid>. Use it together with /proc/<pid>/oom_score_adj to tune which 1730process should be killed in an out-of-memory situation. 1731 1732Please note that the exported value includes oom_score_adj so it is 1733effectively in range [0,2000]. 1734 1735 17363.3 /proc/<pid>/io - Display the IO accounting fields 1737------------------------------------------------------- 1738 1739This file contains IO statistics for each running process. 1740 1741Example 1742~~~~~~~ 1743 1744:: 1745 1746 test:/tmp # dd if=/dev/zero of=/tmp/test.dat & 1747 [1] 3828 1748 1749 test:/tmp # cat /proc/3828/io 1750 rchar: 323934931 1751 wchar: 323929600 1752 syscr: 632687 1753 syscw: 632675 1754 read_bytes: 0 1755 write_bytes: 323932160 1756 cancelled_write_bytes: 0 1757 1758 1759Description 1760~~~~~~~~~~~ 1761 1762rchar 1763^^^^^ 1764 1765I/O counter: chars read 1766The number of bytes which this task has caused to be read from storage. This 1767is simply the sum of bytes which this process passed to read() and pread(). 1768It includes things like tty IO and it is unaffected by whether or not actual 1769physical disk IO was required (the read might have been satisfied from 1770pagecache). 1771 1772 1773wchar 1774^^^^^ 1775 1776I/O counter: chars written 1777The number of bytes which this task has caused, or shall cause to be written 1778to disk. Similar caveats apply here as with rchar. 1779 1780 1781syscr 1782^^^^^ 1783 1784I/O counter: read syscalls 1785Attempt to count the number of read I/O operations, i.e. syscalls like read() 1786and pread(). 1787 1788 1789syscw 1790^^^^^ 1791 1792I/O counter: write syscalls 1793Attempt to count the number of write I/O operations, i.e. syscalls like 1794write() and pwrite(). 1795 1796 1797read_bytes 1798^^^^^^^^^^ 1799 1800I/O counter: bytes read 1801Attempt to count the number of bytes which this process really did cause to 1802be fetched from the storage layer. Done at the submit_bio() level, so it is 1803accurate for block-backed filesystems. <please add status regarding NFS and 1804CIFS at a later time> 1805 1806 1807write_bytes 1808^^^^^^^^^^^ 1809 1810I/O counter: bytes written 1811Attempt to count the number of bytes which this process caused to be sent to 1812the storage layer. This is done at page-dirtying time. 1813 1814 1815cancelled_write_bytes 1816^^^^^^^^^^^^^^^^^^^^^ 1817 1818The big inaccuracy here is truncate. If a process writes 1MB to a file and 1819then deletes the file, it will in fact perform no writeout. But it will have 1820been accounted as having caused 1MB of write. 1821In other words: The number of bytes which this process caused to not happen, 1822by truncating pagecache. A task can cause "negative" IO too. If this task 1823truncates some dirty pagecache, some IO which another task has been accounted 1824for (in its write_bytes) will not be happening. We _could_ just subtract that 1825from the truncating task's write_bytes, but there is information loss in doing 1826that. 1827 1828 1829.. Note:: 1830 1831 At its current implementation state, this is a bit racy on 32-bit machines: 1832 if process A reads process B's /proc/pid/io while process B is updating one 1833 of those 64-bit counters, process A could see an intermediate result. 1834 1835 1836More information about this can be found within the taskstats documentation in 1837Documentation/accounting. 1838 18393.4 /proc/<pid>/coredump_filter - Core dump filtering settings 1840--------------------------------------------------------------- 1841When a process is dumped, all anonymous memory is written to a core file as 1842long as the size of the core file isn't limited. But sometimes we don't want 1843to dump some memory segments, for example, huge shared memory or DAX. 1844Conversely, sometimes we want to save file-backed memory segments into a core 1845file, not only the individual files. 1846 1847/proc/<pid>/coredump_filter allows you to customize which memory segments 1848will be dumped when the <pid> process is dumped. coredump_filter is a bitmask 1849of memory types. If a bit of the bitmask is set, memory segments of the 1850corresponding memory type are dumped, otherwise they are not dumped. 1851 1852The following 9 memory types are supported: 1853 1854 - (bit 0) anonymous private memory 1855 - (bit 1) anonymous shared memory 1856 - (bit 2) file-backed private memory 1857 - (bit 3) file-backed shared memory 1858 - (bit 4) ELF header pages in file-backed private memory areas (it is 1859 effective only if the bit 2 is cleared) 1860 - (bit 5) hugetlb private memory 1861 - (bit 6) hugetlb shared memory 1862 - (bit 7) DAX private memory 1863 - (bit 8) DAX shared memory 1864 1865 Note that MMIO pages such as frame buffer are never dumped and vDSO pages 1866 are always dumped regardless of the bitmask status. 1867 1868 Note that bits 0-4 don't affect hugetlb or DAX memory. hugetlb memory is 1869 only affected by bit 5-6, and DAX is only affected by bits 7-8. 1870 1871The default value of coredump_filter is 0x33; this means all anonymous memory 1872segments, ELF header pages and hugetlb private memory are dumped. 1873 1874If you don't want to dump all shared memory segments attached to pid 1234, 1875write 0x31 to the process's proc file:: 1876 1877 $ echo 0x31 > /proc/1234/coredump_filter 1878 1879When a new process is created, the process inherits the bitmask status from its 1880parent. It is useful to set up coredump_filter before the program runs. 1881For example:: 1882 1883 $ echo 0x7 > /proc/self/coredump_filter 1884 $ ./some_program 1885 18863.5 /proc/<pid>/mountinfo - Information about mounts 1887-------------------------------------------------------- 1888 1889This file contains lines of the form:: 1890 1891 36 35 98:0 /mnt1 /mnt2 rw,noatime master:1 - ext3 /dev/root rw,errors=continue 1892 (1)(2)(3) (4) (5) (6) (n…m) (m+1)(m+2) (m+3) (m+4) 1893 1894 (1) mount ID: unique identifier of the mount (may be reused after umount) 1895 (2) parent ID: ID of parent (or of self for the top of the mount tree) 1896 (3) major:minor: value of st_dev for files on filesystem 1897 (4) root: root of the mount within the filesystem 1898 (5) mount point: mount point relative to the process's root 1899 (6) mount options: per mount options 1900 (n…m) optional fields: zero or more fields of the form "tag[:value]" 1901 (m+1) separator: marks the end of the optional fields 1902 (m+2) filesystem type: name of filesystem of the form "type[.subtype]" 1903 (m+3) mount source: filesystem specific information or "none" 1904 (m+4) super options: per super block options 1905 1906Parsers should ignore all unrecognised optional fields. Currently the 1907possible optional fields are: 1908 1909================ ============================================================== 1910shared:X mount is shared in peer group X 1911master:X mount is slave to peer group X 1912propagate_from:X mount is slave and receives propagation from peer group X [#]_ 1913unbindable mount is unbindable 1914================ ============================================================== 1915 1916.. [#] X is the closest dominant peer group under the process's root. If 1917 X is the immediate master of the mount, or if there's no dominant peer 1918 group under the same root, then only the "master:X" field is present 1919 and not the "propagate_from:X" field. 1920 1921For more information on mount propagation see: 1922 1923 Documentation/filesystems/sharedsubtree.rst 1924 1925 19263.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm 1927-------------------------------------------------------- 1928These files provide a method to access a task's comm value. It also allows for 1929a task to set its own or one of its thread siblings comm value. The comm value 1930is limited in size compared to the cmdline value, so writing anything longer 1931then the kernel's TASK_COMM_LEN (currently 16 chars, including the NUL 1932terminator) will result in a truncated comm value. 1933 1934 19353.7 /proc/<pid>/task/<tid>/children - Information about task children 1936------------------------------------------------------------------------- 1937This file provides a fast way to retrieve first level children pids 1938of a task pointed by <pid>/<tid> pair. The format is a space separated 1939stream of pids. 1940 1941Note the "first level" here -- if a child has its own children they will 1942not be listed here; one needs to read /proc/<children-pid>/task/<tid>/children 1943to obtain the descendants. 1944 1945Since this interface is intended to be fast and cheap it doesn't 1946guarantee to provide precise results and some children might be 1947skipped, especially if they've exited right after we printed their 1948pids, so one needs to either stop or freeze processes being inspected 1949if precise results are needed. 1950 1951 19523.8 /proc/<pid>/fdinfo/<fd> - Information about opened file 1953--------------------------------------------------------------- 1954This file provides information associated with an opened file. The regular 1955files have at least four fields -- 'pos', 'flags', 'mnt_id' and 'ino'. 1956The 'pos' represents the current offset of the opened file in decimal 1957form [see lseek(2) for details], 'flags' denotes the octal O_xxx mask the 1958file has been created with [see open(2) for details] and 'mnt_id' represents 1959mount ID of the file system containing the opened file [see 3.5 1960/proc/<pid>/mountinfo for details]. 'ino' represents the inode number of 1961the file. 1962 1963A typical output is:: 1964 1965 pos: 0 1966 flags: 0100002 1967 mnt_id: 19 1968 ino: 63107 1969 1970All locks associated with a file descriptor are shown in its fdinfo too:: 1971 1972 lock: 1: FLOCK ADVISORY WRITE 359 00:13:11691 0 EOF 1973 1974The files such as eventfd, fsnotify, signalfd, epoll among the regular pos/flags 1975pair provide additional information particular to the objects they represent. 1976 1977Eventfd files 1978~~~~~~~~~~~~~ 1979 1980:: 1981 1982 pos: 0 1983 flags: 04002 1984 mnt_id: 9 1985 ino: 63107 1986 eventfd-count: 5a 1987 1988where 'eventfd-count' is hex value of a counter. 1989 1990Signalfd files 1991~~~~~~~~~~~~~~ 1992 1993:: 1994 1995 pos: 0 1996 flags: 04002 1997 mnt_id: 9 1998 ino: 63107 1999 sigmask: 0000000000000200 2000 2001where 'sigmask' is hex value of the signal mask associated 2002with a file. 2003 2004Epoll files 2005~~~~~~~~~~~ 2006 2007:: 2008 2009 pos: 0 2010 flags: 02 2011 mnt_id: 9 2012 ino: 63107 2013 tfd: 5 events: 1d data: ffffffffffffffff pos:0 ino:61af sdev:7 2014 2015where 'tfd' is a target file descriptor number in decimal form, 2016'events' is events mask being watched and the 'data' is data 2017associated with a target [see epoll(7) for more details]. 2018 2019The 'pos' is current offset of the target file in decimal form 2020[see lseek(2)], 'ino' and 'sdev' are inode and device numbers 2021where target file resides, all in hex format. 2022 2023Fsnotify files 2024~~~~~~~~~~~~~~ 2025For inotify files the format is the following:: 2026 2027 pos: 0 2028 flags: 02000000 2029 mnt_id: 9 2030 ino: 63107 2031 inotify wd:3 ino:9e7e sdev:800013 mask:800afce ignored_mask:0 fhandle-bytes:8 fhandle-type:1 f_handle:7e9e0000640d1b6d 2032 2033where 'wd' is a watch descriptor in decimal form, i.e. a target file 2034descriptor number, 'ino' and 'sdev' are inode and device where the 2035target file resides and the 'mask' is the mask of events, all in hex 2036form [see inotify(7) for more details]. 2037 2038If the kernel was built with exportfs support, the path to the target 2039file is encoded as a file handle. The file handle is provided by three 2040fields 'fhandle-bytes', 'fhandle-type' and 'f_handle', all in hex 2041format. 2042 2043If the kernel is built without exportfs support the file handle won't be 2044printed out. 2045 2046If there is no inotify mark attached yet the 'inotify' line will be omitted. 2047 2048For fanotify files the format is:: 2049 2050 pos: 0 2051 flags: 02 2052 mnt_id: 9 2053 ino: 63107 2054 fanotify flags:10 event-flags:0 2055 fanotify mnt_id:12 mflags:40 mask:38 ignored_mask:40000003 2056 fanotify ino:4f969 sdev:800013 mflags:0 mask:3b ignored_mask:40000000 fhandle-bytes:8 fhandle-type:1 f_handle:69f90400c275b5b4 2057 2058where fanotify 'flags' and 'event-flags' are values used in fanotify_init 2059call, 'mnt_id' is the mount point identifier, 'mflags' is the value of 2060flags associated with mark which are tracked separately from events 2061mask. 'ino' and 'sdev' are target inode and device, 'mask' is the events 2062mask and 'ignored_mask' is the mask of events which are to be ignored. 2063All are in hex format. Incorporation of 'mflags', 'mask' and 'ignored_mask' 2064provide information about flags and mask used in fanotify_mark 2065call [see fsnotify manpage for details]. 2066 2067While the first three lines are mandatory and always printed, the rest is 2068optional and may be omitted if no marks created yet. 2069 2070Timerfd files 2071~~~~~~~~~~~~~ 2072 2073:: 2074 2075 pos: 0 2076 flags: 02 2077 mnt_id: 9 2078 ino: 63107 2079 clockid: 0 2080 ticks: 0 2081 settime flags: 01 2082 it_value: (0, 49406829) 2083 it_interval: (1, 0) 2084 2085where 'clockid' is the clock type and 'ticks' is the number of the timer expirations 2086that have occurred [see timerfd_create(2) for details]. 'settime flags' are 2087flags in octal form been used to setup the timer [see timerfd_settime(2) for 2088details]. 'it_value' is remaining time until the timer expiration. 2089'it_interval' is the interval for the timer. Note the timer might be set up 2090with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' 2091still exhibits timer's remaining time. 2092 2093DMA Buffer files 2094~~~~~~~~~~~~~~~~ 2095 2096:: 2097 2098 pos: 0 2099 flags: 04002 2100 mnt_id: 9 2101 ino: 63107 2102 size: 32768 2103 count: 2 2104 exp_name: system-heap 2105 2106where 'size' is the size of the DMA buffer in bytes. 'count' is the file count of 2107the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter. 2108 21093.9 /proc/<pid>/map_files - Information about memory mapped files 2110--------------------------------------------------------------------- 2111This directory contains symbolic links which represent memory mapped files 2112the process is maintaining. Example output:: 2113 2114 | lr-------- 1 root root 64 Jan 27 11:24 333c600000-333c620000 -> /usr/lib64/ld-2.18.so 2115 | lr-------- 1 root root 64 Jan 27 11:24 333c81f000-333c820000 -> /usr/lib64/ld-2.18.so 2116 | lr-------- 1 root root 64 Jan 27 11:24 333c820000-333c821000 -> /usr/lib64/ld-2.18.so 2117 | ... 2118 | lr-------- 1 root root 64 Jan 27 11:24 35d0421000-35d0422000 -> /usr/lib64/libselinux.so.1 2119 | lr-------- 1 root root 64 Jan 27 11:24 400000-41a000 -> /usr/bin/ls 2120 2121The name of a link represents the virtual memory bounds of a mapping, i.e. 2122vm_area_struct::vm_start-vm_area_struct::vm_end. 2123 2124The main purpose of the map_files is to retrieve a set of memory mapped 2125files in a fast way instead of parsing /proc/<pid>/maps or 2126/proc/<pid>/smaps, both of which contain many more records. At the same 2127time one can open(2) mappings from the listings of two processes and 2128comparing their inode numbers to figure out which anonymous memory areas 2129are actually shared. 2130 21313.10 /proc/<pid>/timerslack_ns - Task timerslack value 2132--------------------------------------------------------- 2133This file provides the value of the task's timerslack value in nanoseconds. 2134This value specifies an amount of time that normal timers may be deferred 2135in order to coalesce timers and avoid unnecessary wakeups. 2136 2137This allows a task's interactivity vs power consumption tradeoff to be 2138adjusted. 2139 2140Writing 0 to the file will set the task's timerslack to the default value. 2141 2142Valid values are from 0 - ULLONG_MAX 2143 2144An application setting the value must have PTRACE_MODE_ATTACH_FSCREDS level 2145permissions on the task specified to change its timerslack_ns value. 2146 21473.11 /proc/<pid>/patch_state - Livepatch patch operation state 2148----------------------------------------------------------------- 2149When CONFIG_LIVEPATCH is enabled, this file displays the value of the 2150patch state for the task. 2151 2152A value of '-1' indicates that no patch is in transition. 2153 2154A value of '0' indicates that a patch is in transition and the task is 2155unpatched. If the patch is being enabled, then the task hasn't been 2156patched yet. If the patch is being disabled, then the task has already 2157been unpatched. 2158 2159A value of '1' indicates that a patch is in transition and the task is 2160patched. If the patch is being enabled, then the task has already been 2161patched. If the patch is being disabled, then the task hasn't been 2162unpatched yet. 2163 21643.12 /proc/<pid>/arch_status - task architecture specific status 2165------------------------------------------------------------------- 2166When CONFIG_PROC_PID_ARCH_STATUS is enabled, this file displays the 2167architecture specific status of the task. 2168 2169Example 2170~~~~~~~ 2171 2172:: 2173 2174 $ cat /proc/6753/arch_status 2175 AVX512_elapsed_ms: 8 2176 2177Description 2178~~~~~~~~~~~ 2179 2180x86 specific entries 2181~~~~~~~~~~~~~~~~~~~~~ 2182 2183AVX512_elapsed_ms 2184^^^^^^^^^^^^^^^^^^ 2185 2186 If AVX512 is supported on the machine, this entry shows the milliseconds 2187 elapsed since the last time AVX512 usage was recorded. The recording 2188 happens on a best effort basis when a task is scheduled out. This means 2189 that the value depends on two factors: 2190 2191 1) The time which the task spent on the CPU without being scheduled 2192 out. With CPU isolation and a single runnable task this can take 2193 several seconds. 2194 2195 2) The time since the task was scheduled out last. Depending on the 2196 reason for being scheduled out (time slice exhausted, syscall ...) 2197 this can be arbitrary long time. 2198 2199 As a consequence the value cannot be considered precise and authoritative 2200 information. The application which uses this information has to be aware 2201 of the overall scenario on the system in order to determine whether a 2202 task is a real AVX512 user or not. Precise information can be obtained 2203 with performance counters. 2204 2205 A special value of '-1' indicates that no AVX512 usage was recorded, thus 2206 the task is unlikely an AVX512 user, but depends on the workload and the 2207 scheduling scenario, it also could be a false negative mentioned above. 2208 22093.13 /proc/<pid>/fd - List of symlinks to open files 2210------------------------------------------------------- 2211This directory contains symbolic links which represent open files 2212the process is maintaining. Example output:: 2213 2214 lr-x------ 1 root root 64 Sep 20 17:53 0 -> /dev/null 2215 l-wx------ 1 root root 64 Sep 20 17:53 1 -> /dev/null 2216 lrwx------ 1 root root 64 Sep 20 17:53 10 -> 'socket:[12539]' 2217 lrwx------ 1 root root 64 Sep 20 17:53 11 -> 'socket:[12540]' 2218 lrwx------ 1 root root 64 Sep 20 17:53 12 -> 'socket:[12542]' 2219 2220The number of open files for the process is stored in 'size' member 2221of stat() output for /proc/<pid>/fd for fast access. 2222------------------------------------------------------- 2223 2224 2225Chapter 4: Configuring procfs 2226============================= 2227 22284.1 Mount options 2229--------------------- 2230 2231The following mount options are supported: 2232 2233 ========= ======================================================== 2234 hidepid= Set /proc/<pid>/ access mode. 2235 gid= Set the group authorized to learn processes information. 2236 subset= Show only the specified subset of procfs. 2237 ========= ======================================================== 2238 2239hidepid=off or hidepid=0 means classic mode - everybody may access all 2240/proc/<pid>/ directories (default). 2241 2242hidepid=noaccess or hidepid=1 means users may not access any /proc/<pid>/ 2243directories but their own. Sensitive files like cmdline, sched*, status are now 2244protected against other users. This makes it impossible to learn whether any 2245user runs specific program (given the program doesn't reveal itself by its 2246behaviour). As an additional bonus, as /proc/<pid>/cmdline is unaccessible for 2247other users, poorly written programs passing sensitive information via program 2248arguments are now protected against local eavesdroppers. 2249 2250hidepid=invisible or hidepid=2 means hidepid=1 plus all /proc/<pid>/ will be 2251fully invisible to other users. It doesn't mean that it hides a fact whether a 2252process with a specific pid value exists (it can be learned by other means, e.g. 2253by "kill -0 $PID"), but it hides process' uid and gid, which may be learned by 2254stat()'ing /proc/<pid>/ otherwise. It greatly complicates an intruder's task of 2255gathering information about running processes, whether some daemon runs with 2256elevated privileges, whether other user runs some sensitive program, whether 2257other users run any program at all, etc. 2258 2259hidepid=ptraceable or hidepid=4 means that procfs should only contain 2260/proc/<pid>/ directories that the caller can ptrace. 2261 2262gid= defines a group authorized to learn processes information otherwise 2263prohibited by hidepid=. If you use some daemon like identd which needs to learn 2264information about processes information, just add identd to this group. 2265 2266subset=pid hides all top level files and directories in the procfs that 2267are not related to tasks. 2268 2269Chapter 5: Filesystem behavior 2270============================== 2271 2272Originally, before the advent of pid namespace, procfs was a global file 2273system. It means that there was only one procfs instance in the system. 2274 2275When pid namespace was added, a separate procfs instance was mounted in 2276each pid namespace. So, procfs mount options are global among all 2277mountpoints within the same namespace:: 2278 2279 # grep ^proc /proc/mounts 2280 proc /proc proc rw,relatime,hidepid=2 0 0 2281 2282 # strace -e mount mount -o hidepid=1 -t proc proc /tmp/proc 2283 mount("proc", "/tmp/proc", "proc", 0, "hidepid=1") = 0 2284 +++ exited with 0 +++ 2285 2286 # grep ^proc /proc/mounts 2287 proc /proc proc rw,relatime,hidepid=2 0 0 2288 proc /tmp/proc proc rw,relatime,hidepid=2 0 0 2289 2290and only after remounting procfs mount options will change at all 2291mountpoints:: 2292 2293 # mount -o remount,hidepid=1 -t proc proc /tmp/proc 2294 2295 # grep ^proc /proc/mounts 2296 proc /proc proc rw,relatime,hidepid=1 0 0 2297 proc /tmp/proc proc rw,relatime,hidepid=1 0 0 2298 2299This behavior is different from the behavior of other filesystems. 2300 2301The new procfs behavior is more like other filesystems. Each procfs mount 2302creates a new procfs instance. Mount options affect own procfs instance. 2303It means that it became possible to have several procfs instances 2304displaying tasks with different filtering options in one pid namespace:: 2305 2306 # mount -o hidepid=invisible -t proc proc /proc 2307 # mount -o hidepid=noaccess -t proc proc /tmp/proc 2308 # grep ^proc /proc/mounts 2309 proc /proc proc rw,relatime,hidepid=invisible 0 0 2310 proc /tmp/proc proc rw,relatime,hidepid=noaccess 0 0 2311